Browse the corpus

of sedatives and is thought to have different outcomes then regular hypoactive delirium. Apart from a burden for the patient, the presence of delirium may also increase the workload for ICU professionals. Typically, workload in the ICU is measured using the Simplified Therapeutic Intervention Scoring System (TISS-28). Objectives: To determine the effect of delirium status on ICU workload. Methods: A retrospective cohort study was performed at the ICU of a university medical center. All ICU patients admitted between December 2011 and July 2013 were classified for their delirium sub-type status, using the CAM-ICU. Peterson criteria were used to distinguish between the delirium subtypes, added with Patel's criterion for rapid reversible sedation-related delirium. Demographics and other relevant data relating to ICU workload were collected.

NEUROMONITORING A391 Physiologic monitoring for neurocritically-ill patients: an international survey of intensivists S. Sivakumar1, F.S. Taccone2, K.A. Desai1, C. Lazaridis1 1Baylor College of Medicine, Neurology/Neurocritical Care, Houston, TX, USA; 2Erasmus Hospital, Intensive Care Medicine, Brussels, Belgium Correspondence: C. Lazaridis - Baylor College of Medicine, Neurology/Neurocritical Care, Houston, TX, USA Introduction In addition to systemic hemodynamics, the management of neurocritically ill patients is often informed by neuromonitoring. In the absence of high-level evidence clinicians are often guided by personal and local expertise. Little is known about practices as they pertain to the use of such monitoring in patients with acute brain injury (ABI). Objectives To investigate practices in bedside monitoring for ABI patients. Particularly interested in differences among “neurointensivists” (NIs; defined here as intensivists whose clinical practice is comprised > 1/3 by neurocritical care) and other intensivists (OIs). Also, to explore patterns specific to traumatic brain injury (TBI) and subarachnoid hemorrhage (SAH), as well as preferences and availability of particular technologies/devices.

tensivists” (NIs; defined here as intensivists whose clinical practice is comprised > 1/3 by neurocritical care) and other intensivists (OIs). Also, to explore patterns specific to traumatic brain injury (TBI) and subarachnoid hemorrhage (SAH), as well as preferences and availability of particular technologies/devices. Methods Electronic survey of 22 items including two case-based scenarios; endorsed by SCCM (9,000 recipients) and ESICM (on-line newsletter) in 2013. A sample size of 370 was calculated based on a population of 10,000 physician members, a 5 % margin error, and 95 % confidence interval. We summarized results using descriptive statistics (proportions with 95 % confidence intervals). A chi-square test was used to compare proportions of responses between NIs and OIs with a significance p < 0.05.

alculated based on a population of 10,000 physician members, a 5 % margin error, and 95 % confidence interval. We summarized results using descriptive statistics (proportions with 95 % confidence intervals). A chi-square test was used to compare proportions of responses between NIs and OIs with a significance p < 0.05. Results There were 655 responders (66 % completion rate); 422(65 %) were classified as OIs and 226(35 %) as NIs. More NIs follow hemodynamic protocols for neurocritically-ill patients (56 % vs. 43 %, p 0.001), in TBI (44.5 % vs. 33.3 %, p 0.007), and in SAH (38.1 % vs. 21.3 %, p < 000.1). For delayed cerebral ischemia (DCI), more NIs target cardiac index (CI) (35 % vs. 21 %, p 0.0001), and fluid responsiveness (62 % vs. 53 %, p 0.03), use more bedside ultrasound (BUS) (42 % vs. 29 %, p 0.005) and arterial waveform analysis (40 % vs. 29 %, p 0.02). For DCI neuromonitoring, NIs use more angiography (57 % vs. 43 %, p 0.004), TCD (46 % vs. 38 %, p 0.0001), and CTP (32 % vs.16 %, p 0.0001). For CPP optimization in TBI, NIs use more arterial waveform analysis (45 % vs. 35 %, p 0.019), and BUS (37 % vs. 27.7 %, p 0.023), while more OIs monitor mixed venous oxygen saturation (54.1 % vs. 45 %, p 0.045). For TBI neuromonitoring, NIs use more PbtO2 (28 % vs. 10 %, p 0.0001). In the case scenario of raised ICP/low PbtO2, most employ analgosedation (47 %) and osmotherapy (38 %). Fewer make use of preserved pressure reactivity, particularly OIs (vasopressor use 23 % vs. 34 %, p 0.014).

aturation (54.1 % vs. 45 %, p 0.045). For TBI neuromonitoring, NIs use more PbtO2 (28 % vs. 10 %, p 0.0001). In the case scenario of raised ICP/low PbtO2, most employ analgosedation (47 %) and osmotherapy (38 %). Fewer make use of preserved pressure reactivity, particularly OIs (vasopressor use 23 % vs. 34 %, p 0.014). Conclusions There is large heterogeneity in the use of monitoring protocols, variables, and technologies/devices. “Neurointensivists” not only employ more neuromonitoring but also more hemodynamic monitoring in patients with acute brain injury. ICP/CPP remain the most commonly followed neuro-variables in TBI patients, with low use of other brain-physiology parameters, suggesting that clinicians make limited efforts to individualize these goals. A392 A prospective observational pilot study of cerebral autoregulation measured by near infrared spectroscopy (NIRS) in patients with septic shock M. Skarzynski1, M. Sekhon2, W. Henderson2, D. Griesdale2 1Centre Hospitalier Régional Orléans, Réaimation Médicale, Orléans, France; 2University of British Columbia, Vancouver, Canada Correspondence: M. Skarzynski - Centre Hospitalier Régional Orléans, Réaimation Médicale, Orléans, France Introduction Impairment of cerebral autoregulation has been proposed as a possible explanation of cognitive dysfunction in patients with septic shock. Although transcranial Doppler has previously been used to assess cerebral autoregulation, this technology can only evaluate at single points in time. In contrast, near-infrared spectroscopy offers continuous assessment of cerebral autoregulation.

sible explanation of cognitive dysfunction in patients with septic shock. Although transcranial Doppler has previously been used to assess cerebral autoregulation, this technology can only evaluate at single points in time. In contrast, near-infrared spectroscopy offers continuous assessment of cerebral autoregulation. Objectives Assess cerebral autoregulation using NIRS in patients admitted to the intensive care unit with septic shock. Methods We included 20 patients admitted with septic shock admitted to the intensive care unit (ICU) at Vancouver General Hospital (VGH). The ICU is a 31-bed mixed medical-surgical unit affiliated with the University of British Columbia. We excluded patients with acute or chronic neurological disorders, end stage liver disease, long-term dialysis, and those admitted following a cardiac arrest. We measured regional cerebral oximetry (rSO2) by NIRS (INVOS®, Covidien, Ireland) for 24 hours. NIRS and mean arterial pressure (MAP) data were collected in real time using ICM + ® brain monitoring software (Cambridge University, UK). ICM+ calculates a moving Pearson correlation coefficient (COx) between 30 consecutive, 10 second average MAP and rSO2 values. Impaired cerebral autoregulation was defined as a COx greater than 0.3. We also defined the impaired autoregulation index (IARindex) as the percentage of monitoring time spent with an impaired autoregulation. The IARindex was calculated for each 6 hours period (H0H6; H6H12;H12H18, H18H24), and for 24 hours.

SO2 values. Impaired cerebral autoregulation was defined as a COx greater than 0.3. We also defined the impaired autoregulation index (IARindex) as the percentage of monitoring time spent with an impaired autoregulation. The IARindex was calculated for each 6 hours period (H0H6; H6H12;H12H18, H18H24), and for 24 hours. Results We analyzed 19 patients, one patient being excluded from analysis due to removal for arterial line [mean (Standard deviation); median (interquartile)] age 67(12), APACHE II score 21(6) median MAP 72 [67–75] mmHg, median rSO2 64 [57–70] %, median end tidal carbon dioxide 30 [27–35] mmHg and median temperature 37.1 [36.8-37.3] °C. After removal of artefacts, the mean monitoring time was 22 h08 (8 h54). All patients had impaired cerebral autoregulation during their monitoring time. The mean IAR index was 17 (9.5) %. During H0H6 and H18H24, the majority of our patients; respectively 53 and 71 % had an IAR index > 10 %. Conclusion According to our data, patients with septic shock had impaired cerebral autoregulation within the first 24 hours of their admission in the ICU. In our patients, we described a variability of distribution of impaired autoregulation according to time. References Schramm P, Klein KU, Falkenberg L, et al. Impaired cerebrovascular autoregulation in patients with severe sepsis and sepsis-associated delirium. Crit Care 2012; 16: R181. Aries MJH, Czosnyka M, Budohoski KP, et al. Continuous determination of optimal cerebral perfusion pressure in traumatic brain injury. Crit. Care Med. 2012.Fig. 1 (abstract 392). IAR index by 6 hours period

Schramm P, Klein KU, Falkenberg L, et al. Impaired cerebrovascular autoregulation in patients with severe sepsis and sepsis-associated delirium. Crit Care 2012; 16: R181. Aries MJH, Czosnyka M, Budohoski KP, et al. Continuous determination of optimal cerebral perfusion pressure in traumatic brain injury. Crit. Care Med. 2012.Fig. 1 (abstract 392). IAR index by 6 hours period A393 Changes in muscle thickness throughout hospitalisation after traumatic brain injury L. Chapple1, A. Deane1,2, L. Williams3, R. Strickland2, K. Lange4, D. Heyland5, M. Chapman1,2 1University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia; 2Royal Adelaide Hospital, Intensive Care, Adelaide, Australia; 3Griffith University, Menzies Health Institute of Queensland, Gold Coast, Australia; 4University of Adelaide, Discipline of Medicine, Adelaide, Australia; 5Kingston General Hospital, Clinical Evaluation Research Unit, Kingston, Canada Correspondence: L. Chapple - University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia Introduction Patients with a traumatic brain injury (TBI) remain in hospital for extended periods. It is likely that muscle mass in these patients diminishes over the hospital admission, yet this has not previously been quantified. Ultrasonography may provide a useful measure of changes in muscle size over the hospital stay. Objectives To quantify changes in ultrasound-derived quadriceps muscle layer thickness (QMLT) and establish the feasibility of repeated ultrasound examinations throughout the entire hospitalisation of patients with a TBI.

A393 Changes in muscle thickness throughout hospitalisation after traumatic brain injury L. Chapple1, A. Deane1,2, L. Williams3, R. Strickland2, K. Lange4, D. Heyland5, M. Chapman1,2 1University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia; 2Royal Adelaide Hospital, Intensive Care, Adelaide, Australia; 3Griffith University, Menzies Health Institute of Queensland, Gold Coast, Australia; 4University of Adelaide, Discipline of Medicine, Adelaide, Australia; 5Kingston General Hospital, Clinical Evaluation Research Unit, Kingston, Canada Correspondence: L. Chapple - University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia Introduction Patients with a traumatic brain injury (TBI) remain in hospital for extended periods. It is likely that muscle mass in these patients diminishes over the hospital admission, yet this has not previously been quantified. Ultrasonography may provide a useful measure of changes in muscle size over the hospital stay. Objectives To quantify changes in ultrasound-derived quadriceps muscle layer thickness (QMLT) and establish the feasibility of repeated ultrasound examinations throughout the entire hospitalisation of patients with a TBI. Methods Adult patients with a moderate-severe TBI (Glasgow Coma Scale 3–12) consecutively admitted to the intensive care unit (ICU) at a single trauma referral centre over 12 months were eligible following informed consent. Ultrasounds of QMLT at the midpoint and two-thirds between the anterior superior iliac spine and top of the patella were conducted weekly during admission. Data were censored at 3-months and are mean (SD) unless otherwise stated.

t a single trauma referral centre over 12 months were eligible following informed consent. Ultrasounds of QMLT at the midpoint and two-thirds between the anterior superior iliac spine and top of the patella were conducted weekly during admission. Data were censored at 3-months and are mean (SD) unless otherwise stated. Results Thirty-three patients [45.4 (16.6) years; 88 % male; 55 % severe TBI; Trauma Injury Severity Score 0.5 (0.3); median APACHE II 18 (IQR: 14–22); admission body mass index 27.2 (6.5) kg/m2] were studied. Primary injury cause was vehicular (58 %) and 67 % of TBIs were multi-trauma. Median length of ICU and ward-based stay was 13.4 [IQR: 6.5-17.9] and 31.8 [9.4-52.4] days, respectively. At 3-months all patients were alive and four remained in hospital. A total of 123 ultrasounds were taken; 30 (24 %) in ICU and 93 (76 %) on the acute ward, equal to 3.7 ultrasounds per patient. Ultrasounds were not possible at 21 % of time-points in hospital due to patient agitation and 61 % at 3-month follow-up due to inability to attend appointment in person. Twenty-eight (85 %) patients had >1 ultrasound taken during their admission, for which the mean baseline measure was 1.78 (0.72) cm and within patient standard deviation 0.3 cm. Nineteen patients (58 %) had >2 ultrasounds during the hospital admission with a baseline measure of 1.78 (0.68) and proximate measure to discharge of 1.53 (0.52) cm. There was a 6.2 (35.8) % decrease between the baseline and proximate discharge measures. QMLT decreased over the first four weeks of hospital admission [week 1 (n = 22): 2.01 (0.84), week 2 (n = 24): 1.66 (0.59), week 3 (n = 19): 1.50 (0.57), week 4 (n = 13): 1.40 (0.38)]. At 3-months the QMLT was 2.00 (0.7) cm (n = 22), with a decrease from baseline of 3.2 (26.0) %.

eline and proximate discharge measures. QMLT decreased over the first four weeks of hospital admission [week 1 (n = 22): 2.01 (0.84), week 2 (n = 24): 1.66 (0.59), week 3 (n = 19): 1.50 (0.57), week 4 (n = 13): 1.40 (0.38)]. At 3-months the QMLT was 2.00 (0.7) cm (n = 22), with a decrease from baseline of 3.2 (26.0) %. Conclusions This is the first study to describe longitudinal change in muscle size over the hospital admission in patients with a moderate-severe TBI. Whilst technically feasible, issues with compliance and missing data and intra-individual variation between measurements are challenges when conducting regular ultrasound measures in this population. Muscle thickness reduced throughout hospitalisation, but improved towards baseline by 3-months.

h a moderate-severe TBI. Whilst technically feasible, issues with compliance and missing data and intra-individual variation between measurements are challenges when conducting regular ultrasound measures in this population. Muscle thickness reduced throughout hospitalisation, but improved towards baseline by 3-months. A394 Acute impairment of saccadic eye movement is associated with cerebral infarction after aneurysmal subarachnoid haemorrhage M.J. Rowland1,2, P. Garry1,2, J. Westbrook1,2, R. Corkill2, C.A. Antoniades1,2, K.T. Pattinson1,2 1University of Oxford, Nuffield Department of Clinical Neurosciences, Oxford, UK; 2Oxford University Hospitals NHS Foundation Trust, Neurosciences Intensive Care Unit, Oxford, UK Correspondence: M.J. Rowland - Oxford University Hospitals NHS Foundation Trust, Neurosciences Intensive Care Unit, Oxford, UK Introduction Cerebral infarction due to delayed cerebral ischaemia (DCI) remains a significant cause of morbidity and mortality following aneurysmal subarachnoid haemorrhage (SAH). Damage to the brain in the first 72 hours (“early brain injury”) is likely to play a key pathophysiological role but remains difficult to quantify objectively and non-invasively at the bedside- especially in those patients who do not require sedation or ventilation. Current diagnostic modalities used in routine clinical practice are either invasive, require ionising radiation or contrast or have a high learning curve and user variability.

ns difficult to quantify objectively and non-invasively at the bedside- especially in those patients who do not require sedation or ventilation. Current diagnostic modalities used in routine clinical practice are either invasive, require ionising radiation or contrast or have a high learning curve and user variability. Objectives We sought to determine whether saccadic eye movements are impaired following SAH and whether measurement of saccadic latency (SL) in the acute period post-aneurysm rupture is associated with the likelihood of developing DCI in patients after SAH. Methods 24 male/female patients (mean age 53.32, range 31–70) with World Federation of Neurosurgeons (WFNS) grade I and II aneurysmal SAH, treated endovascularly within 72 hours of rupture were recruited. DCI and DCI-related cerebral infarction were defined as per consensus guidelines.1 Saccadometry data was collected at three time points: in the first 72 hours, between days 5 and 10 and at three months post-SAH. Data from 10 healthy age/gender matched controls was collected on one occasion for comparison.

re recruited. DCI and DCI-related cerebral infarction were defined as per consensus guidelines.1 Saccadometry data was collected at three time points: in the first 72 hours, between days 5 and 10 and at three months post-SAH. Data from 10 healthy age/gender matched controls was collected on one occasion for comparison. Results Age-adjusted median SL in patients was significantly prolonged in the first 72 hours post-SAH when compared to controls (188.7, 95 % CI = (176.9, 202.2)ms v 160.7, 95 % CI = (145.6, 179.4)ms, p = 0.0054 t-test). By 3 months post-SAH, there was no significant difference in median SL compared to controls (188.7, 95 % CI = (176.9, 202.2)ms v 180.0, 95 % CI = (165.1, 197.8)ms, p = 0.4175 t-test). Patients diagnosed with cerebral infarction due to DCI had a significantly higher age-adjusted median SL in the first 72 hours than those without infarction (240.6, 95 % CI = (216.7, 270.3)ms v 204.1, 95 % CI = (190.7, 219.5)ms, p = 0.0157 t-test). This difference was more pronounced during days 5–10 post-SAH - the peak incidence for DCI (303.7, 95 % CI = (266.7 352.7)ms v 207.6, 95 % CI = (193.7, 223.6)ms, p < 0.0001 t-test). A binary generalised linear model showed that median SL in the first 72 hours was the only significant predictor of cerebral infarction after SAH.

ference was more pronounced during days 5–10 post-SAH - the peak incidence for DCI (303.7, 95 % CI = (266.7 352.7)ms v 207.6, 95 % CI = (193.7, 223.6)ms, p < 0.0001 t-test). A binary generalised linear model showed that median SL in the first 72 hours was the only significant predictor of cerebral infarction after SAH. Conclusions This is the first study to use saccadometry to measure eye movements in patients with SAH during the acute period post-rupture. Results show that median saccadic latency in the first 72 hours following SAH is an independent predictor of the risk of developing cerebral infarction due to DCI. Saccadometry may act as a potential objective biomarker of early brain injury to guide the need for intensive care admission and treatment. References 1. Vergouwen MDI et al. 2010, Stroke;(41):2391–2395 Grant acknowledgment MRC (UK).

Conclusions This is the first study to use saccadometry to measure eye movements in patients with SAH during the acute period post-rupture. Results show that median saccadic latency in the first 72 hours following SAH is an independent predictor of the risk of developing cerebral infarction due to DCI. Saccadometry may act as a potential objective biomarker of early brain injury to guide the need for intensive care admission and treatment. References 1. Vergouwen MDI et al. 2010, Stroke;(41):2391–2395 Grant acknowledgment MRC (UK). A395 The incidence of spreading depolarisations in ischaemic brain injury using non-invasive near-infrared spectroscopy G. Fatania1, A.J. Strong2 1King's College Hospital, Department of Neurosurgery, London, UK; 2King's College London, Institute of Psychiatry, Psychology and Neuroscience, Department of Clinical Neuroscience, London, UK Correspondence: G. Fatania - King's College Hospital, Department of Neurosurgery, London, UK Introduction Spreading depolarisations (SD) occur spontaneously in ischaemic cortex and are implicated in the evolution of the ischaemic penumbra. Subsequent enlargement of the ischaemic lesion causes worse neurological outcomes. Monitoring for SDs is solely performed through invasive electrocorticography (ECoG) which is located when a patient requires emergency surgery. SDs can cause significant haemodynamic changes which may be amenable to non-invasive monitoring.

sequent enlargement of the ischaemic lesion causes worse neurological outcomes. Monitoring for SDs is solely performed through invasive electrocorticography (ECoG) which is located when a patient requires emergency surgery. SDs can cause significant haemodynamic changes which may be amenable to non-invasive monitoring. Objectives This study seeks to develop such a technique using near-infrared spectroscopy (NIRS). NIRS allows surrogate measure of cerebral blood flow by inferring changes in the concentration of oxyhaemoglobin and deoxyhaemoglobin. Methods 5 ischaemic brain injury patients requiring emergency neurosurgery (4 retrospective, 1 prospective) recruited to the COSBID study at King's College Hospital (and monitored with ECoG) underwent concomitant NIRS monitoring. The NIRS data was analysed for significant haemodynamic response to SDs. Results In total, three ECoG-confirmed SDs occurred during NIRS monitoring, each in a different patient. Two acute (one minute) hypoperfusion responses were seen after an SD in one patient. One acute (one minute) hyperperfusion response was seen after an SD in one patient. Three subacute (20 minute) hypoperfusion responses were seen across two patients. Conclusion This preliminary data is promising that NIRS can be used to assess the haemodynamic responses secondary to SDs. More data is required so that the responses can be characterised adequately which could lead to a novel non-invasive monitoring technique.

Results In total, three ECoG-confirmed SDs occurred during NIRS monitoring, each in a different patient. Two acute (one minute) hypoperfusion responses were seen after an SD in one patient. One acute (one minute) hyperperfusion response was seen after an SD in one patient. Three subacute (20 minute) hypoperfusion responses were seen across two patients. Conclusion This preliminary data is promising that NIRS can be used to assess the haemodynamic responses secondary to SDs. More data is required so that the responses can be characterised adequately which could lead to a novel non-invasive monitoring technique. A396 Predicting intracranial pressure and brain tissue oxygen crises in patients with severe traumatic brain injury R.B. Myers1, C. Lazaridis2, C.M. Jermaine1, C.S. Robertson3, C.G. Rusin4 1Rice University, Computer Science, Houston, TX, USA; 2Baylor College of Medicine, Neurology, Neurocritical Care, Houston, TX, USA; 3Baylor College of Medicine, Neurosurgery, Houston, TX, USA; 4Baylor College of Medicine, Pediatric Cardiology, Houston, TX, USA Correspondence: C. Lazaridis - Baylor College of Medicine, Neurology, Neurocritical Care, Houston, TX, USA Introduction Two central physiologic targets in the management of severe traumatic brain injury (TBI) are the intracranial pressure (ICP) and the partial brain tissue oxygen tension (PbtO2). Intracranial pressure and PbtO2 thresholds are incorporated into step-tiered clinical protocols and guidelines however by the time treatment is provided, it may be too late. The ability to predict the onset of these “crisis” events would potentially provide clinicians with valuable time to attempt aborting the episode and/or to appropriately manage it. Combining statistical machine learning with physiologic and clinical insight allows the construction of robust quantitative models to predict future events such as elevated ICP or low PbtO2. Such models provide targets for evidence-based individualized treatment in real time.

episode and/or to appropriately manage it. Combining statistical machine learning with physiologic and clinical insight allows the construction of robust quantitative models to predict future events such as elevated ICP or low PbtO2. Such models provide targets for evidence-based individualized treatment in real time. Objectives Develop computer algorithms that can recognize physiologic patterns in TBI patients that occur in advance of intracranial pressure ICP and PbtO2 crises. The automated early detection of crisis precursors can provide clinicians with time to intervene in order to prevent or mitigate secondary brain injury. Methods A retrospective study was conducted from prospectively collected physiologic data. Intracranial pressure and PbtO2 crisis events were defined as ICP ≥ 20 mmHg lasting at least 15 minutes and PbtO2 values < 10 mmHg for at least 10 minutes, respectively. The physiologic data preceding each crisis event were used to identify precursors associated with crisis onset. Multivariate classification models were applied to recorded data in 30-minute epochs of time to predict crises between 15 and 360 minutes in the future. Our cohort consisted of 817 severe TBI subjects admitted to the neurosurgical intensive care unit of Ben Taub General Hospital in Houston, Texas. Results Our algorithm can predict the onset of an ICP crisis with 30 minutes advance warning and an AUC of 0.86 using only ICP measurements and time since last crisis. An analogous algorithm can predict the start of PbtO2 crises with 30 minutes advanced warning and an AUC of 0.91.

Methods A retrospective study was conducted from prospectively collected physiologic data. Intracranial pressure and PbtO2 crisis events were defined as ICP ≥ 20 mmHg lasting at least 15 minutes and PbtO2 values < 10 mmHg for at least 10 minutes, respectively. The physiologic data preceding each crisis event were used to identify precursors associated with crisis onset. Multivariate classification models were applied to recorded data in 30-minute epochs of time to predict crises between 15 and 360 minutes in the future. Our cohort consisted of 817 severe TBI subjects admitted to the neurosurgical intensive care unit of Ben Taub General Hospital in Houston, Texas. Results Our algorithm can predict the onset of an ICP crisis with 30 minutes advance warning and an AUC of 0.86 using only ICP measurements and time since last crisis. An analogous algorithm can predict the start of PbtO2 crises with 30 minutes advanced warning and an AUC of 0.91. Conclusions We report here novel algorithms that provide accurate and timely predictions of intracranial hypertension and tissue hypoxia crises in patients with severe TBI. Almost all of the information needed to predict the onset of these events is contained within the signal of interest and the time since last crisis. These predictive algorithms offer clinicians the opportunity to prepare and potentially prevent secondary brain injury insults. Grant acknowledgment

Conclusions We report here novel algorithms that provide accurate and timely predictions of intracranial hypertension and tissue hypoxia crises in patients with severe TBI. Almost all of the information needed to predict the onset of these events is contained within the signal of interest and the time since last crisis. These predictive algorithms offer clinicians the opportunity to prepare and potentially prevent secondary brain injury insults. Grant acknowledgment Supported by a training fellowship from the Keck Center of the Gulf Coast Consortia, on Rice University´s NLM Training Program in Biomedical Informatics (grant number T15LM007093) and by the NSF under grant number 0964526.

Conclusions We report here novel algorithms that provide accurate and timely predictions of intracranial hypertension and tissue hypoxia crises in patients with severe TBI. Almost all of the information needed to predict the onset of these events is contained within the signal of interest and the time since last crisis. These predictive algorithms offer clinicians the opportunity to prepare and potentially prevent secondary brain injury insults. Grant acknowledgment Supported by a training fellowship from the Keck Center of the Gulf Coast Consortia, on Rice University´s NLM Training Program in Biomedical Informatics (grant number T15LM007093) and by the NSF under grant number 0964526. A397 Early EEG for outcome prediction of postanoxic coma: validation of undisputable predictive value and cost-effectiveness analysis J. Hofmeijer1,2, L. Sondag2, M.C. Tjepkema-Cloostermans3, A. Beishuizen4, F.H. Bosch5, M.J.A.M. van Putten1,3 1University of Twente, Clinical Neurophysiology, Enschede, Netherlands; 2Rijnstate Hospital, Neurology, Arnhem, Netherlands; 3Medical Spectrum Twente, Clinical Neurophysiology, Enschede, Netherlands; 4Medical Spectrum Twente, Intensive Care, Enschede, Netherlands; 5Rijnstate Hospital, Intensive Care, Arnhem, Netherlands Correspondence: J. Hofmeijer - Rijnstate Hospital, Neurology, Arnhem, Netherlands Introduction Early identification of patients without potential for recovery of brain function may prevent inappropriate treatment of comatose patients after cardiac arrest. We recently showed that evolution of the EEG pattern within the first 24 hours after cardiac arrest robustly contributes to multimodal prediction of either poor or good outcome [1].

f patients without potential for recovery of brain function may prevent inappropriate treatment of comatose patients after cardiac arrest. We recently showed that evolution of the EEG pattern within the first 24 hours after cardiac arrest robustly contributes to multimodal prediction of either poor or good outcome [1]. Objectives We aim to confirm our results and present a cost-effectiveness analysis. Methods 432 consecutive comatose patients after cardiac arrest were included in a prospective cohort study on two intensive care units. Continuous EEG was measured during the first three days. EEGs were visually classified as unfavorable (isoelectric, low-voltage, burst-suppression-with-identical-bursts), intermediate, or favorable (continuous patterns), at 12, 24, 48, and 72 hours by two reviewers, independently. Outcome was dichotomized as good (CPC score 1 or 2) or poor (CPC score 3, 4, or 5) at six months. EEG parameters were related to outcome using logistic regression analysis. Cost-effectiveness in the hospital is currently estimated by decision tree analysis.

, at 12, 24, 48, and 72 hours by two reviewers, independently. Outcome was dichotomized as good (CPC score 1 or 2) or poor (CPC score 3, 4, or 5) at six months. EEG parameters were related to outcome using logistic regression analysis. Cost-effectiveness in the hospital is currently estimated by decision tree analysis. Results Poor outcome occurred 54 % of included patients. Single parameters unequivocally predicting poor outcome in the first 277 patients were an unfavorable EEG pattern at 24 hours, absent pupillary light responses at 48 hours, and absent SSEPs at 72 hours. Together, these had a specificity of 100 % and a sensitivity of 50 %. Favorable EEG patterns at 12 hours were strongly associated with good outcome. EEG beyond 24 hours had no additional predictive value. For the remaining 155 patients, EEG analyses are ongoing. Data on cost-effectiveness, based on the assumption of EEG based treatment discontinuation, will be presented. Conclusions EEG within 24 hours is a robust contributor to prediction of poor or good outcome and should be included in guidelines for treatment of comatose patients after cardiac arrest. Reference 1. Hofmeijer et al. Neurology 2015;85:137–43.

Results Poor outcome occurred 54 % of included patients. Single parameters unequivocally predicting poor outcome in the first 277 patients were an unfavorable EEG pattern at 24 hours, absent pupillary light responses at 48 hours, and absent SSEPs at 72 hours. Together, these had a specificity of 100 % and a sensitivity of 50 %. Favorable EEG patterns at 12 hours were strongly associated with good outcome. EEG beyond 24 hours had no additional predictive value. For the remaining 155 patients, EEG analyses are ongoing. Data on cost-effectiveness, based on the assumption of EEG based treatment discontinuation, will be presented. Conclusions EEG within 24 hours is a robust contributor to prediction of poor or good outcome and should be included in guidelines for treatment of comatose patients after cardiac arrest. Reference 1. Hofmeijer et al. Neurology 2015;85:137–43. A398 Cerebral energy dysfunction and hyperemia during the early brain injury phase following aneurysmal subarachnoid hemorrhage L. Carteron, C. Patet, D. Solari, M. Oddo CHUV, Department of Intensive Care Medicine, Neuroscience Critical Care Research Group, Lausanne, Switzerland Correspondence: L. Carteron - CHUV, Department of Intensive Care Medicine, Neuroscience Critical Care Research Group, Lausanne, Switzerland Introduction Mechanisms of early brain injury (EBI) following aneurysmal subarachnoid hemorrhage (SAH) are poorly understood. Using brain perfusion CT (pCT) and cerebral microdialysis (CMD), we examined the relationship of cerebral energy metabolism with brain perfusion during the EBI phase after SAH in humans.

zerland Introduction Mechanisms of early brain injury (EBI) following aneurysmal subarachnoid hemorrhage (SAH) are poorly understood. Using brain perfusion CT (pCT) and cerebral microdialysis (CMD), we examined the relationship of cerebral energy metabolism with brain perfusion during the EBI phase after SAH in humans. Methods Prospective observational cohort of poor-grade SAH patients monitored with cerebral microdialysis (CMD) who were resuscitated according to current guidelines. Data from brain pCT (performed 44 ± 25 hrs from ictus) were matched to CMD epochs displaying cerebral energy dysfunction, defined by a CMD lactate/pyruvate ratio > 40 and/or lactate > 4 mmol/L. Results A total of 19 pCT and 127 hours of CMD samples (15 patients) were analyzed. The majority (14/19) of pCT were associated with cerebral energy dysfunction, despite main cerebral physiologic variables were within normal range (intracranial pressure 14.2 ± 7.2 mmHg, PbtO2 25 ± 10 mmHg). Energy dysfunction was associated with simultaneous normal (n = 9; 28–65 mL/100 g/min) or hyperemic cerebral blood flow (n = 5; 69–85 mL/100 g/min). Energy dysfunction also correlated with concomitant pathological elevations of glutamate and glycerol, that were more pronounced in hyperemic than normal pCT (LPR 54 ± 12 vs. 42 ± 7 and glycerol 157 ± 76 vs. 95 ± 41 μmol/L, both p < 0.01; glutamate 38 ± 52 vs. 26 ± 24 μmol/L, p = 0.18, t-test for comparisons between groups).

dysfunction also correlated with concomitant pathological elevations of glutamate and glycerol, that were more pronounced in hyperemic than normal pCT (LPR 54 ± 12 vs. 42 ± 7 and glycerol 157 ± 76 vs. 95 ± 41 μmol/L, both p < 0.01; glutamate 38 ± 52 vs. 26 ± 24 μmol/L, p = 0.18, t-test for comparisons between groups). Conclusions EBI after SAH is associated with profound cerebral metabolic and cellular distress, despite normal-hyperemic brain perfusion. Our findings reveal alternative pathophysiological mechanisms to ischemia in the EBI phase of SAH in humans and support therapeutic strategies targeted to energy dysfunction in this setting. GRANTS Supported by research grants from the Swiss National Science Foundation (SNSF), the Novartis Foundation for Biomedical Research, the Société Française d'Anesthésie et de Réanimation (SFAR) and the “Fondation des Gueules Cassées”.

r 2011 and July 2013 were classified for their delirium sub-type status, using the CAM-ICU. Peterson criteria were used to distinguish between the delirium subtypes, added with Patel's criterion for rapid reversible sedation-related delirium. Demographics and other relevant data relating to ICU workload were collected. Results: A total of 3,926 ICU patients were admitted during the study period (Table 4). The age was 62 ± 15 (mean ± SD), of which 64.6 % were male. The APACHE II score was 17 ± 7, and hospital mortality was 11.6 %. Patients were ventilated for median 2 [IQR 2–3] days. The LOS-ICU was median 1 [IQR 1–3] day, and in-hospital LOS was median 5 [IQR 5–17] days. In total 881 (22 %) patients were classified as delirious of which the alternating subtype had the highest incidence. Delirium duration was median 1 [IQR 1–2] day, the dose of haloperidol administered was median 19 [IQR 9–42] mg. The overall TISS-28 score was 28 ± 7. After adjusting for APACHE II score the TISS-28 score did not differ between delirious patients and non-delirious patients, (28.4 ± 6 versus 27.2 ± 8, respectively). In addition, no differences in TISS-28 score were found between the delirium subgroups. Patients with delirium stayed significantly longer on the mechanical ventilator, in the ICU and in-hospital (all p < 0.001).

Conclusions EBI after SAH is associated with profound cerebral metabolic and cellular distress, despite normal-hyperemic brain perfusion. Our findings reveal alternative pathophysiological mechanisms to ischemia in the EBI phase of SAH in humans and support therapeutic strategies targeted to energy dysfunction in this setting. GRANTS Supported by research grants from the Swiss National Science Foundation (SNSF), the Novartis Foundation for Biomedical Research, the Société Française d'Anesthésie et de Réanimation (SFAR) and the “Fondation des Gueules Cassées”. A399 Optic nerve sheath diameter evaluated by transorbital sonography in healthy volunteers from Pakistan M.A. Ali Aga Khan University Hospital, Anaesthesiology, Karachi, Pakistan Introduction Raised intracranial pressure is a common manifestation of severe brain injury. Rapid diagnosis and timely intervention is required to prevent secondary brain damage and death.An accurate, reliable, noninvasive, point-of-care monitoring device to identify presence of intracranial hypertension would be helpful in situations where there is clinical suspicion for intracranial hypertension. Bedside ocular ultrasound is an emerging noninvasive technique to measure optic nerve sheath diameter. Knowledge of the normal range of optic nerve sheath diameter in a healthy population is essential to interpret this measurement as a marker of raised intracranial pressure in clinical practice. Objectives To evaluate normal optic nerve sheath diameter in healthy volunteers in Pakistan.

A399 Optic nerve sheath diameter evaluated by transorbital sonography in healthy volunteers from Pakistan M.A. Ali Aga Khan University Hospital, Anaesthesiology, Karachi, Pakistan Introduction Raised intracranial pressure is a common manifestation of severe brain injury. Rapid diagnosis and timely intervention is required to prevent secondary brain damage and death.An accurate, reliable, noninvasive, point-of-care monitoring device to identify presence of intracranial hypertension would be helpful in situations where there is clinical suspicion for intracranial hypertension. Bedside ocular ultrasound is an emerging noninvasive technique to measure optic nerve sheath diameter. Knowledge of the normal range of optic nerve sheath diameter in a healthy population is essential to interpret this measurement as a marker of raised intracranial pressure in clinical practice. Objectives To evaluate normal optic nerve sheath diameter in healthy volunteers in Pakistan. Methods Hundred healthy volunteers of Pakistani origin, aged more than 18 years were recruited in the study. The ultrasound probe was placed on the superior and lateral aspect of the orbit against the upper eyelid with the eye closed. For each subject, the primary investigator performed three measurements on each eye. The measurements of each eye were then averaged to yield a mean optic nerve sheath diameter (ONSD). Results are presented as mean ± standard deviation (SD). Statistical analysis was performed with SPSS software version 19. Mann Whitney U test was used to compare unpaired variables between genders and Wilcoxon matched pairs signed rank test to compare left and right eyes.

eld a mean optic nerve sheath diameter (ONSD). Results are presented as mean ± standard deviation (SD). Statistical analysis was performed with SPSS software version 19. Mann Whitney U test was used to compare unpaired variables between genders and Wilcoxon matched pairs signed rank test to compare left and right eyes. Results The median ONSD of right eye was 4.84 mm and 95 % of individuals had mean ONSD in the range 4.84-4.97 mm while the median ONSD of left eye was 4.86 mm and 95 % of individuals had mean ONSD in the range 4.85-4.96 mm. There was no difference among the 3 repeated measures of ONSD in each eye. There was no relationship between ONSD with age, gender and measurement taken between left and right eyes. Conclusions 95 % of healthy Pakistani adults have an ONSD less than 4.82 mm. ONSD more than 4.82 mm in this population should be considered abnormal and may reflect raised intracranial pressure. Note: This abstract has been previously published and is available at [3]. It is included here as a complete record of the abstracts from the conference. References 1. Brain Trauma Foundation, American Association of Neurological Surgeons, Congress of Neurological Surgeons, Joint Section on Neurotrauma, Critical Care, AANS/CNS. Guidelines for the management of severe traumatic brain injury. VI. Indications for intracranial pressure monitoring. J Neurotrauma. 2007; 24(Suppl 1):S37-44.

1. Brain Trauma Foundation, American Association of Neurological Surgeons, Congress of Neurological Surgeons, Joint Section on Neurotrauma, Critical Care, AANS/CNS. Guidelines for the management of severe traumatic brain injury. VI. Indications for intracranial pressure monitoring. J Neurotrauma. 2007; 24(Suppl 1):S37-44. 2. Morgenstern LB, Hemphill JC III, Anderson C, Becker K, Broderick JP, Connolly ES Jr, Greenberg SM, Huang JN, Mac-Donald RL, Messe´ SR, Mitchell PH, Selim M, Tamargo RJ, American Heart Association Stroke Council and Council on Cardiovascular Nursing. Guidelines for the management of spontaneous intracerebral hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2010;41:2108–29. 3. Ashgar A, Hasmi M, Hussain A (2015) Optic nerve sheath diameter evaluated by transorbital sonography in health volunteers from Pakistan. Anaesthesia, Pain & Intensive Care 19(3):p282.

2. Morgenstern LB, Hemphill JC III, Anderson C, Becker K, Broderick JP, Connolly ES Jr, Greenberg SM, Huang JN, Mac-Donald RL, Messe´ SR, Mitchell PH, Selim M, Tamargo RJ, American Heart Association Stroke Council and Council on Cardiovascular Nursing. Guidelines for the management of spontaneous intracerebral hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2010;41:2108–29. 3. Ashgar A, Hasmi M, Hussain A (2015) Optic nerve sheath diameter evaluated by transorbital sonography in health volunteers from Pakistan. Anaesthesia, Pain & Intensive Care 19(3):p282. A400 Heart rate variability and multimodal brain monitoring before and after decompressive craniectomy in traumatic brain injury C. Dias1,2, R. Almeida3,4,5, A. Vaz-Ferreira6, J. Silva6, E. Monteiro1,2, A. Cerejo7,8, A.P. Rocha4,9 1Centro Hospitalar São João, Intensive Care, Porto, Portugal; 2Faculty of Medicine, University of Porto, Medicine, Porto, Portugal; 3Faculdade de Ciências, Universidade do Porto, Departamento de Matemática, Porto, Portugal; 4Centro de Matemática, Universidade do Porto, Porto, Portugal; 5The Biomedical Research Networking Center in Bioengineering, Biomaterials and Nanomedicine, Zaragoza, Spain; 6Centro Hospitalar São João, Porto, Portugal; 7Centro Hospitalar São João, Neurosurgery, Porto, Portugal; 8Faculty of Medicine, University of Porto, Clinical Neurosciences, Porto, Portugal; 9Faculdade de Ciências, Universidade do Porto, Porto, Portugal Correspondence: C. Dias - Faculty of Medicine, University of Porto, Medicine, Porto, Portugal Introduction Autonomic control and cerebral autoregulation (CAR) may be disturbed after traumatic brain injury (TBI) and become severely impaired due to intracranial hypertension (IH). Decompressive craniectomy (DC) after TBI is a tiered therapy for patients with IH, but the links between autonomic derangement, intracranial pressure (ICP), impaired cerebral autoregulation and outcome remain poorly explored.

aumatic brain injury (TBI) and become severely impaired due to intracranial hypertension (IH). Decompressive craniectomy (DC) after TBI is a tiered therapy for patients with IH, but the links between autonomic derangement, intracranial pressure (ICP), impaired cerebral autoregulation and outcome remain poorly explored. Objectives We aimed to evaluate the relationship between heart rate variability (HRV), as a surrogate of autonomic control, ICP and CAR before and after DC. Methods We retrospectively studied 9 adult TBI patients, admitted to the Neurocritical Care Unit at Hospital São João, Porto that were submitted to primary or secondary DC and had completed monitoring records 12 h before and 24 h after surgery. Patients were monitored with continuous ECG, ICP, cerebral perfusion pressure, cerebral autoregulation with pressure reactivity and pressure-volume compensatory reserve index (RAP). Automatic delineation of ECG signal was performed using a wavelet-based approach and HRV analysis in time and frequency domain was done according to the guidelines for both periods selected.

rebral perfusion pressure, cerebral autoregulation with pressure reactivity and pressure-volume compensatory reserve index (RAP). Automatic delineation of ECG signal was performed using a wavelet-based approach and HRV analysis in time and frequency domain was done according to the guidelines for both periods selected. Results A total of 48 consecutive TBI patients submitted to DC were screened, but we had to exclude 39 patients because of incomplete monitoring data. The median age was 27 (IQR13), 7 were men, median GCS was 6 (IQR3), median GOS was 4 (IQR2) and hospital mortality was 22 %. DC led to a significantly decrease in ICP maximum value (p = 0.011), CPP (p = 0.038) and RAP (p = 0.008) after surgery. There were no significant differences between values of PRx and HRV variables both in time and frequency domain. However, we found statistically significant correlations between HRV nonparametric variables and multimodal brain monitoring variables (Fig. 2). Before DC normalized low frequency (LFn), high frequency (HFn) and LF/HF ratio are positively correlated with ICP maximum (ICPmax) value, (respectively Rs = 0.5,p = 0.000; Rs = 0.3,p = 0.005;Rs = 0.4,p = 0.0001). After DC the correlation becomes negative but non-significant. Total power (Tp) has a positive and significant correlation with CPP before DC (Rs = 0.5,p = 0.000). RAP is negatively correlated with Tp and HF after DC (Rs = −0.4,p = 0.000;Rs = −0.5,p = 0.000). PRx increases after DC and we could not find any important correlation with HRV.

tion becomes negative but non-significant. Total power (Tp) has a positive and significant correlation with CPP before DC (Rs = 0.5,p = 0.000). RAP is negatively correlated with Tp and HF after DC (Rs = −0.4,p = 0.000;Rs = −0.5,p = 0.000). PRx increases after DC and we could not find any important correlation with HRV. Conclusions Further studies are warranted to better understand the brain changes and autonomic distress associated with high intracranial pressure and decompressive craniectomy. References 1. HRV: Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Soc. of Cardiology and the North American Soc. of Pacing and Electrophysiology. Eur Heart J 1996;17(3):354–81. 2. Sykora M, et al. Autonomic Impairment in Severe TBI: A Multimodal Neuromonitoring Study. Crit Care Med 2016. 3. Kolias AG, et al. Decompressive craniectomy following TBI: developing the evidence base. Br J Neurosurg 2016;30(2):246–50.Fig. 2 (abstract A400). Monitoring time frame analysis before and after DC Fig. 3 (abstract A400). Multimodal Brain Monitoring before and afterDC

2. Sykora M, et al. Autonomic Impairment in Severe TBI: A Multimodal Neuromonitoring Study. Crit Care Med 2016. 3. Kolias AG, et al. Decompressive craniectomy following TBI: developing the evidence base. Br J Neurosurg 2016;30(2):246–50.Fig. 2 (abstract A400). Monitoring time frame analysis before and after DC Fig. 3 (abstract A400). Multimodal Brain Monitoring before and afterDC A401 Value of noninvasive ultrasonographic techniques in assessing increased intracranial pressure in patients with moderate to severe traumatic brain injury A.A. Elsayed1, A.M. Abougabal2, B.N. Beshey1, K.M. Alzahaby1 1Faculty of Medicine, Alexandria University, Critical Care Medicine Department, Alexandria, Egypt; 2Faculty of Medicine, Alexandria University, Radiology Department, Alexandria, Egypt Correspondence: K.M. Alzahaby - Faculty of Medicine, Alexandria University, Critical Care Medicine Department, Alexandria, Egypt Introduction: The idea of a non-invasive method of measuring intracranial pressure (ICP) is captivating, as disadvantages seen in relation to the invasive methods of ICP measuring, that is, hemorrhage, infection, and costly expenses are avoidable. Objectives: Determine the value of ultrasonographic transcranial Doppler (TCD) and optic nerve sheath diameter(ONSD)in assessing increased ICP and outcome of moderate to severe traumatic brain injury patients.

A401 Value of noninvasive ultrasonographic techniques in assessing increased intracranial pressure in patients with moderate to severe traumatic brain injury A.A. Elsayed1, A.M. Abougabal2, B.N. Beshey1, K.M. Alzahaby1 1Faculty of Medicine, Alexandria University, Critical Care Medicine Department, Alexandria, Egypt; 2Faculty of Medicine, Alexandria University, Radiology Department, Alexandria, Egypt Correspondence: K.M. Alzahaby - Faculty of Medicine, Alexandria University, Critical Care Medicine Department, Alexandria, Egypt Introduction: The idea of a non-invasive method of measuring intracranial pressure (ICP) is captivating, as disadvantages seen in relation to the invasive methods of ICP measuring, that is, hemorrhage, infection, and costly expenses are avoidable. Objectives: Determine the value of ultrasonographic transcranial Doppler (TCD) and optic nerve sheath diameter(ONSD)in assessing increased ICP and outcome of moderate to severe traumatic brain injury patients. Methods: 40 patients with moderate to severe traumatic brain injury(GCS ≤ 13) admitted to a university teaching hospital were enrolled. ONSD and TCD measurements were performed daily for 7 days. TCD was performed on both middle cerebral arteries(MCA), recording peak systolic (sFV), end diastolic(dFV)and mean time-averaged(mFV)blood flow velocities in cm/s. The Gosling and King's pulsatility index(PI) was calculated:PI = sFV-dFV/mFV. Marshall and Rotterdam head CT neuroimaging scales were recorded on admission, 48 hours and 5 to 7 days later. Glasgow Outcome Scale(GOS)was assessed six months after discharge for survivors.

n time-averaged(mFV)blood flow velocities in cm/s. The Gosling and King's pulsatility index(PI) was calculated:PI = sFV-dFV/mFV. Marshall and Rotterdam head CT neuroimaging scales were recorded on admission, 48 hours and 5 to 7 days later. Glasgow Outcome Scale(GOS)was assessed six months after discharge for survivors. Results: The PI significantly increased in days 2 and 3 then it showed a non-significant decrease in day 4 and 5, all compared to baseline value. This was followed by a significant decrease in days 6 and 7. Comparing the PI's daily reading with the previous day was significant, increasing from day 1 to day 2, and from day 2 to day 3 then decreasing from day 3 till day 7(p < 0.001). The PI showed a significant direct correlation the Marshall and Rotterdam Scales on days 3 and 7, and Rotterdam Scale on day 7. A significant direct correlation was demonstrated between the ICU LOS and the PI determined on days 4 to7. No significant correlation was found between the mean ONSD and the PI measured on days 1 to 5, but a significant direct correlation was demonstrated between the two parameters on day 6 and day 7. The sFV, dFV and mFV were significantly higher on days 2 to 4 in survivors than in non-survivors. While the PI on day 2 and 3 was significantly lower in survivors than in non-survivors. On day 2, a mFV of ≤27.31 cm/sec was 94.44 % accurate in detecting non-survivors. On day 3, a mFV of ≤6.62 cm/sec was 100 % accurate in detecting non-survivors. On day 2, a PI of >1.26 was 89.47 % accurate in detecting non-survivors. On day 3, a PI of >1.87 was 100 % accurate in detecting non-survivors.

non-survivors. On day 2, a mFV of ≤27.31 cm/sec was 94.44 % accurate in detecting non-survivors. On day 3, a mFV of ≤6.62 cm/sec was 100 % accurate in detecting non-survivors. On day 2, a PI of >1.26 was 89.47 % accurate in detecting non-survivors. On day 3, a PI of >1.87 was 100 % accurate in detecting non-survivors. Conclusions: There was a significant day-to-day changes in the ONSD and PI in TBI patients making serial estimation of clinically/radiologically diagnosed raised ICP possible in the absence of invasive ICP measurement.The correlation between PI and CT imaging of the head together with its cost-effectiveness, safety to the patient and permanent availability makes the sonographic approach highly beneficial. TCD derived parameters predicted survival in TBI patients.TCD derived PI correlated directly with ICU length of stay.Fig. 4 (abstract A401). ROC curve for PI to determine its ability to predi Fig. 5 (abstract A401). ROC curve for mFV to determine its ability to pred Fig. 6 (abstract A401). Correlation between ONSD and PI on day 7

Conclusions: There was a significant day-to-day changes in the ONSD and PI in TBI patients making serial estimation of clinically/radiologically diagnosed raised ICP possible in the absence of invasive ICP measurement.The correlation between PI and CT imaging of the head together with its cost-effectiveness, safety to the patient and permanent availability makes the sonographic approach highly beneficial. TCD derived parameters predicted survival in TBI patients.TCD derived PI correlated directly with ICU length of stay.Fig. 4 (abstract A401). ROC curve for PI to determine its ability to predi Fig. 5 (abstract A401). ROC curve for mFV to determine its ability to pred Fig. 6 (abstract A401). Correlation between ONSD and PI on day 7 A402 Cerebral near-infrared spectroscopy in adults on veno-arterial extracorporeal membrane oxygenation S. Pozzebon, A. Blandino Ortiz, S. Cristallini, O. Lheureux, A. Brasseur, J.-L. Vincent, J. Creteur, F.S. Taccone Erasme University Hospital, Université Libre de Bruxelles, Department of Intensive Care, Brussels, Belgium Correspondence: S. Pozzebon - Erasme University Hospital, Université Libre de Bruxelles, Department of Intensive Care, Brussels, Belgium Introduction: Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is increasingly used to treat severe cardio-pulmonary failure. Among the possible complications, central nervous system (CNS) events, such as stroke or hemorrhage, are associated with long-term neurologic morbidity and increased mortality. Additionally, peripheral VA-ECMO can result in delivery of hypoxic blood to the brain. The use of cerebral near-infrared spectroscopy (NIRS) can detect cerebral hypoperfusion non-invasively.

us system (CNS) events, such as stroke or hemorrhage, are associated with long-term neurologic morbidity and increased mortality. Additionally, peripheral VA-ECMO can result in delivery of hypoxic blood to the brain. The use of cerebral near-infrared spectroscopy (NIRS) can detect cerebral hypoperfusion non-invasively. Methods: We reviewed our institutional VA-ECMO database (n = 159) from November 2008 to December 2015 and identified those patients who were monitored with cerebral NIRS. Sensors were placed on the patients´ foreheads using the Foresight device (CAS Medical Systems, Inc, Branford, USA). Regional saturation (rSO2) was recorded and analysed by calculating the time below different thresholds (60 %, 55 % and 50 %).

er 2015 and identified those patients who were monitored with cerebral NIRS. Sensors were placed on the patients´ foreheads using the Foresight device (CAS Medical Systems, Inc, Branford, USA). Regional saturation (rSO2) was recorded and analysed by calculating the time below different thresholds (60 %, 55 % and 50 %). Results: A total of 39 patients (age: 54 [48–72] years) had cerebral NIRS monitoring during VA-ECMO for cardiogenic shock (n = 19), refractory cardiac arrest (n = 14) and post heart/lung transplantation (n = 6). ECMO was applied for 6 [3–10] days and NIRS monitoring for 3 [2–4] days. Nine (23 %) patients developed an ischemic stroke (8/9 in the anterior circulation) and 6 (15 %) differential hypoxia -lower PaO2 in the upper body than in the lower body, because of normal cardiac output with severe impairment of pulmonary function- during the first 3 days of monitoring. Hospital mortality was 22/39 (56 %). Twenty-seven (69 %) patients had a drop of rSO2 < 60 % for at least 5 % of the NIRS monitoring period, resulting in hemodynamic interventions, which involved increasing pressure, oxygenation, and/or ECMO flow. In the 3/6 patients with differential hypoxia, these interventions were unsuccessful and a veno-arterial-venous ECMO was implemented. Patients developing ischemic stroke had a higher differential right-left rSO2 (11[6–12]% vs. 6 [4–7]%; p = 0.004) than others. Survivors had a shorter rSO2 time < 60 % than non-survivors (6.6[1.2-24.5]% vs. 43 [12–60]%; p = 0.007).

, these interventions were unsuccessful and a veno-arterial-venous ECMO was implemented. Patients developing ischemic stroke had a higher differential right-left rSO2 (11[6–12]% vs. 6 [4–7]%; p = 0.004) than others. Survivors had a shorter rSO2 time < 60 % than non-survivors (6.6[1.2-24.5]% vs. 43 [12–60]%; p = 0.007). Conclusions: Cerebral NIRS monitoring may be helpful in patients undergoing VA-ECMO to detect cerebrovascular events or brain hypoperfusion/reduced oxygen delivery. In these patients cerebral hypoperfusion is associated with a poor outcome. A403 Cerebral regional tissue oxygenation in patients with neurocardiac injury after subarachnoid hemorrhage M. Hravnak1, K. Yousef1, Y. Chang2, E. Crago1, R.M. Friedlander2 1University of Pittsburgh, School of Nursing; Department of Acute & Tertiary Care, Pittsburgh, Pennsylvania, PA, USA; 2University of Pittsburgh, School of Medicine: Department of Neurosurgery, Pittsburgh, PA, USA Correspondence: M. Hravnak - University of Pittsburgh, School of Nursing; Department of Acute & Tertiary Care, Pittsburgh, Pennsylvania, PA, USA Introduction: Although it has been demonstrated previously that patients with aneurysmal subarachnoid hemorrhage (aSAH) may also experience neurocardiac injury, the impact of this complication on cerebral and peripheral tissue perfusion is not well understood.

of Acute & Tertiary Care, Pittsburgh, Pennsylvania, PA, USA Introduction: Although it has been demonstrated previously that patients with aneurysmal subarachnoid hemorrhage (aSAH) may also experience neurocardiac injury, the impact of this complication on cerebral and peripheral tissue perfusion is not well understood. Objectives: We aimed to determine if neurocardiac injury as defined by increased levels of cardiac troponin I (cTnI) was associated with changes in regional cerebral (CrSO2) and peripheral (PrSO2) tissue oxygen saturation as measured with continuous near-infrared spectroscopy (NIRS) across days 1–3 after aSAH.

of Acute & Tertiary Care, Pittsburgh, Pennsylvania, PA, USA Introduction: Although it has been demonstrated previously that patients with aneurysmal subarachnoid hemorrhage (aSAH) may also experience neurocardiac injury, the impact of this complication on cerebral and peripheral tissue perfusion is not well understood. Objectives: We aimed to determine if neurocardiac injury as defined by increased levels of cardiac troponin I (cTnI) was associated with changes in regional cerebral (CrSO2) and peripheral (PrSO2) tissue oxygen saturation as measured with continuous near-infrared spectroscopy (NIRS) across days 1–3 after aSAH. Methods: Longitudinal prospective analysis of 13 patients with aSAH. Inclusion criteria: age 21–75 years, spontaneous aneurysm rupture, Fisher grade >1 and/or Hunt and Hess grade >2, and with NIRS data. Exclusion: traumatic SAH, and recent myocardial dysfunction. cTnI was measured at least daily. Continuous non-invasive CrSO2 and PrSO2 commenced immediately following study enrollment. CrSO2 was acquired utilizing two separate sensors affixed to the left and right forehead, while PrSO2 was obtained utilizing a single sensor attached to either the left or right thenar eminence. Continuous CrSO2 and PrSO2 values were recorded every 5 milliseconds and then averaged on hourly intervals for statistical analyses. The daily peak values of cTnI were used in the analysis as a time-varying variable. The hourly averaged CrSO2 and PrSO2 values were also treated as time varying variables. Generalized estimating equation (GEE) was used to evaluate the association between cTnI and both right and left CrSO2, as well as PrSO2, where cTnI was used as the independent variable and CrSO2/PrSO2 as dependent variables, while controlling for age and gender.

PrSO2 values were also treated as time varying variables. Generalized estimating equation (GEE) was used to evaluate the association between cTnI and both right and left CrSO2, as well as PrSO2, where cTnI was used as the independent variable and CrSO2/PrSO2 as dependent variables, while controlling for age and gender. Results: The 13 patients in the analysis had a mean age of 56.6 years, SD = 8.9, and were predominantly female (85 %). The GEE modeling revealed that higher peak cTnI was significantly associated with lower PrSO2 and higher left-sided CrSO2. Higher cTnI was also associated with a trend toward lower CrSO2 on the right side, but that was not statistically significant (Fig. 7). Conclusions: Neurocardiac injury after aSAH as defined by higher levels of cTnI is associated with impaired peripheral perfusion, but its impact on cerebral perfusion is variable. Further investigation in a larger sample is needed to see if site of the aneurysm, cerebral vascular autoregulation, or vasospasm may play a role. Grant acknowledgment NIH R01NR014221Fig. 7 (abstract A403). ᅟ

Conclusions: Neurocardiac injury after aSAH as defined by higher levels of cTnI is associated with impaired peripheral perfusion, but its impact on cerebral perfusion is variable. Further investigation in a larger sample is needed to see if site of the aneurysm, cerebral vascular autoregulation, or vasospasm may play a role. Grant acknowledgment NIH R01NR014221Fig. 7 (abstract A403). ᅟ A404 Study of the incidence of convulsive and non-convulsive seizures in the acute phase of ischemic cerebrovascular stroke S.A. Abdelmonem1, S.A. Tahon2, T.A. Helmy1, H.S. Meligy1 1University of Alexandria, Criticalcare, Alexandria, Egypt; 2University of Alexandria, Neurology, Alexandria, Egypt Correspondence: S.A. Abdelmonem - University of Alexandria, Criticalcare, Alexandria, Egypt Introduction: Stroke is a major health problem including both ischemic and hemorrhagic types. Although a long-recognized clinical phenomenon, there remain many questions regarding the epidemiology of seizures and epilepsy after ischemic stroke, their effect on outcome, and their treatment [1]. Objectives: This pilot study assesses the incidence of seizures in acute ischemic cerebrovascular disease.

A404 Study of the incidence of convulsive and non-convulsive seizures in the acute phase of ischemic cerebrovascular stroke S.A. Abdelmonem1, S.A. Tahon2, T.A. Helmy1, H.S. Meligy1 1University of Alexandria, Criticalcare, Alexandria, Egypt; 2University of Alexandria, Neurology, Alexandria, Egypt Correspondence: S.A. Abdelmonem - University of Alexandria, Criticalcare, Alexandria, Egypt Introduction: Stroke is a major health problem including both ischemic and hemorrhagic types. Although a long-recognized clinical phenomenon, there remain many questions regarding the epidemiology of seizures and epilepsy after ischemic stroke, their effect on outcome, and their treatment [1]. Objectives: This pilot study assesses the incidence of seizures in acute ischemic cerebrovascular disease. Methods: The study was carried out on all patients presented within the first 24 of ischemic cerebrovascular stroke admitted to the units of Critical Care Medicine, Alexandria Main University Hospital, during a period of 6 months. EEG was performed in the first 24 of presentation, at the end of the first and second week of admission and if the level of consciousness deteriorated at any time during the acute phase of cerebral infarction and not explained by CT findings or any metabolic derangement.

niversity Hospital, during a period of 6 months. EEG was performed in the first 24 of presentation, at the end of the first and second week of admission and if the level of consciousness deteriorated at any time during the acute phase of cerebral infarction and not explained by CT findings or any metabolic derangement. Results: Among all the study group, the incidence of overall seizures was 20 %, of them, incidence of non-convulsive seizures were 13.3 % compared to 6.6 % of the patients showed convulsive seizures. Most seizures occurred on the first 24 hours of ischemia (66.7 %) compared to those occurring at the end of first week (25 %) and those at the end of the second week (8.3 %). Conclusions: Ischemic stroke is considered as a risk factor for the development of seizures and status epilepticus both convulsive and non-convulsive types especially during the first twenty-four hours. References 1- WHO. Stroke 1989. Recommendations on stroke prevention, and therapy. Report of the WHO Task Force on stroke and other cerebrovascular disorders. Stroke; a journal of cerebral circulation 1989; 20:1407–31 Grant acknowledgment Thanks to all members of the Department of Critical Care Medicine, University of Alexandria

References 1- WHO. Stroke 1989. Recommendations on stroke prevention, and therapy. Report of the WHO Task Force on stroke and other cerebrovascular disorders. Stroke; a journal of cerebral circulation 1989; 20:1407–31 Grant acknowledgment Thanks to all members of the Department of Critical Care Medicine, University of Alexandria Contemporary issues in infection and sepsis I A405 Does plasma from septic patients influence tumor cell growth? F. Puig1, I. Dunn-Siegrist1, J. Pugin1,2 1University of Geneva, Department of Microbiology and Molecular Medicine (MIMOL), Geneva, Switzerland; 2Geneva University Hospital, Division of Intensive Care, Geneva, Switzerland Correspondence: F. Puig - University of Geneva, Department of Microbiology and Molecular Medicine (MIMOL), Geneva, Switzerland Introduction: Sepsis is associated with immune suppression, including depressed cellular immunity. A decrease in surveillance by immune cells could be one of the mechanisms explaining the growing perception that patients surviving sepsis are at increased risk of developing cancers. However, other mechanisms may be involved in this increased rate of tumors in sepsis survivors, involving humoral factors. Sepsis is also associated with marked up- or down-regulation of various soluble mediators affecting cell growth. We hypothesized that some of these mediators may promote the proliferation of tumors. Objective: To identify and isolate this (those) plasma factor(s) enhancing tumor cell growth.

not differ between delirious patients and non-delirious patients, (28.4 ± 6 versus 27.2 ± 8, respectively). In addition, no differences in TISS-28 score were found between the delirium subgroups. Patients with delirium stayed significantly longer on the mechanical ventilator, in the ICU and in-hospital (all p < 0.001). Conclusions: Severity of illness but not delirium contributes in a higher workload in ICU patients. Furthermore, there were no differences in subgroups of delirium regarding workload. Importantly, however, delirium does burden ICU patients in the ICU.Table 4 (abstract A421). Outcome measures stratified to delirium status Not delirious Hyperactive Hypoactive Alternating Rapid reversible N=3045 N=68 N=322 N=330 N=161 Mean TISS28 27±8 27±6 28±6 29±6 28±6 APACHE II score 17±6 19±6 21±7 21±7 17±5 Days ventilated 1 [1–2] 2 [1–3] 3 [2–10] 6 [2–14] 2 [1–2] Delirium duration 0 1 [1–2] 1 [1–2] 1 [1–1] 1 [0–1] Cumulative dose of haloperidol (mg) 0 9 [4–16] 17 [8–30] 24 [11–54] 1 [1–1] ICU stay in days 1 [1–2] 2 [1–5] 5 [2–11] 7 [3–17] 1 [1–3] Hospital stay in days 7 [4–14] 9 [5–17] 16 [9–31] 22 [12–38] 7 [6–14]

Contemporary issues in infection and sepsis I A405 Does plasma from septic patients influence tumor cell growth? F. Puig1, I. Dunn-Siegrist1, J. Pugin1,2 1University of Geneva, Department of Microbiology and Molecular Medicine (MIMOL), Geneva, Switzerland; 2Geneva University Hospital, Division of Intensive Care, Geneva, Switzerland Correspondence: F. Puig - University of Geneva, Department of Microbiology and Molecular Medicine (MIMOL), Geneva, Switzerland Introduction: Sepsis is associated with immune suppression, including depressed cellular immunity. A decrease in surveillance by immune cells could be one of the mechanisms explaining the growing perception that patients surviving sepsis are at increased risk of developing cancers. However, other mechanisms may be involved in this increased rate of tumors in sepsis survivors, involving humoral factors. Sepsis is also associated with marked up- or down-regulation of various soluble mediators affecting cell growth. We hypothesized that some of these mediators may promote the proliferation of tumors. Objective: To identify and isolate this (those) plasma factor(s) enhancing tumor cell growth. Methods: Heparinized human blood samples were collected from patients with septic shock and in healthy donors. Human adenocarcinoma epithelial cells from lung (A549), colon (SW620), breast (Hs578T), and liver (HepG2) and human monocytic leukemia cells (THP-1 and HL-60) were grown to 70 % confluence and incubated with 5 - 15 % septic or healthy plasma, or control media. After 24 hours, cell proliferation was evaluated by MTT Cell Proliferation Assay. Septic and healthy plasma fractionation was performed using classical biochemical techniques.

monocytic leukemia cells (THP-1 and HL-60) were grown to 70 % confluence and incubated with 5 - 15 % septic or healthy plasma, or control media. After 24 hours, cell proliferation was evaluated by MTT Cell Proliferation Assay. Septic and healthy plasma fractionation was performed using classical biochemical techniques. Results: Septic plasma supported significantly higher epithelial cell proliferation than plasma from healthy subjects (+48 % in A549, +59 % in SW620, +41 % in Hs578T, and +29 % in HepG2 cells with 15 % plasma). Interestingly, this effect was not observed in tumoral cells of monocytic origin. Adding healthy plasma to septic plasma decreased the tumor epithelial cell proliferation effect seen with septic plasma alone in a dose-dependent manner. Initial plasma fractionation studies suggest that the proliferation factor(s) in septic plasma remains in an immunoglobulin- and albumin-free plasma fraction, is(are) not precipitable by 50 % ammonium sulfate, and is(are) a trypsin-sensitive protein(s) > 50 kDa. Conclusions: The identification of such plasma protein(s) affecting tumor cell growth could be of great interest in the fields of sepsis and cancer biology. Grant acknowledgment Unrestricted research grant from Novartis Research Foundation (15B094).

Results: Septic plasma supported significantly higher epithelial cell proliferation than plasma from healthy subjects (+48 % in A549, +59 % in SW620, +41 % in Hs578T, and +29 % in HepG2 cells with 15 % plasma). Interestingly, this effect was not observed in tumoral cells of monocytic origin. Adding healthy plasma to septic plasma decreased the tumor epithelial cell proliferation effect seen with septic plasma alone in a dose-dependent manner. Initial plasma fractionation studies suggest that the proliferation factor(s) in septic plasma remains in an immunoglobulin- and albumin-free plasma fraction, is(are) not precipitable by 50 % ammonium sulfate, and is(are) a trypsin-sensitive protein(s) > 50 kDa. Conclusions: The identification of such plasma protein(s) affecting tumor cell growth could be of great interest in the fields of sepsis and cancer biology. Grant acknowledgment Unrestricted research grant from Novartis Research Foundation (15B094). A406 High Flow Nasal Cannula (HFNC) as an alternative to noninvasive ventilation (NIV) in acute respiratory failure (ARF) in immunosuppressed patients - an Indian post liver transplant experience S. Gupta, D. Govil, S. Srinivasan, S.J. Patel, J.K. N, A. Gupta, D.S. Tomar, M. Shafi, R. Harne, D.P. Arora, N. Talwar, S. Mazumdar Medanta - The Medicity, Gurgaon, India Correspondence: S. Gupta - Medanta - The Medicity, Gurgaon, India Introduction: In immunocompromised patients, acute respiratory failure (ARF) is associated with high mortality [1] and many studies have confirmed the utility of prolonged non invasive ventilation (NIV) in such condition. Lately High Flow Nasal Cannula (HFNC) has been used in ARF in various patient subsets but not specifically in post transplant patients.

tients, acute respiratory failure (ARF) is associated with high mortality [1] and many studies have confirmed the utility of prolonged non invasive ventilation (NIV) in such condition. Lately High Flow Nasal Cannula (HFNC) has been used in ARF in various patient subsets but not specifically in post transplant patients. Objectives: To compare HFNC vs NIV as the modality to manage ARF in postoperative hypoxemia in post liver transplant patients. Methods: This was a pilot study conducted in a Liver Transplant intensive care unit (ICU) of a tertiary care hospital in India. We randomly assigned 20 consecutive post transplant patients who developed respiratory failure in the post-operative period to either HFNC group (n = 10) or to NIV group (n = 10). The HFNC was initiated at a flow rate of 60 l/min whereas NIV was set at EPAP of 5 cm and IPAP at 10 cm. Both the device setting and oxygen titration was done according to arterial blood gas (ABG) analysis. Apart from ABG analysis, we assessed the COMFORT scale as well as the RASS and CAM-ICU scale of the patient and also assessed the total nutritional deficit at the end of 48-hr therapy duration. The need for invasive mechanical ventilation (IMV) was also noted for both the groups.

g to arterial blood gas (ABG) analysis. Apart from ABG analysis, we assessed the COMFORT scale as well as the RASS and CAM-ICU scale of the patient and also assessed the total nutritional deficit at the end of 48-hr therapy duration. The need for invasive mechanical ventilation (IMV) was also noted for both the groups. Results: None of the patients in the HFNC group required need for IMV whereas 2 patients in NIV group had to be intubated. Patients in the HFNC group had better average PaO2 as compared to NIV (98.2 vs 72.6 mm Hg) respectively. The patients in the NIV group received more sedation with dexmedetomidine as compared to HFNC group (22 vs 10 hrs) respectively. The patients in the HFNC group were more comfortable as compared to NIV group and two patients in the NIV group developed delirium for which they required IMV. The average RASS score was 0 to +1 in the HFNC group whereas it ranged from −2 to +2 in the NIV group. All patients were fed either orally or enterally but the NIV group consumed less feeding due to the inability to feed orally and apprehension of aspiration due to aerophagia when fed enterally. The NIV group received 52 % less calories as compared to HFNC group in 48-hr period. Conclusions: HFNC is may be an excellent armamentarium for managing hyperemic respiratory failure in immunocompromised patients with reduced risk of intubation, more comfortable to the patients and little interruption in providing adequate nutrition. References

Results: None of the patients in the HFNC group required need for IMV whereas 2 patients in NIV group had to be intubated. Patients in the HFNC group had better average PaO2 as compared to NIV (98.2 vs 72.6 mm Hg) respectively. The patients in the NIV group received more sedation with dexmedetomidine as compared to HFNC group (22 vs 10 hrs) respectively. The patients in the HFNC group were more comfortable as compared to NIV group and two patients in the NIV group developed delirium for which they required IMV. The average RASS score was 0 to +1 in the HFNC group whereas it ranged from −2 to +2 in the NIV group. All patients were fed either orally or enterally but the NIV group consumed less feeding due to the inability to feed orally and apprehension of aspiration due to aerophagia when fed enterally. The NIV group received 52 % less calories as compared to HFNC group in 48-hr period. Conclusions: HFNC is may be an excellent armamentarium for managing hyperemic respiratory failure in immunocompromised patients with reduced risk of intubation, more comfortable to the patients and little interruption in providing adequate nutrition. References 1. Lemaire V, Mokart D, Mayaux J, Lambert J, Rabbat A, Demoule A, Azoulay E. The effects of a 2-h trial of high-flow oxygen by nasal cannula versus venturi mask in immunocompromised patients with hypoxemic acute respiratory failure: a multicentre randomized trial. Crit Care 2015 Nov 2;19:380Fig. 8 (abstract A406). HFNC vs NIV

ire V, Mokart D, Mayaux J, Lambert J, Rabbat A, Demoule A, Azoulay E. The effects of a 2-h trial of high-flow oxygen by nasal cannula versus venturi mask in immunocompromised patients with hypoxemic acute respiratory failure: a multicentre randomized trial. Crit Care 2015 Nov 2;19:380Fig. 8 (abstract A406). HFNC vs NIV A407 Intra-abdominal hypertension increases the frequency of ventilator associated pneumonia E.E. Papakrivou, D. Makris, E. Manoulakas, B. Tsolaki, B. Karadodas, E. Zakynthinos University of Thessaly School of Medicine, Department of Critical Care Medicine, Larisa, Greece Correspondence: E.E. Papakrivou - University of Thessaly School of Medicine, Department of Critical Care Medicine, Larisa, Greece Introduction: To study the effect of intra-abdominal hypertension (IAH) on the frequency of ventilator associated pneumonia (VAP) in critical care patients with risk factors for IAH. Methods: This one-center prospective study was conducted in the ICU of the University Hospital of Larissa, Greece. Consecutive critical care patients were recruited if they presented risk factors for IAH. Patients were evaluated systematically for IAH and VAP for a 28-day period. Results: Twenty-four out of 59 (41 %) patients presented IAH and 28 (47.4 %) presented VAP; seventeen (70.83 %) patients presented VAP following IAH. Multivariate analysis showed that VAP [1.18(1.10-1.22) (p = 0.004)], COPD [1.28(1.09-1.86) (p = 0.001)] and the use of antacids [9.54(2.74-33.19) (p = 0.016)] were independently associated with IAP.

59 (41 %) patients presented IAH and 28 (47.4 %) presented VAP; seventeen (70.83 %) patients presented VAP following IAH. Multivariate analysis showed that VAP [1.18(1.10-1.22) (p = 0.004)], COPD [1.28(1.09-1.86) (p = 0.001)] and the use of antacids [9.54(2.74-33.19) (p = 0.016)] were independently associated with IAP. Conclusion: IAH may have an adverse impact on the frequency of VAP in critically ill patients with risk factors for IAH. A408 Azithromycin modulates inflammatory response in a murine model of Pseudomonas aeruginosa severe infection I. Palacios Garcia, A. Diaz Martin, V. Sanchez Encinares, M. Pachón Ibañez, J. Garnacho Montero, G. Labrador, T. Cebrero Cangueiro Hospital Virgen del Rocio, Sevilla, Spain Correspondence: I. Palacios Garcia - Hospital Virgen del Rocio, Sevilla, Spain Objectives: Macrolides, apart from its antibiotic properties, are able to modulate the inflammatory response: inhibit production and secretion of pro-inflammatory cytokines (IL-1, IL-6, IL-8 and TNF-α), increase levels of anti-inflammatory cytokines (IL-10) and inhibit secretion of nitric oxide (NO). Our working hypothesis is that the use of azithromycin (AZM) with ceftazidime (CFZ) in a mouse model of severe sepsis by P. aeruginosa, modulates the inflammatory response.

ammatory cytokines (IL-1, IL-6, IL-8 and TNF-α), increase levels of anti-inflammatory cytokines (IL-10) and inhibit secretion of nitric oxide (NO). Our working hypothesis is that the use of azithromycin (AZM) with ceftazidime (CFZ) in a mouse model of severe sepsis by P. aeruginosa, modulates the inflammatory response. Methods: lethal sepsis model mouse with a clinical P. aeruginosa strain (Pa4) is characterized. i) CFZ (dose 100 mg / kg / ip / 12 h), ii) AZM (30 mg / kg / ip / 24 h): To study the inflammatory response, the following treatment groups (n = 15) for 72 hours were performed iii) COMB: CFZ + AZM, iv) control of infected animals and untreated group (CON) and v) group of uninfected mice. TNF-α determinations, IL-6, IL-10 and nitrite / nitrate (NOx) routine employed as a surrogate marker of NO metabolism in mouse plasma by ELISA (commercially available kits) ratio were performed. Results: We compare the TNF-α, IL-10 and NOx plasma concentrations, in each group (4 and 8 hours post-treatment) related to those obtained in the CON group. The COMB (AZM + CFZ) group showed lower plasma concentrations of TNF-α (pg/ml) than AZM and CFZ groups: [CON: 1477–716; COMB: 720–567; AZM: 1911–1663; CFZ: 793–666]. Plasma concentrations of IL-10 (pg/ml) were higher in the COMB and AZM groups than in the CFZ one: [CON: 1868–1761; COMB: 1541–2035; AZM: 1860–2002; CFZ: 1898–1886]. NOx concentrations (M) observed were lower in the COMB group than in AZM and CFZ ones: [CON: 76–47; COMB: 59–28; AZM: 61–51; CFZ: 35–42].

The COMB (AZM + CFZ) group showed lower plasma concentrations of TNF-α (pg/ml) than AZM and CFZ groups: [CON: 1477–716; COMB: 720–567; AZM: 1911–1663; CFZ: 793–666]. Plasma concentrations of IL-10 (pg/ml) were higher in the COMB and AZM groups than in the CFZ one: [CON: 1868–1761; COMB: 1541–2035; AZM: 1860–2002; CFZ: 1898–1886]. NOx concentrations (M) observed were lower in the COMB group than in AZM and CFZ ones: [CON: 76–47; COMB: 59–28; AZM: 61–51; CFZ: 35–42]. Conclusions: These results suggest the immunomodulatory capability of AZM as an adjunct treatment to appropriate antibiotic. Further studies are needed to infer these findings to human setting. Grant acknowledgment This project (PI10/01563) was funded by the "Health Research Fund" Health Institute Carlos III. A409 Intravenous glutamine increases risk of death in severe sepsis V. Poulose1, J. Koh1, J.W. Kam2 1Changi General Hospital, Respiratory & Critical Care Medicine, Singapore, Singapore; 2Changi General Hospital, Clinical Trials & Research Unit, Singapore, Singapore Correspondence: V. Poulose - Changi General Hospital, Respiratory & Critical Care Medicine, Singapore, Singapore Introduction: Intravenous glutamine can have beneficial effects on critically ill patients by preserving gut barrier and improving immune function. Objectives: We wanted to prove the benefit of intravenous glutamine in patients admitted to medical intensive care unit (MICU) with severe sepsis and receiving enteral nutrition.

A409 Intravenous glutamine increases risk of death in severe sepsis V. Poulose1, J. Koh1, J.W. Kam2 1Changi General Hospital, Respiratory & Critical Care Medicine, Singapore, Singapore; 2Changi General Hospital, Clinical Trials & Research Unit, Singapore, Singapore Correspondence: V. Poulose - Changi General Hospital, Respiratory & Critical Care Medicine, Singapore, Singapore Introduction: Intravenous glutamine can have beneficial effects on critically ill patients by preserving gut barrier and improving immune function. Objectives: We wanted to prove the benefit of intravenous glutamine in patients admitted to medical intensive care unit (MICU) with severe sepsis and receiving enteral nutrition. Methods: Randomized, single center, double -blind, placebo-controlled, pilot study on patients admitted to the MICU who met the criteria for severe sepsis. In the intervention arm, intravenous glutamine was given for 5 days at a dose of 0.5 g/kg body weight/day. All patients were fed enterally as per the MICU feeding protocol. The primary outcomes were 28-day mortality and the occurrence of new infections. We also looked at severity scores (SOFA), ICU length of stay (LOS), hospital LOS and duration of mechanical ventilation.

en for 5 days at a dose of 0.5 g/kg body weight/day. All patients were fed enterally as per the MICU feeding protocol. The primary outcomes were 28-day mortality and the occurrence of new infections. We also looked at severity scores (SOFA), ICU length of stay (LOS), hospital LOS and duration of mechanical ventilation. Results: Thirty nine patients were randomized to receive glutamine (n = 19) or placebo (n = 20). The glutamine group had less disease severity than placebo (median SOFA score 8 versus 11, p =0.038). There was no difference in 28-day mortality between the glutamine and placebo groups (42 % vs 15 %, p = 0.06). When adjusted for disease severity, the glutamine arm had 5.6 times higher death rates (95 % CI 1.1-30.2, p = 0.044). The glutamine group had lesser incidence of new infections (0 % vs 30 %, p = 0.02). There was no difference in ICU LOS, hospital LOS or the duration of mechanical ventilation. Conclusions: Intravenous glutamine increases mortality risk in ICU patients with severe sepsis, although it reduces risk of new infections Grant acknowledgement SingHealth Foundation Research Grant

Results: Thirty nine patients were randomized to receive glutamine (n = 19) or placebo (n = 20). The glutamine group had less disease severity than placebo (median SOFA score 8 versus 11, p =0.038). There was no difference in 28-day mortality between the glutamine and placebo groups (42 % vs 15 %, p = 0.06). When adjusted for disease severity, the glutamine arm had 5.6 times higher death rates (95 % CI 1.1-30.2, p = 0.044). The glutamine group had lesser incidence of new infections (0 % vs 30 %, p = 0.02). There was no difference in ICU LOS, hospital LOS or the duration of mechanical ventilation. Conclusions: Intravenous glutamine increases mortality risk in ICU patients with severe sepsis, although it reduces risk of new infections Grant acknowledgement SingHealth Foundation Research Grant A410 Sequential vitamin d measurement in patients with septic shock: could vitamin D levels be suppressed in septic shock? H. Yeter1, A. Kara2, O. Aktepe3, A. Topeli4 1Hacettepe University, Internal Medicine, Ankara, Turkey; 2Hacettepe University, Intensive Care Medicine, Ankara, Turkey; 3Hacettepe University, Internal medicine, Ankara, Turkey; 4Hacettepe University, Intensive care medicine, Ankara, Turkey Correspondence: H. Yeter - Hacettepe University, Internal Medicine, Ankara, Turkey Objective: Sepsis is characterized by dysregulated immune response to infection leading to organ dysfunction. Vitamin D plays a pivot role in the immune system and low levels of vitamin D have been shown to be associated with worse outcome in septic patients. In this study we tested vitamin D levels sequentially and aimed to define the vitamin D response in septic shock.

mune response to infection leading to organ dysfunction. Vitamin D plays a pivot role in the immune system and low levels of vitamin D have been shown to be associated with worse outcome in septic patients. In this study we tested vitamin D levels sequentially and aimed to define the vitamin D response in septic shock. Methods: Between September 2014 and January 2016, 41patients with septic shock were included in the study. Patients were excluded if they had a disease affecting calcium and vitamin D metabolism such as malignancy, chronic kidney disease, parathyroid disorders, pancreatitis, tumor lysis syndrome, rhabdomyolysis, renal tubular disorders and pregnancy. We measured vitamin D levels in day 1 and day 5after diagnosis of septic shock.

nts were excluded if they had a disease affecting calcium and vitamin D metabolism such as malignancy, chronic kidney disease, parathyroid disorders, pancreatitis, tumor lysis syndrome, rhabdomyolysis, renal tubular disorders and pregnancy. We measured vitamin D levels in day 1 and day 5after diagnosis of septic shock. Results: The median (min-max) age of 41 patients was 67 (19–88). 21of these patients were male. The median APACHE II score was 28(11–45). Day 1 and day 5 median vitamin D levels were 6.8 ng/ml(1–30) and 12 (2–29) ng/ml, respectively. Baseline corrected calcium and ionized calcium levels were 9.04 mg/dl (4.70-11.50) and 1.09 mmol/L (0.80-1.22). 21of the 41 septic shock patients died within 28 day. The APACHE II scores were similar between the survivor and non-survivor groups [27.5 (11–40), 30 (11–45), p = 0.13]. Baseline median corrected calcium and ionized calcium levels of survivors vs. non-survivors were 9.01 mg/dl (8.3-10.0) vs 9.16 mg/dl (4.7-11.5) and 1.07 mmol/L (0.90-1.18) vs. 1.10 mmol/L (0.80-1.22). Day 1 median vitamin D levels of survivors and non-survivors were 8.7 ng/ml(4.3-30.4) and5.3 ng/ml (1.0-21.7), respectively(p = 0.047). Day 5 vitamin D levels were not statistically different between survivors and non-survivors (n = 17; 12.3 ng/ml and n = 7; 5.7 ng/ml, p = 0.37).Vitamin D levels increased in the survivor group from 8.7 ng/ml in day 1 to 12.4 ng/m lin day 5, however the difference was not statistically different (p = 0.18). Vitamin D levels did not change in the non-survivor group from day 1 to day 5 (5.3 ng/ml to 5.7 ng/ml, p = 0.89).Kaplan Meier survival analysis revealed that patients with vitamin D levels ≥ 6.8 ng/ml (median value) had increased 28-day survival as compared to patients with low vitamin D levels (<6.8 ng/ml) (long rank test p = 0.012).

did not change in the non-survivor group from day 1 to day 5 (5.3 ng/ml to 5.7 ng/ml, p = 0.89).Kaplan Meier survival analysis revealed that patients with vitamin D levels ≥ 6.8 ng/ml (median value) had increased 28-day survival as compared to patients with low vitamin D levels (<6.8 ng/ml) (long rank test p = 0.012). Conclusion: Our study showed that vitamin D response might be different between surviving and non-surviving patients during the course of septic shock. Surviving patients had higher day 1vitamin D levels as compared to non-surviving patients. Increase in vitamin D level from day 1 to 5 suggests a clinically significant albeit statistically insignificant association of vitamin D and survival, in line with high vitamin D levels in baseline comparisons of the groups, favoring survivors, due to the limited sample size. Further studies in larger groups are clearly warranted.

itamin D level from day 1 to 5 suggests a clinically significant albeit statistically insignificant association of vitamin D and survival, in line with high vitamin D levels in baseline comparisons of the groups, favoring survivors, due to the limited sample size. Further studies in larger groups are clearly warranted. A411 Central line associated blood stream infections in the obese and overweight critically ill patients (preliminary data) I. Tsolakoglou1, G. Intas2, P. Stergiannis3, A.A. Kolaros4, E. Chalari5, E. Athanasiadou6, A. Martika1, G. Fildisis7 1General Hospital of Thessaloniki Agios Pavlos, ICU, Thessaloniki, Greece; 2General Hospital of Nikaia-Pireus AG. Panteleimon, Actinotherapy, Pireus, Greece; 3General Hospital of Athens “Agioi Anargyroi, ICU, Athens, Greece; 4General Hospital of Thessaloniki Theageneio, ICU, Thessaloniki, Greece; 5General Hospital of Nikaia-Pireus AG. Panteleimon, Anesthesiology, Pireus, Greece; 6General Hospital of Thessaloniki Agios Pavlos, Nursing Department Management, Thessaloniki, Greece, 7University of Athens, Faculty of Nursing, Critical Care Directorate, Athens, Greece Correspondence: I. Tsolakoglou - General Hospital of Thessaloniki Agios Pavlos, ICU, Thessaloniki, Greece Introduction: Central-Line-Associated Bloodstream Infections (CLABSI) have been studied extensively in ICU patients. There are no data in the literature regarding a potential association between obesity and CLABSI. Objectives: To test the hypothesis that CLABSI depends on the presence of obesity in critically ill patients.

A411 Central line associated blood stream infections in the obese and overweight critically ill patients (preliminary data) I. Tsolakoglou1, G. Intas2, P. Stergiannis3, A.A. Kolaros4, E. Chalari5, E. Athanasiadou6, A. Martika1, G. Fildisis7 1General Hospital of Thessaloniki Agios Pavlos, ICU, Thessaloniki, Greece; 2General Hospital of Nikaia-Pireus AG. Panteleimon, Actinotherapy, Pireus, Greece; 3General Hospital of Athens “Agioi Anargyroi, ICU, Athens, Greece; 4General Hospital of Thessaloniki Theageneio, ICU, Thessaloniki, Greece; 5General Hospital of Nikaia-Pireus AG. Panteleimon, Anesthesiology, Pireus, Greece; 6General Hospital of Thessaloniki Agios Pavlos, Nursing Department Management, Thessaloniki, Greece, 7University of Athens, Faculty of Nursing, Critical Care Directorate, Athens, Greece Correspondence: I. Tsolakoglou - General Hospital of Thessaloniki Agios Pavlos, ICU, Thessaloniki, Greece Introduction: Central-Line-Associated Bloodstream Infections (CLABSI) have been studied extensively in ICU patients. There are no data in the literature regarding a potential association between obesity and CLABSI. Objectives: To test the hypothesis that CLABSI depends on the presence of obesity in critically ill patients. Methods: We conducted an 18-month observational study on 576 critically ill patients, in three general ICUs in Greece. All patients had inserted a triple-lumen catheter in a central vein (internal jugular, femoral or subclavian). Body Mass Index (BMI) was determined by a dietitian on ICU admission. BMI was categorized a priori as < 18.5 (underweight), 18.5-24.9 (normal weight), 25–29.9 (overweight), and >30 (obese). CLABSI was diagnosed by examining the catheter's tip and a blood sample. Multivariate logistic regression analysis was used to estimate the association between BMI groups and CLABSI.

on ICU admission. BMI was categorized a priori as < 18.5 (underweight), 18.5-24.9 (normal weight), 25–29.9 (overweight), and >30 (obese). CLABSI was diagnosed by examining the catheter's tip and a blood sample. Multivariate logistic regression analysis was used to estimate the association between BMI groups and CLABSI. Results: From the 576 critically ill patients, (258 men, 318 women) mean aged 62.3 ± 18.4 years, 28 (4.9 %) were underweight, 220 (38.2 %) normal weight, 234 (40.6 %) overweight and 94 (16.3 %) obese. CLABSI was diagnosed in 156 (27.1 %) patients. Overweight and obese patients had significant higher CLABSI rates than the other patients (p < 0.05). Obese patients had significantly less survival rates (p < 0.05). Patient's data according to the BMI category are shown on Table 1. Additional adjustment for obesity-central line catheter association for the presence of CLABSI attenuates the obesity- central line catheter association: underweight CLABSI OR = 1.79 (95 % CI 1.54-2.03; p = 0.006); normal weight CLABSI OR = 1.86 (95 % CI 1.80-1.93; p = 0.003); overweight CLABSI OR = 1.63 (95 % CI 1.54-1.72; p = 0.001); obese- CLABSI OR = 1.64 (95 % CI 1.50-1.78, p = 0.001). Conclusions: Obesity appears to be associated with the presence of CLABSI in critically ill patients. This could be partially attributed to the more efforts made by physicians to insert the catheter in the obese patients than the other patients. REFERENCE(S)

Results: From the 576 critically ill patients, (258 men, 318 women) mean aged 62.3 ± 18.4 years, 28 (4.9 %) were underweight, 220 (38.2 %) normal weight, 234 (40.6 %) overweight and 94 (16.3 %) obese. CLABSI was diagnosed in 156 (27.1 %) patients. Overweight and obese patients had significant higher CLABSI rates than the other patients (p < 0.05). Obese patients had significantly less survival rates (p < 0.05). Patient's data according to the BMI category are shown on Table 1. Additional adjustment for obesity-central line catheter association for the presence of CLABSI attenuates the obesity- central line catheter association: underweight CLABSI OR = 1.79 (95 % CI 1.54-2.03; p = 0.006); normal weight CLABSI OR = 1.86 (95 % CI 1.80-1.93; p = 0.003); overweight CLABSI OR = 1.63 (95 % CI 1.54-1.72; p = 0.001); obese- CLABSI OR = 1.64 (95 % CI 1.50-1.78, p = 0.001). Conclusions: Obesity appears to be associated with the presence of CLABSI in critically ill patients. This could be partially attributed to the more efforts made by physicians to insert the catheter in the obese patients than the other patients. REFERENCE(S) Tagliabue C, Principi N, Giavoli C, Esposito S. Obesity: impact of infections and response to vaccines. Eur J Clin Microbiol Infect Dis. 2016 Mar;35(3):325–31.Table 1 (abstract A411). Patient's characteristics according to BMI categor

Conclusions: Obesity appears to be associated with the presence of CLABSI in critically ill patients. This could be partially attributed to the more efforts made by physicians to insert the catheter in the obese patients than the other patients. REFERENCE(S) Tagliabue C, Principi N, Giavoli C, Esposito S. Obesity: impact of infections and response to vaccines. Eur J Clin Microbiol Infect Dis. 2016 Mar;35(3):325–31.Table 1 (abstract A411). Patient's characteristics according to BMI categor Underweight (N=28) Normal weight (N=220) Overweight (N=234) Obese (N=94) p Apache II score 23.7±7.7 20.5±6.9 23.2±6.9 26.8±5.7 0.001 Femoral vein catheter insertion, n (%) 10 (35.7%) 30 (13.6%) 28 (11.9%) 14 (14.9%) 0.025 Subclavian vein catheter insertion, n (%) 10 (35.7%) 142 (64.5%) 156 (65%) 54 (61.4%) 0.025 Jugular vein catheter insertion, n (%) 8 (28.6%) 48 (21.9%) 56 (21.1%) 20 (23.7%) 0.025 N of attempts for catheter insertion 1.3±0.6 1.1±0.2 2.1±0.6 3.3±1.2 0.028 CLABSI, n (%) 6 (21.4%) 30 (13.6%) 86 (36.8%) 34 (36.2%) 0.001 ICU LOS 11.4±11.8 15.1±10.7 23.1±16.5 24.4±10.4 0.001 Hospital LOS 16.2±14.1 21.1±10.6 28.9±12.3 29.3±15.4 0.001 Survival, n (%) 16 (42.9%) 148 (67.3%) 94 (40.2%) 8 (8.5%) 0.001

.025 N of attempts for catheter insertion 1.3±0.6 1.1±0.2 2.1±0.6 3.3±1.2 0.028 CLABSI, n (%) 6 (21.4%) 30 (13.6%) 86 (36.8%) 34 (36.2%) 0.001 ICU LOS 11.4±11.8 15.1±10.7 23.1±16.5 24.4±10.4 0.001 Hospital LOS 16.2±14.1 21.1±10.6 28.9±12.3 29.3±15.4 0.001 Survival, n (%) 16 (42.9%) 148 (67.3%) 94 (40.2%) 8 (8.5%) 0.001 A412 Reactive oxygen species (ROS) production and leukocyte immunoglobulin-like receptors (LILR) expression by immune cells in pleural fluid (PL) and blood (BL) in critical care septic patients V. Faivre1, C. Mengelle1, B. Favier2, D. Payen1,3 1Université Paris Diderot, Sorbonne Paris Cité, INSERM UMR1160, Paris, France; 2Université Paris Sud / CEA, Inserm U1184, Fontenay Aux Roses, France; 3APHP, Lariboisiere University Hospital, Surgical ICU, Paris, France Correspondence: V. Faivre - Université Paris Diderot, Sorbonne Paris Cité, INSERM UMR1160, Paris, France Introduction: Sepsis induces a hyperinflammatory phase with a concomitant immunosuppression that predominates after initial phase and exposes to a failure of controlling primary infection, or to an increased risk of secondary infections. Blood monocyte (Mo) HLA-DR expression (MHC Class II) seems to characterize well innate immunodepression. The LILR expression on immune cells interacting with HLA class I induces a functional inhibition of neutrophils (PMN) as shown by the expression of LILR B2 modified during sepsis1.

ased risk of secondary infections. Blood monocyte (Mo) HLA-DR expression (MHC Class II) seems to characterize well innate immunodepression. The LILR expression on immune cells interacting with HLA class I induces a functional inhibition of neutrophils (PMN) as shown by the expression of LILR B2 modified during sepsis1. Objectives: To investigate the expression of LILR subtypes on blood Mo CD16- (classical Mo) and Mo CD16+ (non-classical patroller) as on PMNs and eosinophils (Eo) vs expression on Pl cells, more related to lung tissue inflammation2, during sepsis. Methods: In patients having severe sepsis or septic shock: 1/(flow cytometry)LILR A2, LILR B1, LILR B2, LILR B3, LILR B4 and HLA I on Mo CD16-, Mo CD16+, PMN (CD16+ Granulocytes (Gr)) and CD16- Gr in Bl and Pl. 2/ (chemiluminescence3) Spontaneous and stimulated ROS production. Measurements done at the 1st collection of PL (T1) and a week after (T2) when possible. Statistical analysis: non parametric tests. Ethical committee: CPP Ile De France VII, n° PP 15–010.

o CD16+, PMN (CD16+ Granulocytes (Gr)) and CD16- Gr in Bl and Pl. 2/ (chemiluminescence3) Spontaneous and stimulated ROS production. Measurements done at the 1st collection of PL (T1) and a week after (T2) when possible. Statistical analysis: non parametric tests. Ethical committee: CPP Ile De France VII, n° PP 15–010. Results: 11 patients were enrolled, vs 7 healthy volunteers (HV, Bl only). Compared to HV, in Bl, HLA-DR expression decreased in patients on Mo CD16- (p < 0,05) with a higher ROS production (p < 0,05). HLA-DR expression was higher on Mo CD16+. All LILR expressions were significantly higher on Mo CD16-, Mo CD16+, PMNs, except for LILRB1, with a decrease in their ligand HLA-I (p < 0,05). On Gr CD16-, LILR B1 only was increased. Comparison between Bl and Pl showed no differences except for LILRB3 expression being higher on PMNs with a higher HLA-DR expression on Mo CD16-. All LILR expressions decreased between T1 and T2 (p < 0,05) regardless the patients prognosis. Conclusions: The increased LILR expression in Bl myeloid cells during sepsis was similar in Pl cells, with a decrease along time without relation with prognosis. LILR expression could be used both in Mo and PMNs in association with Bl Mo HLA-DR to characterize immunodepression and the potential therapy for immunomodulation. References 1. Baudhuin J et al. PNAS 2013, 110: 17957–17962. 2. Marie C et al. Am. J. Respir. Crit. Care Med. 1997, 156: 1515–1522. 3. Lukaszewicz et al. Ann Crit Care 2012; 2: 10. Grant acknowledgement

Conclusions: The increased LILR expression in Bl myeloid cells during sepsis was similar in Pl cells, with a decrease along time without relation with prognosis. LILR expression could be used both in Mo and PMNs in association with Bl Mo HLA-DR to characterize immunodepression and the potential therapy for immunomodulation. References 1. Baudhuin J et al. PNAS 2013, 110: 17957–17962. 2. Marie C et al. Am. J. Respir. Crit. Care Med. 1997, 156: 1515–1522. 3. Lukaszewicz et al. Ann Crit Care 2012; 2: 10. Grant acknowledgement Fondation pour la Recherche Médicale, Commissariat à l'Energie Atomique, Ministère de l'Enseignement Supérieur et de la Recherche. A413 The impact of red blood cell (RBC) transfusion on sphingosine-1-phosphate (S1P) levels of critically ill patients A. Poppe1, M.S. Winkler1, E. Mudersbach2, J. Schreiber3, M.-L. Wruck1, E. Schwedhelm2, S. Kluge3, C. Zöllner1 1University Medical Center Hamburg-Eppendorf, Department of Anaesthesiology, Hamburg, Germany; 2University Medical Center Hamburg-Eppendorf, Institute of Clinical Pharmacology and Toxicology, Hamburg, Germany; 3University Medical Center Hamburg-Eppendorf, Department of Intensive Care Medicine, Hamburg, Germany Correspondence: A. Poppe - University Medical Center Hamburg-Eppendorf, Department of Anaesthesiology, Hamburg, Germany Introduction: Sphingosine-1-phosphate (S1P) is a G-protein coupled signaling lipid. In particular S1P has been shown to reduce sepsis induced endothelial leakage and cytokine release from immune cells and to regulate vascular tone.

ersity Medical Center Hamburg-Eppendorf, Department of Anaesthesiology, Hamburg, Germany Introduction: Sphingosine-1-phosphate (S1P) is a G-protein coupled signaling lipid. In particular S1P has been shown to reduce sepsis induced endothelial leakage and cytokine release from immune cells and to regulate vascular tone. Among others these processes are responsible for the high mortality of critically ill patients. Therefore S1P has been studied in critically ill patients; lately low S1P concentrations have been associated with sepsis severity by our group [1]. S1P is mainly released by haematopoietic cells such as erythrocytes (EC) and thrombocytes (TC) [2]. EC produce and secrete S1P continuously, whereas TC produce S1P but release S1P only upon activation [3]. Objective: We were interested if S1P concentration in blood of critically ill patients may be influenced by red blood cell (RBC) transfusion. Methods: S1P concentration in serum and EDTA plasma of 28 critically ill patients was measured before and after transfusion of one RBC preservation at several times (pre and 30 min, 180 min, 24 h post). S1P was also measured in the transfused RBC preservation. Differences between transfusion of “fresh” (RBC ≤ 15d) and “old“(RBC > 15d) RBC were analysed. Clinical data like SOFA scores, haemoglobin (HB) and haematocrit (Hct) values of all patients were monitored repeatedly.

Methods: S1P concentration in serum and EDTA plasma of 28 critically ill patients was measured before and after transfusion of one RBC preservation at several times (pre and 30 min, 180 min, 24 h post). S1P was also measured in the transfused RBC preservation. Differences between transfusion of “fresh” (RBC ≤ 15d) and “old“(RBC > 15d) RBC were analysed. Clinical data like SOFA scores, haemoglobin (HB) and haematocrit (Hct) values of all patients were monitored repeatedly. Results: S1P concentration was higher in RBC preservations compared to serum and plasma levels in patients (P < 0,05). Serum S1P levels were higher compared to plasma S1P levels (p < 0,05), though serum and plasma levels correlated positively during time. S1P concentration in plasma declined 30 min after transfusion (p < 0,05) and increased to pre levels within 24 h. Serum levels did not change. Comparing “fresh” and “old” blood preservations, S1P concentration was significantly higher in “fresh” RBC (P < 0,05). However in patients, blood S1P values did not differ if “fresh” or “old” RBCs have been transfused. Conclusions: RBC preservations contain high amounts of S1P compared to patient´s endogenous S1P. S1P concentration in RBC preservations correlates negatively with duration of storage. Transfusion of RBC influences plasma S1P negatively within 30 min. S1P declines immediately after transfusion and increases thereafter during the next 24 h. Serum S1P levels are not changed by RBC transfusion. However alteration of plasma S1P levels by RBC transfusion had no influence on the clinical outcome of our patients. References

Conclusions: RBC preservations contain high amounts of S1P compared to patient´s endogenous S1P. S1P concentration in RBC preservations correlates negatively with duration of storage. Transfusion of RBC influences plasma S1P negatively within 30 min. S1P declines immediately after transfusion and increases thereafter during the next 24 h. Serum S1P levels are not changed by RBC transfusion. However alteration of plasma S1P levels by RBC transfusion had no influence on the clinical outcome of our patients. References 1. Winkler, M.S., et al., Decreased serum concentrations of sphingosine-1-phosphate in sepsis. Crit Care, 2015. 19: p. 372. 2. Hanel, P., P. Andreani, and M.H. Graler, Erythrocytes store and release sphingosine 1-phosphate in blood. FASEB J, 2007. 21(4): p. 1202–9. 3. Yatomi, Y., et al., Sphingosine-1-phosphate: a platelet-activating sphingolipid released from agonist-stimulated human platelets. Blood, 1995. 86(1): p. 193–202.

1. Winkler, M.S., et al., Decreased serum concentrations of sphingosine-1-phosphate in sepsis. Crit Care, 2015. 19: p. 372. 2. Hanel, P., P. Andreani, and M.H. Graler, Erythrocytes store and release sphingosine 1-phosphate in blood. FASEB J, 2007. 21(4): p. 1202–9. 3. Yatomi, Y., et al., Sphingosine-1-phosphate: a platelet-activating sphingolipid released from agonist-stimulated human platelets. Blood, 1995. 86(1): p. 193–202. A414 Intensive care adult and pediatric patients at risk of mortality: inflammatory - immunity and metabolic profiles T. Tavladaki1, A.M. Spanaki1, H. Dimitriou2, E. Kondili3, C. Choulaki4, E. Meleti2, D. Kafetzopoulos4, D. Georgopoulos3, G. Briassoulis1 1University of Crete, Medical School, University Hospital, PICU, Heraklion, Greece; 2University of Crete, Medical School, Paediatric Haematology Oncology, Heraklion, Greece; 3University of Crete, Medical School, University Hospital, ICU, Heraklion, Greece; 4Institute of Molecular Biology and Biotechnology (IMBB), Foundation for Research and Technology - Hellas (FORTH), Heraklion, Greece Correspondence: T. Tavladaki - University of Crete, Medical School, University Hospital, PICU, Heraklion, Greece Introduction: Biomarkers influenced by tissue utilization, clearance, and metabolic derangements have not been shown to be able to identify complex cellular abnormalities of critical illness pathobiology.1 Currently there are no comparative data on metabolic and innate immunity - inflammatory changes in adult (A) and pediatric (P) patients with early-onset sepsis or systemic inflammatory response syndrome (SIRS).

have not been shown to be able to identify complex cellular abnormalities of critical illness pathobiology.1 Currently there are no comparative data on metabolic and innate immunity - inflammatory changes in adult (A) and pediatric (P) patients with early-onset sepsis or systemic inflammatory response syndrome (SIRS). Objectives: To determine the «at risk of mortality» profiles in A and P intensive care unit (ICU) patients in comparison to ICU-survivors and healthy (H) controls. To assess possible associations of heat shock proteins (HSP) and Resistin and Adiponectin hormone changes or oxygen consumption (VO2), dioxide production (VCO2), energy expenditure (EE) and metabolic pattern alterations with mortality in the early phase of critical illness. Methods: Seventy-eight adults (S/22; SIRS/23; H/33) and 67 children (S/18; SIRS/23; H/26) mechanically ventilated were included in the study. Blood samples were obtained within 24-hours upon admission. Mean Fluorescence Intensity (MFI) for HSP expression in monocytes (m) or neutrophils (n) was determined by Flow Cytometry. Restistin, Adiponectin and extracellular (e) HSP72 were measured using ELISA and energy-expenditure (EE) by E-COVX. Genomic DNA was extracted with PureLink Genomic DNA kit to detect HSP72 SNPs.

on. Mean Fluorescence Intensity (MFI) for HSP expression in monocytes (m) or neutrophils (n) was determined by Flow Cytometry. Restistin, Adiponectin and extracellular (e) HSP72 were measured using ELISA and energy-expenditure (EE) by E-COVX. Genomic DNA was extracted with PureLink Genomic DNA kit to detect HSP72 SNPs. Results: More patients with septic shock (p < 0.005) or patients with lactate >2 mmol/L (p < 0.02) were recorded among non-survivors (A 12.7 % vs. P 7.4 %) compared to survivors. APACHEII, eHSP72, troponin, and lactate levels were higher and VO2, VCO2, EE, metabolic pattern, and albumin lower among non-survivors compared to survivors in both groups (p < 0.05). Also, non-survivors P had higher PELOD, Resistin, and anion gap and lower EF (p < 0.05); non-survivors A showed a trend for higher eHSP72, nHSP72, and Resistin levels. In both age groups, the rs6457452 and rs1061581 HSP72 haplotypes were not related to mortality. In a logistic regression model, high lactate in A and resistin levels in P were independently associated with mortality (p < 0.01). For predicting sepsis-3 in both groups, low VO2, VCO2, metabolic pattern, and albumin and high lactate levels achieved a receiver operating characteristic curve >0.7 (p < 0.05). Conclusions: Intensive care adult and pediatric patients at risk of mortality show similar inflammatory - immunity and metabolic profiles not influenced by the rs6457452 and rs1061581 HSP72 SNPs. High lactate levels and hypometabolism better predict mortality in both groups. Grant acknowledgement

Results: More patients with septic shock (p < 0.005) or patients with lactate >2 mmol/L (p < 0.02) were recorded among non-survivors (A 12.7 % vs. P 7.4 %) compared to survivors. APACHEII, eHSP72, troponin, and lactate levels were higher and VO2, VCO2, EE, metabolic pattern, and albumin lower among non-survivors compared to survivors in both groups (p < 0.05). Also, non-survivors P had higher PELOD, Resistin, and anion gap and lower EF (p < 0.05); non-survivors A showed a trend for higher eHSP72, nHSP72, and Resistin levels. In both age groups, the rs6457452 and rs1061581 HSP72 haplotypes were not related to mortality. In a logistic regression model, high lactate in A and resistin levels in P were independently associated with mortality (p < 0.01). For predicting sepsis-3 in both groups, low VO2, VCO2, metabolic pattern, and albumin and high lactate levels achieved a receiver operating characteristic curve >0.7 (p < 0.05). Conclusions: Intensive care adult and pediatric patients at risk of mortality show similar inflammatory - immunity and metabolic profiles not influenced by the rs6457452 and rs1061581 HSP72 SNPs. High lactate levels and hypometabolism better predict mortality in both groups. Grant acknowledgement This research has been co-financed by the European Union (European Social Fund (ESF)) and Greek national funds through the Operational Program ´´Education and Lifelong Learning´´ of the National Strategic Reference Framework (NSRF)-Research Funding Program: THALES.

Conclusions: Intensive care adult and pediatric patients at risk of mortality show similar inflammatory - immunity and metabolic profiles not influenced by the rs6457452 and rs1061581 HSP72 SNPs. High lactate levels and hypometabolism better predict mortality in both groups. Grant acknowledgement This research has been co-financed by the European Union (European Social Fund (ESF)) and Greek national funds through the Operational Program ´´Education and Lifelong Learning´´ of the National Strategic Reference Framework (NSRF)-Research Funding Program: THALES. A415 Prognosis value of immunoglobulins IGG, IGA and IGM in patients with severe sepsis or septic shock A. García-de la Torre1, M.V. de la Torre-Prados2, T. Tsvetanova-Spasova2, P. Nuevo-Ortega2, C. Rueda-Molina2, A. Fernández-Porcel2, E. Camara-Sola2, L. Salido-Díaz2, A. García-Alcántara2 1University Hospital Virgen de la Victoria / IBIMA, Clinical Chemistry Department, Málaga, Spain; 2University Hospital Virgen de la Victoria / IBIMA, Department of Intensive Care Unit, Málaga, Spain Correspondence: A. García-de la Torre - University Hospital Virgen de la Victoria / IBIMA, Clinical Chemistry Department, Málaga, Spain Introduction: Despite the promise of Ig therapy, the last Surviving Sepsis Guidelines (2012) does not suggest using intravenous Ig in adults with severe sepsis or septic shock, requiring more research in this field. Objectives: Assess the prognostic value of serum immunoglobulins (Igs): IgG, IgA and IgM determined within 24 hours from severe sepsis (SS) or septic shock (SSh) onset.

A415 Prognosis value of immunoglobulins IGG, IGA and IGM in patients with severe sepsis or septic shock A. García-de la Torre1, M.V. de la Torre-Prados2, T. Tsvetanova-Spasova2, P. Nuevo-Ortega2, C. Rueda-Molina2, A. Fernández-Porcel2, E. Camara-Sola2, L. Salido-Díaz2, A. García-Alcántara2 1University Hospital Virgen de la Victoria / IBIMA, Clinical Chemistry Department, Málaga, Spain; 2University Hospital Virgen de la Victoria / IBIMA, Department of Intensive Care Unit, Málaga, Spain Correspondence: A. García-de la Torre - University Hospital Virgen de la Victoria / IBIMA, Clinical Chemistry Department, Málaga, Spain Introduction: Despite the promise of Ig therapy, the last Surviving Sepsis Guidelines (2012) does not suggest using intravenous Ig in adults with severe sepsis or septic shock, requiring more research in this field. Objectives: Assess the prognostic value of serum immunoglobulins (Igs): IgG, IgA and IgM determined within 24 hours from severe sepsis (SS) or septic shock (SSh) onset. Methods: A cohort study was performed in 133 critically ill adult patients admitted to the Intensive Care Unit (ICU). Demographic variables, severity score, mortality at 28 days were recorded, those patients who had been administered the first dose of intravenous immunoglobulins were excluded. The Igs were determined by nephelometry, Dimension Vista.® Siemens. Statistical analysis was performed using SPSS 15.0 for Windows (SPSS Inc. Chicago, IL, USA).

bles, severity score, mortality at 28 days were recorded, those patients who had been administered the first dose of intravenous immunoglobulins were excluded. The Igs were determined by nephelometry, Dimension Vista.® Siemens. Statistical analysis was performed using SPSS 15.0 for Windows (SPSS Inc. Chicago, IL, USA). Results: Of the 133 patients enrolled in the study 16.5 % met SS criteria and 83.5 % SSh. The mean age was 62 [inter-quartile range (IR): 48.5-70.5] years, 62.4 % were men. The main sources of infection were: respiratory tract 36.8 % and intra-abdomen 28.6 % The mediam stay in the ICU was 6 [3.5-11] days and mortality at 28 days was 21.8 % (n = 29). Median serum IgG Igs were 792 [607 to 976.5] mg / dl, IgA 204 [147.5 to 315] mg / dl, and IgM 76 [46.5 to 132] mg / dl. Patients who died had significantly higher levels of clinical severity with APACHE II 29.8 vs 24.1, 8.9 vs 12.1 Sequential Organ Failure Assessment (SOFA) and organ dysfunction number 4.6 vs 3.6 and in serum IgA: 323 vs 195 mg / dl. This difference was not statistically significant in levels of IgG 822 vs. 774 mg / dl or IgM 86 vs. 72.5 mg / dl. Conclusions: Serum levels of IgG, IgA and IgM were higher in patients who died, but only IgA showed a prognostic value when was measured within 24 hours from SS or SSh onset. References 1. Alejandria MM, Lansang MAD, Dans LF, Mantaring III JB. Intravenous immunoglobulin for treating sepsis, severe sepsis and septic shock (Review). The Cochrane Library 2013, Issue 9, 1–107.

Conclusions: Serum levels of IgG, IgA and IgM were higher in patients who died, but only IgA showed a prognostic value when was measured within 24 hours from SS or SSh onset. References 1. Alejandria MM, Lansang MAD, Dans LF, Mantaring III JB. Intravenous immunoglobulin for treating sepsis, severe sepsis and septic shock (Review). The Cochrane Library 2013, Issue 9, 1–107. A416 Similar adult and pediatric inflammatory - immunity and metabolic Sepsis3 profiles compared to SIRS and healthy controls T. Tavladaki1, A.M. Spanaki1, H. Dimitriou2, E. Kondili3, C. Choulaki4, D.E. Meleti2, D. Kafetzopoulos4, D. Georgopoulos3, G. Briassoulis1 1University of Crete, Medical School, University Hospital, PICU, Heraklion, Greece; 2University of Crete, Medical School, Paediatric Haematology Oncology, Heraklion, Greece; 3University of Crete, Medical School, University Hospital, ICU, Heraklion, Greece; 4Institute of Molecular Biology and Biotechnology (IMBB), Foundation for Research and Technology - Hellas (FORTH), Heraklion, Greece Correspondence: T. Tavladaki - University of Crete, Medical School, University Hospital, PICU, Heraklion, Greece Introduction: The new sepsis definition, known as Sepsis-3, shifted the diagnostic focus from infection with systemic inflammation to infection triggered organ failure, and it does away with the systemic inflammatory response syndrome criteria (SIRS). The role of cellular innate immunity and inflammatory-metabolic abnormalities in S and SIRS development has not been clarified.

shifted the diagnostic focus from infection with systemic inflammation to infection triggered organ failure, and it does away with the systemic inflammatory response syndrome criteria (SIRS). The role of cellular innate immunity and inflammatory-metabolic abnormalities in S and SIRS development has not been clarified. Objectives: To examine early heat shock proteins (HSP) and Resistin and Adiponectin hormone changes, along with clinical, inflammatory, and metabolic profiles, in adult (A) and pediatric (P) intensive care unit (ICU) patients with sepsis (S). To compare the S profiles with those of SIRS or healthy-controls (H) in each group. Methods: . Seventy-eight adults (S/22; SIRS/23; H/33) and 67 children (S/18; SIRS/23; H/26) mechanically ventilated were included in the study. Blood samples were obtained within 24-hours upon admission. Mean Fluorescence Intensity (MFI) for HSP expression in monocytes (m) or neutrophils (n) was determined using 4-colour Flow Cytometry. Restistin, Adiponectin and extracellular (e) HSP72 using ELISA. E-COVX for energy-expenditure (EE). Genomic DNA was extracted by using the PureLink Genomic DNA kit to detect the polymorphic HSP72 SNPs rs1061581 and rs6457452.

P expression in monocytes (m) or neutrophils (n) was determined using 4-colour Flow Cytometry. Restistin, Adiponectin and extracellular (e) HSP72 using ELISA. E-COVX for energy-expenditure (EE). Genomic DNA was extracted by using the PureLink Genomic DNA kit to detect the polymorphic HSP72 SNPs rs1061581 and rs6457452. Results: Similarly in A and P ICU-patients, APACHEII, SAPS3, HR, CRP, lactate, creatinine, and resistin, were higher and EF, mHSP72, VO2, VCO2, EE, metabolic pattern, glucose, and albumin lower in S compared to SIRS and/or H (p < 0.05). Only PELOD and MAP in P and INR, nHSP72 and glucose in A differed among groups (p < 0.05). For predicting sepsis-3 in both groups CRP, resistin, lactate, and eHSP72 in P, achieved a receiver operating characteristic curve (AUROC) > 0.80 (p < 0.05). In a logistic regression model resistin and mHSP72 independently discriminated S from SIRS in children only (p < 0.001). In A and P, genotype HSP72 analysis did not disclose any group differences regarding the rs6457452 or the rs1061581 haplotypes. Conclusions: Sepsis presents with repressed innate immunity and metabolism and enhanced inflammatory response in adult and pediatric septic patients. These changes may help in depicting a common for adults and children new sepsis-3 profile discriminated from the one of SIRS. Grant acknowledgement

Results: Similarly in A and P ICU-patients, APACHEII, SAPS3, HR, CRP, lactate, creatinine, and resistin, were higher and EF, mHSP72, VO2, VCO2, EE, metabolic pattern, glucose, and albumin lower in S compared to SIRS and/or H (p < 0.05). Only PELOD and MAP in P and INR, nHSP72 and glucose in A differed among groups (p < 0.05). For predicting sepsis-3 in both groups CRP, resistin, lactate, and eHSP72 in P, achieved a receiver operating characteristic curve (AUROC) > 0.80 (p < 0.05). In a logistic regression model resistin and mHSP72 independently discriminated S from SIRS in children only (p < 0.001). In A and P, genotype HSP72 analysis did not disclose any group differences regarding the rs6457452 or the rs1061581 haplotypes. Conclusions: Sepsis presents with repressed innate immunity and metabolism and enhanced inflammatory response in adult and pediatric septic patients. These changes may help in depicting a common for adults and children new sepsis-3 profile discriminated from the one of SIRS. Grant acknowledgement This research has been co-financed by the European Union (European Social Fund (ESF)) and Greek national funds through the Operational Program ´´Education and Lifelong Learning´´ of the National Strategic Reference Framework (NSRF)-Research Funding Program: THALES.

Conclusions: Sepsis presents with repressed innate immunity and metabolism and enhanced inflammatory response in adult and pediatric septic patients. These changes may help in depicting a common for adults and children new sepsis-3 profile discriminated from the one of SIRS. Grant acknowledgement This research has been co-financed by the European Union (European Social Fund (ESF)) and Greek national funds through the Operational Program ´´Education and Lifelong Learning´´ of the National Strategic Reference Framework (NSRF)-Research Funding Program: THALES. A417 Role of biomarkers in early postoperative period of lung transplantation B. Suberviola1, J. Riera2, L. Rellan3, M. Sanchez4, J.C. Robles5, E. Lopez6, R. Vicente7, E. Miñambres1, M. Santibañez8 1University Hospital Marques de Valdecilla, Santander, Spain; 2University Hospital Vall d´Hebron, Barcelona, Spain; 3University Hospital A Coruña, La Coruña, Spain; 4University Hospital Puerta de Hierro, Madrid, Spain; 5University Hospital Reina Sofia, Cordoba, Spain; 6University Hospital 12 de Octubre, Madrid, Spain; 7University Hospital La Fe, Valencia, Spain; 8Nursing School of University of Cantabria, Santander, Spain Correspondence: B. Suberviola - University Hospital Marques de Valdecilla, Santander, Spain Introduction: The major reported causes of death within the first 30 days after lung transplantation (LT) are graft failure and non-CMV infections. Delay in starting effective antimicrobiological treatment in case of infections is associated with worse outcome. It is imperative to diagnose the presence of infectious complications after LT in a timely fashion. Biomarkers have been proposed as a tool to improve early diagnosis of infections in this population.

ions. Delay in starting effective antimicrobiological treatment in case of infections is associated with worse outcome. It is imperative to diagnose the presence of infectious complications after LT in a timely fashion. Biomarkers have been proposed as a tool to improve early diagnosis of infections in this population. Objectives: To determine how sequential measurements of PCT and CRP could improve the diagnosis of early infectious complications after LT. Methods: Multicentre prospective observational study between September 2014 and September 2015. The study included all 7 centres authorized to perform lung transplantation in Spain.Biomarker measurements were executed on ICU admission (day 1) and daily till 7th postoperative day. Patients were split into two groups depending on the presence of infectious complications.

y between September 2014 and September 2015. The study included all 7 centres authorized to perform lung transplantation in Spain.Biomarker measurements were executed on ICU admission (day 1) and daily till 7th postoperative day. Patients were split into two groups depending on the presence of infectious complications. Results: Two hundred and thirty three consecutive patients (148 men, median age 56 [range 47 to 61] years) underwent single (n = 108, 46.4 %) or bilateral (n = 125, 53.6 %) LT. In the whole population, both biomarkers, PCT and CRP, presented similar kinetics during study period with an initial increase in their levels, a plasma peak recorded in the first 48 hours and a progressive decline over the next days. PCT plasma levels were similar for PGD grades 1 and 2 and increased significantly in the group of patients with PGD grade 3. CRP levels were similar in all groups. The median PCT levels were significantly lower in patients without infection than in patients with 'Infection' for all days of follow up (Table 2). During the first 5 postoperative days (day 0 to day 4), PCT levels beyond median value were statistically associated with higher risk of infection. Thus, PCT levels beyond 0.50 ng/ml on ICU admission or 1.17 ng/ml on postoperative day 1 were associated with an increase in risk of infection of two-fold and three-fold respectively. Associations remain significant after adjusting by sex, age, need of postoperative ECMO, creatinine levels and type of lung transplant. However, after stratifying by primary graft dysfunction presentation these associations disappeared with respect to the group of patients with PGD 3 and increased in the NO PGD 3 group.Therefore, in the absence of PGD 3, previously described cut-off levels for day 0 and day 1 were significantly associated with an increase of three-fold and four-fold in the risk of infection development.

these associations disappeared with respect to the group of patients with PGD 3 and increased in the NO PGD 3 group.Therefore, in the absence of PGD 3, previously described cut-off levels for day 0 and day 1 were significantly associated with an increase of three-fold and four-fold in the risk of infection development. Conclusions: In absence of severe PGD, PCT peak levels are predictor of infectious complications development during the first postoperative week. PGD grade 3 produce an important increase of PCT levels and interfere with PCT infectious diagnosis ability. PCT was superior to CRP in infections diagnosis during this period. Grant acknowledgement Supported by a grant from the Mutua Madrileña Foundation (FMM 14/01).Table 2 (abstract A417). Median levels for PCT levels for each day Overall NO Infection Infection in trasplant recipient NO Dysfunction grade 3 Primary graft dysfunction grade 3 N=233 N=181 N=52 N=205 N=28 PCT LEVELS Median Median Median P value Median Median P value ICU Admission 0.50 0.36 2.00 <0.001 0.47 4.57 <0.001 Day 1 1.17 1.01 3.83 <0.001 1.07 4.90 <0.001 Day 2 1.14 0.94 2.73 <0.001 1.01 2.61 0.006 Day 3 0.70 0.55 1.70 <0.001 0.69 1.32 0.402 Day 4 0.56 0.50 0.78 0.002 0.54 1.11 0.105 Day 5 0.28 0.20 0.40 0.006 0.28 0.48 0.680 Day 6 0.20 0.20 0.30 0.026 0.20 0.39 0.447

value ICU Admission 0.50 0.36 2.00 <0.001 0.47 4.57 <0.001 Day 1 1.17 1.01 3.83 <0.001 1.07 4.90 <0.001 Day 2 1.14 0.94 2.73 <0.001 1.01 2.61 0.006 Day 3 0.70 0.55 1.70 <0.001 0.69 1.32 0.402 Day 4 0.56 0.50 0.78 0.002 0.54 1.11 0.105 Day 5 0.28 0.20 0.40 0.006 0.28 0.48 0.680 Day 6 0.20 0.20 0.30 0.026 0.20 0.39 0.447 PERIOPERATIVE INTENSIVE CARE AND SEDATION A418 A pilot study to assess the impact of anxiety and depression during admission in post-operative intensive care patients M. Le Guen, J. Moore, N. Mason Central Manchester Foundation Trust, Critical Care Department, Manchester, UK Correspondence: M. Le Guen - Central Manchester Foundation Trust, Critical Care Department, Manchester, UK Introduction: The development of neuropsychological dysfunction in survivors of critical care has been widely documented in the literature (1). In contrast, the impact on patient outcomes of anxiety and depression at the time of critical care admission has been less well documented (2). Objectives: We conducted a pilot study to determine the feasibility of assessing critical care patients' anxiety and depression during their admission and the potential impact of these morbidities on patient outcomes.

PERIOPERATIVE INTENSIVE CARE AND SEDATION A418 A pilot study to assess the impact of anxiety and depression during admission in post-operative intensive care patients M. Le Guen, J. Moore, N. Mason Central Manchester Foundation Trust, Critical Care Department, Manchester, UK Correspondence: M. Le Guen - Central Manchester Foundation Trust, Critical Care Department, Manchester, UK Introduction: The development of neuropsychological dysfunction in survivors of critical care has been widely documented in the literature (1). In contrast, the impact on patient outcomes of anxiety and depression at the time of critical care admission has been less well documented (2). Objectives: We conducted a pilot study to determine the feasibility of assessing critical care patients' anxiety and depression during their admission and the potential impact of these morbidities on patient outcomes. Methods: Patients were recruited following admission to a general intensive care unit post surgical intervention. Patients had a Hospital Anxiety and Depression Scale (HADS) assessment on day 7 and day 15 post operatively, In addition demographic (gender, age, elective vs emergency surgery, cancer vs non-cancer surgery, prior anti-depressant use), intervention and outcome (length of stay, infection, additional oxygen requirement, in-dwelling catheter in situ) data was collected.

ale (HADS) assessment on day 7 and day 15 post operatively, In addition demographic (gender, age, elective vs emergency surgery, cancer vs non-cancer surgery, prior anti-depressant use), intervention and outcome (length of stay, infection, additional oxygen requirement, in-dwelling catheter in situ) data was collected. Results: One hundred and twenty two patients were recruited to the study and 69 (56.6 %) had a HADS assessment on day 7. Amongst the 52 patients that did not have a HADS assessment completed, 27 (52 %) were discharged prior to day 7 and 16 (31 %) either refused or were not appropriate for assessment. The completed HADS assessments indicated that 29 % and 30 % of patients had significant anxiety and depression scores respectively. Increased day 7 median anxiety scores were associated with female gender (6 vs 4, P = 0.04), anti-depressant use (8 vs 5, P = 0.01), oxygen requirement (9 vs 5, P = 0.049), infection (7 vs 4, P = 0.05) and urinary catheter use (8 vs 4, P = 0.04). Increased median depression scores were associated with female gender (7 vs 5, P = 0.04) and hospital stay longer than 14 days (7 vs 5, P = 0.007). Twenty-four patients had a further HADS assessment on day 15, with 35 patients having been discharged from hospital in the intervening period. Based on Wilcoxon Signed Ranks tests, anxiety but not depression scores appeared to decrease during the patient's hospital stay (z = 2.03, P = 0.04 and z = 0.25, P = 0.8, respectively).

our patients had a further HADS assessment on day 15, with 35 patients having been discharged from hospital in the intervening period. Based on Wilcoxon Signed Ranks tests, anxiety but not depression scores appeared to decrease during the patient's hospital stay (z = 2.03, P = 0.04 and z = 0.25, P = 0.8, respectively). Conclusions: The assessment of anxiety and depression amongst peri-operative critical care patients is feasible. Increased anxiety levels appear associated with a number of interventions. Increased depression scores appear associated with increased hospital length of stay. Anxiety but not depression levels appear to decrease over time during patients' hospital stays. References 1. Parker, Ann M., et al. "Posttraumatic Stress Disorder in Critical Illness Survivors: A Metaanalysis*." Critical care medicine 43.5 (2015): 1121–1129. 2. Rincon, Hernan G., et al. "Prevalence, detection and treatment of anxiety, depression, and delirium in the adult critical care unit." Psychosomatics 42.5 (2001): 391–396. Grant acknowledgement No grants were accepted or used to support this work

1. Parker, Ann M., et al. "Posttraumatic Stress Disorder in Critical Illness Survivors: A Metaanalysis*." Critical care medicine 43.5 (2015): 1121–1129. 2. Rincon, Hernan G., et al. "Prevalence, detection and treatment of anxiety, depression, and delirium in the adult critical care unit." Psychosomatics 42.5 (2001): 391–396. Grant acknowledgement No grants were accepted or used to support this work A419 Opioid use after propofol or sevoflurane anesthesia: a randomized trial M. Windpassinger1, O. Plattner1, E. Mascha2, D.I. Sessler2, Outcomes Research 1Medical University of Vienna, Anesthesiology & Intensiv Care, Vienna, Austria; 2Cleveland Clinic, Outcomes Research, Cleveland, OH, USA Correspondence: M. Windpassinger - Medical University of Vienna, Anesthesiology & Intensiv Care, Vienna, Austria Introduction: Postoperative pain might be ameliorated by substituting propofol for sevoflurane anesthesia. Support for this theory comes from human pain models in which propofol reduced hyperalgesia and allodynia in response to pinprick and electric stimulation.1 Benefit may result from central and peripheral analgesic effects of sub-hypnotic doses of propofol, as well as suppression of spinal sensitization.2,3 Additional analgesic effects of propofol likely result from interactions with N-methyl-d-aspartate (NMDA), non-NMDA receptors, and via activation of gamma-aminobutyric acid A (GABAA) receptors in the dorsal root ganglion nociceptor cells.

ub-hypnotic doses of propofol, as well as suppression of spinal sensitization.2,3 Additional analgesic effects of propofol likely result from interactions with N-methyl-d-aspartate (NMDA), non-NMDA receptors, and via activation of gamma-aminobutyric acid A (GABAA) receptors in the dorsal root ganglion nociceptor cells. Consistent with multiple analgesic mechanisms, some studies report that patients anesthetized with propofol have less postoperative pain than those anesthetized with volatile anesthestics.Other studies, though, do not support a postoperative analgesic effect of intraoperative propofol. Objectives: Because it remains unclear whether intraoperative propofol analgesia ameliorates postoperative pain, we tested the primary hypothesis that postoperative opioid requirements are greater in patients anesthetized with sevoflurane than propofol. Methods: Ninety patients having open vein stripping were randomized to either sevoflurane or propofol anesthesia. Pain was treated with bolus piritramid and patient-controlled morphine hydrochloride. The primary outcome was total opioid use from the end of surgery until the first postoperative morning. Pain scores (11-point Likert verbal response score) were recorded by a blinded investigator at 30-minute intervals for the initial 4 hours and on the first postoperative morning.

nt-controlled morphine hydrochloride. The primary outcome was total opioid use from the end of surgery until the first postoperative morning. Pain scores (11-point Likert verbal response score) were recorded by a blinded investigator at 30-minute intervals for the initial 4 hours and on the first postoperative morning. Results: Sevoflurane was not superior to propofol on postoperative opioid consumption, giving a ratio of means (95 % interim-adjusted CI) of 0.91 (0.33, 2.4), P = 0.74. Medians [quartiles] of morphine sulfate equivalents were 9.8 mg [4, 19] in the sevoflurane group and 10 [6, 20] mg in the propofol group. In addition, no difference on pain score over time was found between two groups, with a mean difference on an 11-point scale of 0.20 (95 % interim-adjusted CI: -0.36,0.73, P = 0.31). Conclusions: Intraoperative sevoflurane did not reduce postoperative analgesia, which is consistent with most previous reports. References 1. Bandschapp O., Filitz J., Ihmsen H et al. Analgesic and antihyperalgesic properties of propofol in a human pain model. Anesthesiology 2010;113:421–428 2. Jewett BA, Gibbs LM, Tarasuik A, Kendig J. Propofol and barbiturate depression of spinal nociceptive neurotransmission. Anesthesiology, 1992, Vol: 77:1148–1149 3. Cheng S., Yeh J., Flood P. Anaesthesia Matters:Patients anaesthetized with propofol have less postoperative pain than those anesthetized with isoflurane Anesthesia and Analgesia 2008, Vol:106:264–269

2. Jewett BA, Gibbs LM, Tarasuik A, Kendig J. Propofol and barbiturate depression of spinal nociceptive neurotransmission. Anesthesiology, 1992, Vol: 77:1148–1149 3. Cheng S., Yeh J., Flood P. Anaesthesia Matters:Patients anaesthetized with propofol have less postoperative pain than those anesthetized with isoflurane Anesthesia and Analgesia 2008, Vol:106:264–269 A420 Comparison between fuzzy and quadratic models in the development of the EEG based consciousness index qCON U. Melia1, J. Fontanet1, J.P. van den Berg2, M.M.R.F. Struys2,3, H.E.M. Vereecke2,3, E.W. Jensen1,4 1Quantium Medical, Barcelona, Spain; 2University Medical Center Groningen, University of Groningen, Gronigen, Netherlands; 3Ghent University, Ghent, Belgium; 4Center for Biomedical Engineering Research, Barcelona, Spain Correspondence: U. Melia - Quantium Medical, Barcelona, Spain Introduction: In the last decades, several methods have been developed for the noninvasive assessment of the level of consciousness during general anesthesia by processing physiological signals. The performance of the developed indices depends on a large number of factors such as the choice of the signal, the selected parameters, the inter-individual variability, etc. However, it is not clear how the type of the model that is used for the index development and calculation can affect its performances.

signals. The performance of the developed indices depends on a large number of factors such as the choice of the signal, the selected parameters, the inter-individual variability, etc. However, it is not clear how the type of the model that is used for the index development and calculation can affect its performances. Objectives: The objective of the study was to compare two electroencephalographic (EEG) derived depth of anesthesia indexes that were developed with two different models: an adaptive neuro-fuzzy inference system (qCON ANFIS) and a quadratic equation model (qCON QE). The models have the same inputs based on the energy on four EEG frequency bands. Methods: After IRB approval, 71 patients scheduled for elective surgery were randomized in four groups. Anesthesia was induced using effect-site controlled, target controlled infusion with effect size concentration of propofol set to 8.6, 5.9, 3.6 or 2 μg/mL while the corresponding remifentanil was set to 1, 2, 4 and 8 ng/mL respectively. Data were used from 2.5 minutes before to 11 minutes after starting pumps. Patients enrolled were adults from both genders with age 53 ± 13 years, weight 79 ± 14 kg and height 174 ± 9 cm (mean ± standard deviation). The EEG was recorded with qCON 2000 monitor (Quantium Medical, Barcelona, Spain) at a sampling frequency of 1024 Hz. The qCON ANFIS and qCON QE indices were calculated on one-second EEG windows.

Methods: After IRB approval, 71 patients scheduled for elective surgery were randomized in four groups. Anesthesia was induced using effect-site controlled, target controlled infusion with effect size concentration of propofol set to 8.6, 5.9, 3.6 or 2 μg/mL while the corresponding remifentanil was set to 1, 2, 4 and 8 ng/mL respectively. Data were used from 2.5 minutes before to 11 minutes after starting pumps. Patients enrolled were adults from both genders with age 53 ± 13 years, weight 79 ± 14 kg and height 174 ± 9 cm (mean ± standard deviation). The EEG was recorded with qCON 2000 monitor (Quantium Medical, Barcelona, Spain) at a sampling frequency of 1024 Hz. The qCON ANFIS and qCON QE indices were calculated on one-second EEG windows. The agreement between the two anesthesia indexes was evaluated with correlation (R2), prediction probability (Pk) and Bland Altman analysis by calculating the bias and the standard deviation (SD) of the differences between the two indexes. Limits of agreement were defined as the bias ± 1.96SD in which 95 % of the differences between the two indexes are expected to lie. It is considered a clinically acceptable level of agreement when two indexes would differ by less than 10 units. Results: Fig. 9 shows the density plot of qCON QE vs. qCON ANFIS. Table 3 shows the values of the R2 and Pk between qCON ANFIS and qCON QE. Figure. 10 shows the results of the Bland Altman analysis of qCON ANFIS versus qCON QE.

The agreement between the two anesthesia indexes was evaluated with correlation (R2), prediction probability (Pk) and Bland Altman analysis by calculating the bias and the standard deviation (SD) of the differences between the two indexes. Limits of agreement were defined as the bias ± 1.96SD in which 95 % of the differences between the two indexes are expected to lie. It is considered a clinically acceptable level of agreement when two indexes would differ by less than 10 units. Results: Fig. 9 shows the density plot of qCON QE vs. qCON ANFIS. Table 3 shows the values of the R2 and Pk between qCON ANFIS and qCON QE. Figure. 10 shows the results of the Bland Altman analysis of qCON ANFIS versus qCON QE. Conclusions: The two indexes qCON ANFIS and qCON QE that were developed with two different models showed a strong correlation (R2 > 0.95, Pk > 0.95) and a good agreement (Bias < 1 and limits of agreement < 10). In conclusions, changing the model of the qCON calculation from a fuzzy based model to a quadratic equation do not affect the index performances in depth of anesthesia assessment, with the advantage of working with a quadratic model structure that has less coefficients than the fuzzy model.Table 3 (abstract A420). Values of R2, Pk (Standard Deviation SE) Indexes R2 Pk SE qCON ANFIS vs. qCONQE 0.9591 0.9637 0.0003 Fig. 9 (abstract A420). Scatter density plot of qCOn QE vs. qCO Fig. 10 (abstract A420). Bland Altman analysis: qCON QE vs. qCON

Conclusions: The two indexes qCON ANFIS and qCON QE that were developed with two different models showed a strong correlation (R2 > 0.95, Pk > 0.95) and a good agreement (Bias < 1 and limits of agreement < 10). In conclusions, changing the model of the qCON calculation from a fuzzy based model to a quadratic equation do not affect the index performances in depth of anesthesia assessment, with the advantage of working with a quadratic model structure that has less coefficients than the fuzzy model.Table 3 (abstract A420). Values of R2, Pk (Standard Deviation SE) Indexes R2 Pk SE qCON ANFIS vs. qCONQE 0.9591 0.9637 0.0003 Fig. 9 (abstract A420). Scatter density plot of qCOn QE vs. qCO Fig. 10 (abstract A420). Bland Altman analysis: qCON QE vs. qCON A421 ICU workload in different groups of delirious patients P.J.T. Rood, F. van de Schoor, K. van Tertholen, P. Pickkers, M. van den Boogaard Radboud University Nijmegen Medical Centre, Intensive Care, Nijmegen, Netherlands Correspondence: P.J.T. Rood - Radboud University Nijmegen Medical Centre, Intensive Care, Nijmegen, Netherlands Introduction: Delirium occurs frequently in ICU patients and is associated with numerous adverse effects. Apart from the hyperactive, hypoactive and mixed subtype, recently the 'rapid reversible sedation-related delirium' was described which disappears after discontinuing of sedatives and is thought to have different outcomes then regular hypoactive delirium. Apart from a burden for the patient, the presence of delirium may also increase the workload for ICU professionals. Typically, workload in the ICU is measured using the Simplified Therapeutic Intervention Scoring System (TISS-28).

17±5 Days ventilated 1 [1–2] 2 [1–3] 3 [2–10] 6 [2–14] 2 [1–2] Delirium duration 0 1 [1–2] 1 [1–2] 1 [1–1] 1 [0–1] Cumulative dose of haloperidol (mg) 0 9 [4–16] 17 [8–30] 24 [11–54] 1 [1–1] ICU stay in days 1 [1–2] 2 [1–5] 5 [2–11] 7 [3–17] 1 [1–3] Hospital stay in days 7 [4–14] 9 [5–17] 16 [9–31] 22 [12–38] 7 [6–14] A422 A colour coded, targeted RASS (Richmond Agitation-Sedation Scale) protocol to improve sedation practice Z.J. Beardow, H. Redhead, K. Paramasivam Leeds Teaching Hospitals NHS Trust, Adult Critical Care, Leeds, UK Correspondence: Z.J. Beardow - Leeds Teaching Hospitals NHS Trust, Adult Critical Care, Leeds, UK Introduction: Continuous infusions of sedative analgesia prolong the duration of mechanical ventilation and increase the risk of complications. Numerous scoring systems exist that assess depth of sedation, with current guidelines supporting the use of the RASS (Richmond Agitation-Sedation Scale) [1]. Studies highlight that caring for more awake patients is more demanding and increases workload for healthcare staff [2,3]). A preliminary audit was undertaken on our General Intensive Care Unt (GICU) in 2014–2015 to identify whether patients were optimal sedated, as determined by RASS. Objectives: As a result of the previous audit three main areas for improvement were identified. 1. To achieve optimum levels of sedation as determined by RASS. 2. To achieve a gradual reduction in sedation to achieve spontaneous breathing on the ventilator at the earliest opportunity. 3. To achieve predictable and timely sedation holds.

Objectives: As a result of the previous audit three main areas for improvement were identified. 1. To achieve optimum levels of sedation as determined by RASS. 2. To achieve a gradual reduction in sedation to achieve spontaneous breathing on the ventilator at the earliest opportunity. 3. To achieve predictable and timely sedation holds. Methods: A ´targeted RASS´ tool was developed, which was simple, colour coded and kept at the patient bed-side. A protocol for gathering data from audited sedation days was developed. The protocol was instigated over a 5 week period from 4/1/2016 to 5/2/2016. All sedated patients were included, except for those having treatment withdrawn. On each patient, data was collected at the beginning and end of each shift. Including baseline and target RASS, times of RASS allocation made and sedation drug usage. Data was collated on a spreadsheet within the ICU. Results: Total number of admissions were 156 patients, total number of sedated, ventilated days were 46. 38 out of the 46 patients had targets set during ward rounds (82 %) There was no difficulty allocating patients to groups. 11 patients were allocated to RED which achieved 100 % compliance. 13 patients were allocated to AMBER which achieved 63 % compliance. Reduction in sedation drugs and RASS at the end of each shift was, on average, reduced. 14 patients were allocated to GREEN which achieved a 32 % compliance. Sedation drugs were largely stopped or reduced. On average the RASS was lower. Successful extubation occurred in 5 patients. Agitation leading to re-sedation occurred in 8 patients.

13 patients were allocated to AMBER which achieved 63 % compliance. Reduction in sedation drugs and RASS at the end of each shift was, on average, reduced. 14 patients were allocated to GREEN which achieved a 32 % compliance. Sedation drugs were largely stopped or reduced. On average the RASS was lower. Successful extubation occurred in 5 patients. Agitation leading to re-sedation occurred in 8 patients. Conclusion: A simple, clear and effective tool for communication between medical and nursing staff can improve practice and be translated into a real reduction in sedation drug administration and lower levels of sedation. References 1. Barr, J et al. 2013, Clinical Practice guidelines for the management of pain, agitation and delirium in adult patients in the intensive care medicine.41.263 2. Everingham et al. 2014 ´Targeting sedation: the lived experience of the intensive care nurse. Journal of clinical nursing 23 3. Tingsvik et al. 2013 Meeting the challenge: ICU nurses´s experiences of lightly sedated patients. Australian Critical Care 26

1. Barr, J et al. 2013, Clinical Practice guidelines for the management of pain, agitation and delirium in adult patients in the intensive care medicine.41.263 2. Everingham et al. 2014 ´Targeting sedation: the lived experience of the intensive care nurse. Journal of clinical nursing 23 3. Tingsvik et al. 2013 Meeting the challenge: ICU nurses´s experiences of lightly sedated patients. Australian Critical Care 26 A423 EEG-based delirium monitoring: a prospective, multicentre validation study T. Numan1, M. van den Boogaard2, A.M. Kamper3, P. Rood2, L.M. Peelen1, P.M. Zeman4, A.J. Slooter1 1University Medical Center Utrecht, Intensive Care Center, Utrecht, Netherlands; 2Radboud University Nijmegen Medical Centre, Intensive Care Medicine, Nijmegen, Netherlands; 3Isala Zwolle, Geriatrics, Zwolle, Netherlands; 4abvsciences, Vancouver, Canada Correspondence: A.J. Slooter _ University Medical Center Utrecht, Intensive Care Center, Utrecht, Netherlands Introduction: Delirium is common in Intensive Care Unit (ICU) patients and negatively affects outcome. However, recognition of delirium by ICU physicians is poor [1]. To improve detection, delirium screening tools have been developed, but these are subjective and insensitive in routine, daily practice [2]. Previously, we showed that patients with and without delirium could very well be distinguished with one minute of bipolar electroencephalography (EEG) and relative delta power as an indicator of slowing of background activity [3]. Based on these findings, a prototype EEG-based delirium monitor was built, with real-time artefact removal and an improved detection algorithm.

elirium could very well be distinguished with one minute of bipolar electroencephalography (EEG) and relative delta power as an indicator of slowing of background activity [3]. Based on these findings, a prototype EEG-based delirium monitor was built, with real-time artefact removal and an improved detection algorithm. Objectives: To validate an EEG-based delirium monitor in patients at risk of delirium. Methods: In this ongoing multicentre study, we will include 154 frail, surgical patients aged ≥60 years. Preoperatively (T-1) and on postoperative day 1 to 3 an EEG recording is performed together with a standardized cognitive assessment that is a video-taped. Diagnosis of delirium is based on the video by two, or in case of disagreement, three delirium experts (psychiatrists, geriatricians or neurologists) independent of each other, and independent of the EEG recording. EEGs are automatically analysed after artefact removal using waveform analyses, as well as analysis of relative delta power, and compared with the classification of the delirium experts (gold standard), to obtain sensitivity, specificity and the predictive values. Preliminary results: Among 45 patients (75 % men, mean age 76 years, SD 7 years), 5 patients were diagnosed as delirious. The figure shows the development of delirium in an example patient and the findings of the waveform analysis on the consecutive days after surgery. Sensitivity and specificity of the waveform- and relative delta power analyses were promising, and final results will be presented at the ESICM meeting.

nosed as delirious. The figure shows the development of delirium in an example patient and the findings of the waveform analysis on the consecutive days after surgery. Sensitivity and specificity of the waveform- and relative delta power analyses were promising, and final results will be presented at the ESICM meeting. Conclusions: Our findings suggest that EEG-based monitoring is promising for objective and reliable detection of delirium in routine daily practice. References 1. Eijk, M. M. J. van et al. Comparison of delirium assessment tools in a mixed intensive care unit. Crit. Care Med.37, 1881–5 (2009). 2. Eijk, M. M. J. van et al. Routine use of the confusion assessment method for the intensive care unit: a multicenter study. Am J Respir Crit Care Med.184, 340–4 (2011). 3. Van der Kooi, A. W. et al. Delirium detection using EEG: what and where to measure. Chest147, 94–101 (2015).Fig. 11 (abstract A423). Classification by EEG and expert

2. Eijk, M. M. J. van et al. Routine use of the confusion assessment method for the intensive care unit: a multicenter study. Am J Respir Crit Care Med.184, 340–4 (2011). 3. Van der Kooi, A. W. et al. Delirium detection using EEG: what and where to measure. Chest147, 94–101 (2015).Fig. 11 (abstract A423). Classification by EEG and expert A424 Unsuspected serotonin toxicity in the ICU C.E. van Ewijk1, G.E. Jacobs2,3, A.R.J. Girbes1 1VU University Medical Center, Intensive Care, Amsterdam, Netherlands; 2VU University Medical Center, Psychiatry, Amsterdam, Netherlands; 3Centre for Human Drugs Research, Leiden, Netherlands Correspondence: C.E. van Ewijk - VU University Medical Center, Intensive Care, Amsterdam, Netherlands Introduction: Delirium is a frequently occurring syndrome in patients admitted to the Intensive Care Unit (ICU) or Medium Care Unit (MCU), yet the pathophysiology remains poorly understood. An excess of central serotonin can lead to an altered mental status, associated with autonomic hyperactivity, and neuromuscular excitation. (Boyer EW et al. 2005, Buckley NA et al. 2014) Drugs with serotonergic properties are frequently and for prolonged periods administered to ICU/MCU patients. Therefore central serotonergic toxicity may constitute a predisposing, contributing or precipitating factor in the emergence of delirium.

neuromuscular excitation. (Boyer EW et al. 2005, Buckley NA et al. 2014) Drugs with serotonergic properties are frequently and for prolonged periods administered to ICU/MCU patients. Therefore central serotonergic toxicity may constitute a predisposing, contributing or precipitating factor in the emergence of delirium. Purpose: to determine the number of patients admitted to the ICU or MCU who are diagnosed with delirium, and who show characteristics of serotonin toxicity (Sternbach 1991, Dunkley EJC et al. 2003) in association with the administration of serotonergic drugs. Methods: During a 10-week prospective observational cohort study in the ICU and MCU, patients aged 18 or older, diagnosed with delirium in the ICU or MCU were included. Patients were considered as delirious in case of a positive CAM-ICU and/or at the start of haloperidol prescription on suspicion of delirium. Once included, patients were screened for recent administered serotonergic drugs and screened for physical signs associated with serotonin toxicity by a standardized physical examination by a specifically trained physician. Results: A total of 61 patients diagnosed with delirium were enrolled. In 44 out of 61 patients (72,1 %) the use of drugs potentially contributing to serotonergic toxicity was recorded. Out of 44 patients, 7 (16 %) patients showed physical signs of serotonin toxicity and in addition met the Hunter serotonin toxicity criteria, suggesting the presence of serotonergic toxicity. None of these patients were recognized as such by the treating physicians.

y contributing to serotonergic toxicity was recorded. Out of 44 patients, 7 (16 %) patients showed physical signs of serotonin toxicity and in addition met the Hunter serotonin toxicity criteria, suggesting the presence of serotonergic toxicity. None of these patients were recognized as such by the treating physicians. CONCLUSIONS: A significant proportion of delirious patients in the ICU might in fact be classified as suffering from central serotonin toxicity. The awareness of potential serotonin toxicity is low among physicians. References 1. Boyer EW, Shannon M (2005) The serotonin syndrome. N Engl J Med 352:1112–20. 2. Buckley NA et al. (2014) Serotonin syndrome. Bmj 1626(February):15–8. 3. Sternbach H (1991) The Serotonin Syndrome. Am J Psychiatry 148(June):705–13 4. Dunkley EJC et al. (2003) The hunter serotonin toxicity criteria: Simple and accurate diagnostic decision rules for serotonin toxicity. QJM - Mon J Assoc Physicians 96:635–42. A425 Prospective study to determine the incidence and risk factors for delirium in cancer patients in ICU S.N. Myatra, M.M. Harish, N.R. Prabu, S. Siddiqui, A.P. Kulkarni, J.V. Divatia Tata Memorial Hospital, Mumbai, India Correspondence: MM M. Harish - Tata Memorial Hospital, Mumbai, India Introduction: Delirium in ICU associated with both short and long term adverse outcomes. Cancer patients admitted to ICU have several risk factors to develop delirium. Most published data on delirium are about general ICU patients and there are no large studies in cancer patients in ICU.

emorial Hospital, Mumbai, India Introduction: Delirium in ICU associated with both short and long term adverse outcomes. Cancer patients admitted to ICU have several risk factors to develop delirium. Most published data on delirium are about general ICU patients and there are no large studies in cancer patients in ICU. Objectives: We conducted prospective study to know the incidence and risk factors associated with delirium in cancer patients. Methods: Adult cancer patients in ICU were included over a period of 6 months. Demographic data, type/stage of cancer, treatment details including drugs administered were noted. Hemodynamic status, mechanical ventilation, presence of sepsis, patient's sleep and mobilization status were also noted for the assessment of risk factors associated with delirium. Sedation was scored using the Richmond Agitation-Sedation Scale (RASS) and delirium was assessed using the Confusion Assessment Method (CAM-ICU) per shift by the bedside nurses, if they were responsive to verbal commands (RASS score of −3 or lighter level of sedation). If delirium was positive even in one shift in a day, it was considered as a delirium day.

Agitation-Sedation Scale (RASS) and delirium was assessed using the Confusion Assessment Method (CAM-ICU) per shift by the bedside nurses, if they were responsive to verbal commands (RASS score of −3 or lighter level of sedation). If delirium was positive even in one shift in a day, it was considered as a delirium day. Results: Patients with comorbidities showed higher incidence of ICU delirium as compared to those without {73(63.5 %) vs 42(36.5 %)}. Tobacco consumers showed higher incidence of delirium as compared to those without {85(73.1 %) vs 30(26.1 %)}. Recent chemotherapy showed positive correlation with delirium as compared to those who did not receive chemotherapy or received it more than one month back {63(54.8 %) vs 51(44.3 %) vs 1(0.9 %)}. Sleep deprived patients had higher incidence of delirium as compared to patients without {90(78.3 %) vs 25(21.7 %)}. There was a significant association of Midazolam with delirium as compared to those who did not receive the drug{104(90.4 %) vs 11(9.6 %), p-0.000}. Patients who had delirium had median length of stay 7 days as compared to 3 days among without delirium. On multivariate analysis of the risk factors comorbidity(P-0.003,OR-13.3), tobacco consumption(P-0.015,OR-5.95), Midazolam(P-0.000,OR-47.50), mechanical ventilation(P-0.022,OR-9.09) and sleep deprivation (P-0.030,OR-7.18) showed statistically significant association with the delirium. Delirium also showed significant association with increased length of both ICU (P-0.000,OR-1.73) and hospital stay (P-0.004,OR-1.02).

015,OR-5.95), Midazolam(P-0.000,OR-47.50), mechanical ventilation(P-0.022,OR-9.09) and sleep deprivation (P-0.030,OR-7.18) showed statistically significant association with the delirium. Delirium also showed significant association with increased length of both ICU (P-0.000,OR-1.73) and hospital stay (P-0.004,OR-1.02). Conclusion: Incidence of ICU delirium among cancer patients was 70.6 %. Recent chemotherapy mechanical ventilation, Midazolam, comorbidities, tobacco and sleep deprivation were the independent predictor of ICU delirium in cancer patients. ICU delirium increase both ICU and hospital length of stay. References 1. American Psychiatric Association, Diagnostic and Statistical Manual, 4th ed, APA Press, Washington, DC 1994. 2. Ely ICM 2001;27,1892-1900, Ely JAMA 2004;291:1753–1762, 3. Lim SM, CCM 2004;32:2254–2259 A426 Characteristics of critical patients with and without delirium submitted to mobilize early active L.D. Murbach, M.A. Leite, E.F. Osaku, C.R.L.M. Costa, M. Pelenz, N.M. Neitzke, M.M. Moraes, J.L. Jaskowiak, M.M.M. Silva, R.S. Zaponi, L.R.L. Abentroth, S.M. Ogasawara, A.C. Jorge, P.A.D. Duarte Western Parana State University Hospital, Cascavel, Brazil Correspondence: L.D. Murbach - Western Parana State University Hospital, Cascavel, Brazil Introduction: Delirium is defined as an acute cognitive impairment, inattention and disorganized thoughts, may be associated with the use of benzodiazepines, opioid administration, mechanical ventilation duration and immobilization, the management of it depends on multidisciplinary team strategies.

al, Cascavel, Brazil Introduction: Delirium is defined as an acute cognitive impairment, inattention and disorganized thoughts, may be associated with the use of benzodiazepines, opioid administration, mechanical ventilation duration and immobilization, the management of it depends on multidisciplinary team strategies. Objective: The purpose of this study was to compare the results of patients with and without Delirium that were submitted to an early mobilization protocol in the Intensive Care Unit (ICU). Methods: Cohort study in the general ICU of a University Hospital, Southern Brazil. The study was conducted between February and December 2013. Were excluded patients using mechanical ventilation for less than 24 hours, under 18 years old, patients submitted to passive exercises or who did not have early mobilization. Only patients that received active and active-assisted early mobilization were included. Patients were divided between Delirium and no-Delirium groups. Data were reported as mean and standard deviation. The variables were compared through of Mann–Whitney test, t-test parametric and the statistical significance level was 5 % (p ≤ 0.05). Results: In the period were admitted 474 patients of which 127 were included. The incidence of Delirium was 48 %. The main causes of admission in groups with delirium and without delirium were medical non-neurological (34 % vs. 33 %), medical neurological (22 % vs. 9 %) and Traumatic Brain Injury (22 % vs. 26 %). There were no significant. The data are demonstrated in Tables 5 and 6.

e included. The incidence of Delirium was 48 %. The main causes of admission in groups with delirium and without delirium were medical non-neurological (34 % vs. 33 %), medical neurological (22 % vs. 9 %) and Traumatic Brain Injury (22 % vs. 26 %). There were no significant. The data are demonstrated in Tables 5 and 6. Conclusion: In this cohort study, we found no differences in the outcomes of patients with or without delirium submitted to early active mobilization. Most of patients had good response to an early mobilization protocol.Table 5 (abstract A426). ᅟ Delirium (61) Non Delirium (66) p-value Male (n-%) 43 (70) 43 (65) Age (n-%) 49±18.25 43±19.55 0.10 APACHE II 24±4.94 24 ± 6.51 0.49 SOFA 12±3.28 11 ± 3.73 0.17 Death ICU (n-%) 2 (3) 8 (13) 0.131 Table 6 (abstract A426). ᅟ Delirium (61) Non Delirium (66) p-value MV (hours) 179±124.09 175±183.61 0.26 Sedation time (hours) 100±80.52 108±122.18 0.62 Midazolam (mg) 2532±3168.53 3258±5828.26 0.68 Fentanyl (μg) 45804±70319.75 41422±84192.46 0.25 Propofol (mg) 3289±9146.06 4570±9392.23 0.89 Haloperidol (mg) 39±38.85 54±40.33 0.15 Glasgow Coma Scale - ICU discharge 14±1.42 13±1.92 0.57 Functional independence measure 61±31.34 68±30.03 0.40

rs) 100±80.52 108±122.18 0.62 Midazolam (mg) 2532±3168.53 3258±5828.26 0.68 Fentanyl (μg) 45804±70319.75 41422±84192.46 0.25 Propofol (mg) 3289±9146.06 4570±9392.23 0.89 Haloperidol (mg) 39±38.85 54±40.33 0.15 Glasgow Coma Scale - ICU discharge 14±1.42 13±1.92 0.57 Functional independence measure 61±31.34 68±30.03 0.40 A427 Incidence and risk factors of delirium in an oncological intensive care unit N. Hernández-Sánchez, L.A. Sánchez-Hurtado, F.J. García-Guillen, S.A. Ñamendys-Silva Instituto Nacional de Cancerología, Department of Critical Care Medicine, Mexico, Mexico Correspondence: N. Hernández-Sánchez -Instituto Nacional de Cancerología, Department of Critical Care Medicine, Mexico, Mexico Introduction: The presence of delirium in the intensive care unit (ICU) is often underestimated due to its variable clinical presentation. The incidence and risk factors of delirium have not been studied in critically ill cancer patients. Objective: To evaluate the incidence and risk factors of delirium in critically ill cancer patients. Methods: The present study was an observational and descriptive study that included 69 critically ill cancer patients admitted to ICU of the Instituto Nacional de Cancerología of México, between December 2015 and March 2016. Delirium was diagnosed by daily evaluation using the Confusion Assessment Method for the ICU (CAM-ICU). The clinical data was obtained from the patient´s medical record. No interventions were done.

cancer patients admitted to ICU of the Instituto Nacional de Cancerología of México, between December 2015 and March 2016. Delirium was diagnosed by daily evaluation using the Confusion Assessment Method for the ICU (CAM-ICU). The clinical data was obtained from the patient´s medical record. No interventions were done. Results: Of the 69 critically ill cancer patients included in this study, 35 (50.7 %) were men; with a median age of 49.55 years; 2 (7.2 %) had history of delirium, and 2 (2.9 %) had delirium at ICU admission. Solid tumors were the most frequent malignances 49 (71 %), and 20 (29 %) had hematological malignancies. The principal admission diagnosis to ICU was septic shock in 27 (39.1 %). During the first 24 hours in ICU, 20 (68.1 %) required mechanical ventilation, 25 (36.1 %) received midazolam as a sedative, and 47 (68.1 %) fentanyl as an analgesic.

t frequent malignances 49 (71 %), and 20 (29 %) had hematological malignancies. The principal admission diagnosis to ICU was septic shock in 27 (39.1 %). During the first 24 hours in ICU, 20 (68.1 %) required mechanical ventilation, 25 (36.1 %) received midazolam as a sedative, and 47 (68.1 %) fentanyl as an analgesic. The incidence of delirium was 15.9 %. The hyperactive type delirium was the most common. There was no difference between risk factors (sex, age, comorbidities, history of delirium, use of analgesic, sedation or corticosteroids, and severity of critical illness) in patients who presented delirium and those who did not. The most important independent risk factors for delirium were the total number of ventilated days during ICU stay (P = 0.009) and the length of stay in ICU (P < 0.001). The following variables were independent predictors of delirium: the total number of ventilated days during an ICU stay, relative risk (RR) 1.09 (95 % confidence interval (CI); 0.98 to 1.22) and the length of stay in ICU, RR: 1.173 (95 % CI; 1.04 to 1.32). The ICU mortality was 10.1 %, no patients with delirium died. Conclusion: The incidence of delirium in the oncologic ICU is low, the length of stay in ICU and the total number of ventilated days during an ICU stay were the most important risk factors for its presentation; There is no relation between delirium and ICU mortality. References 1. Reade M. et al. Sedation and delirium in the intensive care unit. N Engl J Med 2014; 370:444–454

Conclusion: The incidence of delirium in the oncologic ICU is low, the length of stay in ICU and the total number of ventilated days during an ICU stay were the most important risk factors for its presentation; There is no relation between delirium and ICU mortality. References 1. Reade M. et al. Sedation and delirium in the intensive care unit. N Engl J Med 2014; 370:444–454 2. Uchida M. et al. Prevalence, course and factors associated with delirium in ederly patients with advanced cancer: a longitudinal observational study. Jpn J Clin Oncol, 2015, 45 (10): 934–940

Conclusion: The incidence of delirium in the oncologic ICU is low, the length of stay in ICU and the total number of ventilated days during an ICU stay were the most important risk factors for its presentation; There is no relation between delirium and ICU mortality. References 1. Reade M. et al. Sedation and delirium in the intensive care unit. N Engl J Med 2014; 370:444–454 2. Uchida M. et al. Prevalence, course and factors associated with delirium in ederly patients with advanced cancer: a longitudinal observational study. Jpn J Clin Oncol, 2015, 45 (10): 934–940 A428 Comparison of opium tincture and methadone on pain and agitation control in inhalational opium addicted patients admitted to critical care unit: a pilot study B. Maghsoudi1, M. Emami1, M.B. Khosravi1, F. Zand1, H.R. Tabatabaie2, M. Masjedi1, G. Sabetiyan1, A. Mokri3, Nurses of the Central and General ICUs of Shiraz Namazi Hospital 1Shiraz University of Medical Sciences, Shiraz Anesthesiology and Critical Care Research Center, Shiraz, Islamic Republic of Iran; 2Department of Epidemiology, School of Public Health - Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran; 3National Center for Addiction Studies, Tehran, Islamic Republic of Iran Correspondence: B. Maghsoudi - Shiraz University of Medical Sciences, Shiraz Anesthesiology and Critical Care Research Center, Shiraz, Islamic Republic of Iran Introduction: Addiction is one of the main problems in our society, Iran,about 3 % of the whole population. Many of critically ill patients needing intensive care units (ICU)admission suffer from underlying disease such as ischemic heart disease, hypertension, diabetes and so on, and are opium addicted too. Consequently, the control of pain and agitation in the intensive care unit plays an important role in early discharge and improving patient´s outcome.

needing intensive care units (ICU)admission suffer from underlying disease such as ischemic heart disease, hypertension, diabetes and so on, and are opium addicted too. Consequently, the control of pain and agitation in the intensive care unit plays an important role in early discharge and improving patient´s outcome. Objectives: Comparing opium tincture and methadone on the control of pain, agitation, and ICU outcomes in opium addicted ICU admitted patients. Methods: In this randomized pilot study, a total of 50 inhalant opium addicted patients aged 18–84 years, enrolled into the study and they were divided into two groups of 25, the opium tincture(T) and methadone group(M). Patients received morphine infusion according to needs before allowing enteral drugs (at most one day). After that(as day one of study), in the T group, patients received 10 cc opium tincture in dark tea(total 50 cc) through nasogastric (NGT) tube and 10 cc normal saline intravenously (IV) every 8 hours. The M group received 50 cc dark tea via NGT plus 10 mg mehadone diluting to 10 cc IV every 8 hours.Then patient´s pain and agitation were measured and compared based on (BPS) behavioral pain scale and Richmond agitation sedation scale (RASS), and delirium by CAM-ICU. If BPS scale was desirable (−0, 1), the same dose continued, but if it was ≤ −1 or ≥ 2, the doses was decreased or increased by 2.5 (cc or mg) respectively at next interval. Rescue medications were IV morphine, midazolam and haloperidol.

pain scale and Richmond agitation sedation scale (RASS), and delirium by CAM-ICU. If BPS scale was desirable (−0, 1), the same dose continued, but if it was ≤ −1 or ≥ 2, the doses was decreased or increased by 2.5 (cc or mg) respectively at next interval. Rescue medications were IV morphine, midazolam and haloperidol. Results: Current study shows that the use of opium tincture in the pain management(by BPS) of opium addicted ICU patients had significant desirable impact and, (p-value = 0.041) compared with methadone especially after deleting confounding variables (morphine, midazolam and haloperidol) (P-value = 0.019).But considering agitation (RASS) there was no difference before (p-value = 0.684) and after the deleting confounding variables (p-value = 0.630) between the two groups. Extubation time, duration of ICU and hospital stay was not different between the two groups. Conclusions: In inhalational opium addicted patients admitted to ICU enable of starting oral medications, pain is better controlled with Opium Tincture in comparison to Methadone. References 1. Comparison of tincture of opium and methadone to control opioid withdrawal in a Thai treatment centre.Jittiwutikarn J, Ali R, White JM, Bochner F, Somogyi AA, Foster DJ.Br J Clin Pharmacol. 2004 Nov;58(5):536–41 2. Flexible dosing of tincture of opium in the management of opioid withdrawal: pharmacokinetics and pharmacodynamics.Somogyi AA, Larsen M, Abadi RM, Jittiwutikarn J, Ali R, White JM. Br J Clin Pharmacol. 2008 Nov;66(5):640–7 Grant acknowledgement

1. Comparison of tincture of opium and methadone to control opioid withdrawal in a Thai treatment centre.Jittiwutikarn J, Ali R, White JM, Bochner F, Somogyi AA, Foster DJ.Br J Clin Pharmacol. 2004 Nov;58(5):536–41 2. Flexible dosing of tincture of opium in the management of opioid withdrawal: pharmacokinetics and pharmacodynamics.Somogyi AA, Larsen M, Abadi RM, Jittiwutikarn J, Ali R, White JM. Br J Clin Pharmacol. 2008 Nov;66(5):640–7 Grant acknowledgement We acknowledge Shiraz university of medical sciences and Iranian center for addiction Studies for their support. A429 Sevoflurane sedation in critically ill patients under extracorporeal membrane oxygenation J. Troubleyn, M. Diltoer, R. Jacobs, D.N. Nguyen, E. De Waele, J. De Regt, P.M. Honoré, V. Van Gorp, H.D. Spapen University Hospital, Vrije Universiteit Brussel, Brussels, Belgium Correspondence: J. Troubleyn - University Hospital, Vrije Universiteit Brussel, Brussels, Belgium Introduction: Patients with severe ARDS or intractable cardiogenic shock are increasingly treated with veno-venous or arterio-venous extracorporeal membrane oxygenation (ECMO). During ECMO, intravenous (IV) sedation with midazolam (MIDA) and/or propofol (PROP) may be inadequate because these drugs become highly sequestrated in the circuit. The volatile anaesthetic sevoflurane (SEVO) has been shown to be a potential alternative to IV sedation in cardiac surgery and general ICU patients. Objectives: We assessed the efficacy and safety of prolonged use of SEVO in a cohort of hemodynamically unstable critically ill patients under ECMO.

A429 Sevoflurane sedation in critically ill patients under extracorporeal membrane oxygenation J. Troubleyn, M. Diltoer, R. Jacobs, D.N. Nguyen, E. De Waele, J. De Regt, P.M. Honoré, V. Van Gorp, H.D. Spapen University Hospital, Vrije Universiteit Brussel, Brussels, Belgium Correspondence: J. Troubleyn - University Hospital, Vrije Universiteit Brussel, Brussels, Belgium Introduction: Patients with severe ARDS or intractable cardiogenic shock are increasingly treated with veno-venous or arterio-venous extracorporeal membrane oxygenation (ECMO). During ECMO, intravenous (IV) sedation with midazolam (MIDA) and/or propofol (PROP) may be inadequate because these drugs become highly sequestrated in the circuit. The volatile anaesthetic sevoflurane (SEVO) has been shown to be a potential alternative to IV sedation in cardiac surgery and general ICU patients. Objectives: We assessed the efficacy and safety of prolonged use of SEVO in a cohort of hemodynamically unstable critically ill patients under ECMO. Methods: Patients initially sedated with IV MIDA and/or PROP and then switched to SEVO, were retrospectively reviewed. The liquid form of SEVO was infused via a modified heat-moisture exchanger (AnaConDa®, Sedana Medical AB, Uppsala, Sweden) placed between the Y-piece and the endotracheal tube. SEVO was given as an initial bolus of 1.2 mL followed by a continuous infusion started at 5 mL/h. Infusion rate was adjusted to reach a measured end-tidal SEVO (ETS) concentration between 0.8 and 1.4 %. Sedation level aimed to obtain a RASS score of −3 to −4. Mean blood pressure and norepinephrine requirement were evaluated before start of SEVO and after respectively 2 h and 6 h of treatment. ETS was measured after stabilization, and at 24 h and 72 h. Wilcoxon signed rank test was used to compare variables over time. Values were expressed as means ± SD.

S score of −3 to −4. Mean blood pressure and norepinephrine requirement were evaluated before start of SEVO and after respectively 2 h and 6 h of treatment. ETS was measured after stabilization, and at 24 h and 72 h. Wilcoxon signed rank test was used to compare variables over time. Values were expressed as means ± SD. Results: Twenty-one patients (12 males, age 53 ± 15 years; 16 ARDS and 5 cardiogenic shock) were studied. Mean APACHE II score was 25 ± 5. MIDA and PROP dose before start of SEVO were respectively 2.1 ± 2.6 mg/min and 2.4 ± 0.9 μg/kg/min. Ten patients were curarized. Initiation of SEVO allowed immediate cessation of IV sedation and curarization. Opioid analgesia was continued unchanged. SEVO treatment was not associated with changes in MAP (73 ± 12 vs. 72 ± 10 mmHg; p = 0.56) or norepinephrine need (0.11 ± 0.06 vs. 0.13 ± 0.19; p = 0.83). As compared with baseline (1.1 ± 0.3 %), ETS remained unchanged after 24 h (1.2 ± 0.3 %) and 72 h (1.3 ± 0.4 %). Patients were treated for 13 ± 9 days. Twelve patients (57 %) survived. No adverse events were noted. Random 8 h-samples of ambient air in the bedside breathing zone revealed SEVO concentrations well below the National Institute for Occupational Safety and Health 2 ppm limit. Conclusions: In hemodynamically unstable patients under ECMO, prolonged SEVO sedation is an effective, well-tolerated, and safe alternative for "classic" IV MIDA and PROP-based sedation.

Results: Twenty-one patients (12 males, age 53 ± 15 years; 16 ARDS and 5 cardiogenic shock) were studied. Mean APACHE II score was 25 ± 5. MIDA and PROP dose before start of SEVO were respectively 2.1 ± 2.6 mg/min and 2.4 ± 0.9 μg/kg/min. Ten patients were curarized. Initiation of SEVO allowed immediate cessation of IV sedation and curarization. Opioid analgesia was continued unchanged. SEVO treatment was not associated with changes in MAP (73 ± 12 vs. 72 ± 10 mmHg; p = 0.56) or norepinephrine need (0.11 ± 0.06 vs. 0.13 ± 0.19; p = 0.83). As compared with baseline (1.1 ± 0.3 %), ETS remained unchanged after 24 h (1.2 ± 0.3 %) and 72 h (1.3 ± 0.4 %). Patients were treated for 13 ± 9 days. Twelve patients (57 %) survived. No adverse events were noted. Random 8 h-samples of ambient air in the bedside breathing zone revealed SEVO concentrations well below the National Institute for Occupational Safety and Health 2 ppm limit. Conclusions: In hemodynamically unstable patients under ECMO, prolonged SEVO sedation is an effective, well-tolerated, and safe alternative for "classic" IV MIDA and PROP-based sedation. A430 A pilot study using bispectral index (BIS) to adjust sedation in patients with no neurological pathology in elderly critical care patients R.S. Contreras, N.D. Toapanta, G. Moreno, J. Sabater, H. Torrado, M. Gonzalez, M. Marin, E. Farigola, A. Gonzalez, J. Fernandez, A. Vera, X. Gisbert, C. Juliá, J. Uya, L. Corral, Sedation an Delirium Group Hospital Universitari de Bellvitge Hospital Universitari de Bellvitge, Intensive Care Medicine, Barcelona, Spain Correspondence: R.S. Contreras - Hospital Universitari de Bellvitge, Intensive Care Medicine, Barcelona, Spain Introduction: Patients admitted to the intensive care unit (ICU) usually require use of hypnotics and sedatives to ensure comfort and proper adaptation to mechanical ventilation. An important requirement for an adequate sedation is frequent and proper assessment of its depth. Inadequate sedation can lead to problems of over-sedation, under-sedation and/or delirium in ICU, especially in elderly patients.

f hypnotics and sedatives to ensure comfort and proper adaptation to mechanical ventilation. An important requirement for an adequate sedation is frequent and proper assessment of its depth. Inadequate sedation can lead to problems of over-sedation, under-sedation and/or delirium in ICU, especially in elderly patients. Objective: To compare the total dose of sedative use and the rate of over-sedation in patients over 65 years admitted to the ICU, adjusting sedation by monitoring with BIS® versus monitoring with the exclusive use of sedation scales. Methods: A randomized, clinical trial including patients over 65 years who were admitted to the ICU affected with medical or surgical pathology of non neurological etiology who required sedation for more than 24 hours to maintain adaptation to mechanical ventilation. Patients were randomized into two groups: the intervention group using BIS monitoring to adjust sedation in order to maintain values between 50–60 and; the control group in which sedation was adjusted with the exclusive use of Richmond Agitation-Sedation Scale (RASS) to maintain RASS −2. The study was approved by the institution's Research Ethics Committee. Statistical analysis: We used the chi square test to compare categorical data and proportions, and the T-test, Mann–Whitney test or Kruskal-Wallis, as appropriate, to compare continuous variables. An alpha level of 0.05 was used to determine statistical significance. All data in the present study was analyzed using SPSS version 22.0 (SPSS Inc, Chicago, USA).

uare test to compare categorical data and proportions, and the T-test, Mann–Whitney test or Kruskal-Wallis, as appropriate, to compare continuous variables. An alpha level of 0.05 was used to determine statistical significance. All data in the present study was analyzed using SPSS version 22.0 (SPSS Inc, Chicago, USA). Results: Until now 56 patients have been randomized. 27 (48 %) patients in the control group and 29 (52 %) in the intervention group. Related to the general characteristics, demographics (age, weight, height, gender), medical history (hypertension, diabetes, cognitive impairment, lung disease and chronic renal failure), pathology and admission severity scores (APACHE II and SOFA) there were not statistically significant differences between groups. Besides analgesia; midazolam, propofol or remifentanil were also used in both groups (without significant differences). There were not statistically significant differences in midazolam and propofol, total daily dose, days of mechanical ventilation, ventilator associated pneumonia, delirium, ICU mortality and ICU length of stay. Conclusions: No statistically significant differences were observed between the groups of this pilot study using BIS to adjust the sedation. We attribute these findings mainly to the lack of adherence to the protocol. References 1. Sedation and analgesia in the critically ill adult. Fraser GL, Riker RR - Curr Opin Anaesthesiol - April 1, 2007; 20 (2); 119–23 2. Acknowledgment

Conclusions: No statistically significant differences were observed between the groups of this pilot study using BIS to adjust the sedation. We attribute these findings mainly to the lack of adherence to the protocol. References 1. Sedation and analgesia in the critically ill adult. Fraser GL, Riker RR - Curr Opin Anaesthesiol - April 1, 2007; 20 (2); 119–23 2. Acknowledgment We acknowledge ICU physicians, nurses and personnel for their contribution and especially to Antonio Diaz-Prieto for his contribution to this study designing the database program. A431 An audit of sedation use and sedation scores in intensive care I. Elias-Jones, L. Gemmell, A. MacKay Queen Elizabeth University hospital, Anaesthetics and Intensive Care, Glasgow, UK Correspondence: I. Elias-Jones - Queen Elizabeth University hospital, Anaesthetics and Intensive Care, Glasgow, UK Introduction: How much sedation is given, and for how long, is important in determining patient outcome as both over sedation and under sedation can have potentially deleterious consequences. Over sedation can increase time on ventilator support and prolong ICU duration of stay. Under sedation can cause hyper-catabolism, immunosupporession, hypercoagulability and increase sympathetic activity.

determining patient outcome as both over sedation and under sedation can have potentially deleterious consequences. Over sedation can increase time on ventilator support and prolong ICU duration of stay. Under sedation can cause hyper-catabolism, immunosupporession, hypercoagulability and increase sympathetic activity. Objectives: The aim of this audit was to look at our use of sedation in a large 18bed teaching unit in Glasgow. We aimed to look at propofol dosage, to particularly look if patients exceeded their maximum hourly dose, sedation scoring tools and whether it was regularly assessed, the addition of other sedative agents and when, and whether sedation load was associated with increased ventilator days. Methods: We looked at a snapshot of 40patients picked at random admitted to the Queen Elizabeth University hospital, Glasgow over the previous six months. The Richmond-Agitation Sedation scale was used as the sedation scoring tool.

Objectives: The aim of this audit was to look at our use of sedation in a large 18bed teaching unit in Glasgow. We aimed to look at propofol dosage, to particularly look if patients exceeded their maximum hourly dose, sedation scoring tools and whether it was regularly assessed, the addition of other sedative agents and when, and whether sedation load was associated with increased ventilator days. Methods: We looked at a snapshot of 40patients picked at random admitted to the Queen Elizabeth University hospital, Glasgow over the previous six months. The Richmond-Agitation Sedation scale was used as the sedation scoring tool. Results: Data was collected for 40patients. 57.5 % were male. The median age was 57 years. Weight and height were not documented for 21patients. 6 patients were given in excess of the recommended dose of propofol according to weight (mg/kg/hr). These were six of the patients in which weight was actually documented so exceeding maximum recommended daily dosing could have been more of an issue than is described. All patients were sedated with propofol and an opiate, either remifentanil or alfentanil. Half of the patients were found to be over sedated in the time period, with RASS scores of −4 or −5. Six out of the forty patients did not have sedation scores documented daily as part of their care. When other agents were considered, they were considered relatively late in the patient stay (days 4–5). CK and triglycerides were never checked when propofol had been used in high doses for a significant length of time. Average length of ventilation was 4 days (12 hours - 12 days) with unsurprisingly higher propofol use seen in the patients ventilated for longer periods of time. Excess sedation was more common in the younger patient with a past history of drug or alcohol dependence.

used in high doses for a significant length of time. Average length of ventilation was 4 days (12 hours - 12 days) with unsurprisingly higher propofol use seen in the patients ventilated for longer periods of time. Excess sedation was more common in the younger patient with a past history of drug or alcohol dependence. Conclusions: Delirium and over sedation is associated with an increase in morbidity and mortality. As can be seen from our snapshot of data, we are poor at documenting daily sedation scores, with a significant proportion of our patients being documented as over sedated. Propofol use is excessive, and could perhaps lead to iatrogenic complications and morbidity. Patients who were heavily sedated were ventilated for longer. We are now working on a sedation protocol within our unit to improve our practice. References 1. K.Rowe. Sedation in the Intensive Care Unit. CEACCP 2008:8(2): 50–55

Conclusions: Delirium and over sedation is associated with an increase in morbidity and mortality. As can be seen from our snapshot of data, we are poor at documenting daily sedation scores, with a significant proportion of our patients being documented as over sedated. Propofol use is excessive, and could perhaps lead to iatrogenic complications and morbidity. Patients who were heavily sedated were ventilated for longer. We are now working on a sedation protocol within our unit to improve our practice. References 1. K.Rowe. Sedation in the Intensive Care Unit. CEACCP 2008:8(2): 50–55 Perioperative intensive care and fluid management A432 Peri operative use of plasmalyte in kidney transplant recipients reduces the incidence of RRT in the immediate post-operative period D. Randall1, A. Adwaney1, M. Blunden1, J.R. Prowle1,2, C.J. Kirwan1,2 1Barts Health NHS Trust, Renal and Transplant Medicine, London, UK; 2Barts Health NHS Trust, Adult Critical Care, London, UK Correspondence: C.J. Kirwan - Barts Health NHS Trust, Renal and Transplant Medicine, London, UK Introduction: There is evidence that the type of intravenous fluid used in patients who are critically ill or who have undergone major surgery may have an impact on the progression and recovery of Acute Kidney Injury. Patients undergoing renal transplantation are not specifically included in this work. Our centre changed the general peri-operative fluid regimen from routine use of normal saline (NS) to Plasmalyte (balanced solution containing physiological sodium, chloride, potassium, magnesium with an acetate buffer), including all patients undergoing renal transplantation. We hypothesised that this change would reduce peri-operative hyperkalaemia and thus reduce the need for immediate renal replacement therapy (RRT).

asmalyte (balanced solution containing physiological sodium, chloride, potassium, magnesium with an acetate buffer), including all patients undergoing renal transplantation. We hypothesised that this change would reduce peri-operative hyperkalaemia and thus reduce the need for immediate renal replacement therapy (RRT). Methods: The peri-operative fluid management guidance was changed in March 2013. We performed a single centre retrospective observational study of consecutive renal transplant recipient's pre and post policy change to assess its impact on transplant patients. The local renal database, patient notes, a computer based pathology system and clinic letters were used to collect data. We excluded patients with incomplete data of the peri-operative period, who had not received the standard fluid regimen of the time and those with an immediate surgical complication that rendered dialysis inevitable. Results: There were 47 transplant recipients in the NS group and 29 in the Plasmalyte group. The percentage of patients not requiring RRT in the first 48 hours post operatively in the Plasmalyte group were significantly higher than in the NS group (D1 97 vs 81 % p 0.046, D2 94 vs. 72 %, p 0.025) (Fig. 12). There were no cases of hyperkalaemia in the immediate post operative period for patients who had received Plasmalyte (P < 0.0001) (Fig. 13).

iring RRT in the first 48 hours post operatively in the Plasmalyte group were significantly higher than in the NS group (D1 97 vs 81 % p 0.046, D2 94 vs. 72 %, p 0.025) (Fig. 12). There were no cases of hyperkalaemia in the immediate post operative period for patients who had received Plasmalyte (P < 0.0001) (Fig. 13). A greater urine output in the first 24 hours after surgery was observed in the plasmalyte group, 2195mls (95 % CI:1423–2966) vs. 913.8mls (95 % CI:563.3-1264, P < 0.0001) which may indicate earlier onset of recovery from the post transplant AKI. Baseline characteristics were similar between each group except there were a higher proportion of DCD donors in the NS group compared to the Plasmalyte group (36 % vs. 17 %, p < 0.001). Conclusions: Renal transplant recipients who did not have an immediate surgical complication and received only Plasmalyte in the peri-operative period, had a lower incidence of RRT in the first 48 hours following surgery than those who received normal saline. This may be attributed to the lower incidence of immediate hyperkalaemia in the Plasmalyte group.Fig. 12 (abstract A432). ᅟ Fig. 13 (abstract A432). ᅟ

Conclusions: Renal transplant recipients who did not have an immediate surgical complication and received only Plasmalyte in the peri-operative period, had a lower incidence of RRT in the first 48 hours following surgery than those who received normal saline. This may be attributed to the lower incidence of immediate hyperkalaemia in the Plasmalyte group.Fig. 12 (abstract A432). ᅟ Fig. 13 (abstract A432). ᅟ A433 Enhanced recovery after surgery plus reduces pulmonary morbidity after major surgery - 1 year on N. Thomas1, A. Martin2, H. Owen2, L. Darwin2, D. Conway1, D. Atkinson1, M. Sharman1, J. Moore1 1Central Manchester Foundation NHS Trust, Manchester, UK; 2Health Education North West, Manchester, UK Correspondence: N. Thomas - Central Manchester Foundation NHS Trust, Manchester, UK Introduction: Pulmonary complications after major surgery are common affecting up to 30 % of patients, increasing hospital length of stay, short and long-term mortality1. Despite this, very little specific action is taken to prevent post-operative pulmonary complications (PPC) within major surgery programmes. Even the Enhanced recovery after surgery [ERAS] programme, does not specifically address this. Objectives: Central Manchester Foundation Trust (CMFT) successfully implemented ERAS+ aimed at reducing PPC in 2014 and we now have results to show its sustained benefit.

A433 Enhanced recovery after surgery plus reduces pulmonary morbidity after major surgery - 1 year on N. Thomas1, A. Martin2, H. Owen2, L. Darwin2, D. Conway1, D. Atkinson1, M. Sharman1, J. Moore1 1Central Manchester Foundation NHS Trust, Manchester, UK; 2Health Education North West, Manchester, UK Correspondence: N. Thomas - Central Manchester Foundation NHS Trust, Manchester, UK Introduction: Pulmonary complications after major surgery are common affecting up to 30 % of patients, increasing hospital length of stay, short and long-term mortality1. Despite this, very little specific action is taken to prevent post-operative pulmonary complications (PPC) within major surgery programmes. Even the Enhanced recovery after surgery [ERAS] programme, does not specifically address this. Objectives: Central Manchester Foundation Trust (CMFT) successfully implemented ERAS+ aimed at reducing PPC in 2014 and we now have results to show its sustained benefit. Methods: Using quality improvement methodology, a multidisciplinary team developed an innovative surgical pathway combined with elements of ERAS. Working with the Boston Medical Centre, we produced ICOUGH UK2 consisting of - Incentive spirometry, Coughing and deep breathing, Oral care, Understanding (patient and family education), Getting out of bed and Head-of-bed elevation. Additional innovations within the ERAS+ include: Prehabilitation/Reablement tools; SURGERY SCHOOL - patient and family preparation session and ICOUGH UK multimedia resources including ICOUGH TV. Pathway innovation occurred during sponsored listening sessions with patients and relatives. We introduced ICOUGH prescriptions attached to drug charts to promote compliance and this was audited weekly. All elective surgical patients requiring critical care admission were screened for the development of PPC on days 1, 3, 5, 7 and 15 after surgery using standard definitions3. We prospectively measured baseline PPC between April 2013–4 and repeated PPC audit following introduction of ERAS+ in September 2014 through to Jan 2016.

tive surgical patients requiring critical care admission were screened for the development of PPC on days 1, 3, 5, 7 and 15 after surgery using standard definitions3. We prospectively measured baseline PPC between April 2013–4 and repeated PPC audit following introduction of ERAS+ in September 2014 through to Jan 2016. Results: ERAS+ has been successfully developed and embedded in a large NHS university teaching hospital. More than 200 critical care nurses have been trained in ERAS+. Surgery school has been embedded, with over 350 patients attending; qualitative feedback has been overwhelmingly positive. Since September 2014, utilising ERAS+ within critical care in greater than 500 patients has resulted in a sustained reduction of 40 % in PPCs. Conclusions: ERAS+ implementation into a large teaching hospital has successfully reduced PPCs in major surgery patients. ERAS+ can be rapidly adopted by European healthcare, with a significant reduction in patient morbidity and hospital length of stay. This innovation aligns with domains 1, 3–5 of the NHS Outcome Framework4. References Moonesinghe et al. 'Survival after postoperative morbidity: a longitudinal observational cohort study' BJA 2014 113; 6: 977–84 1. ICOUGH-http://www.cmft.nhs.uk/information-for-patients-visitors-and-carers/enhanced-recovery-programme/icoughuk 2. Kroenke et al. Operative risk in patients with severe obstructive pulmonary disease. Arch Intern Med. 1992;152:967–971

Moonesinghe et al. 'Survival after postoperative morbidity: a longitudinal observational cohort study' BJA 2014 113; 6: 977–84 1. ICOUGH-http://www.cmft.nhs.uk/information-for-patients-visitors-and-carers/enhanced-recovery-programme/icoughuk 2. Kroenke et al. Operative risk in patients with severe obstructive pulmonary disease. Arch Intern Med. 1992;152:967–971 3. NHS Outcome Framework - https://www.gov.uk/government/publications/nhs-outcomes-framework-2015-to-2016Fig. 14 (abstract A433). Incidence of PPC in major elective surgery patient

1. ICOUGH-http://www.cmft.nhs.uk/information-for-patients-visitors-and-carers/enhanced-recovery-programme/icoughuk 2. Kroenke et al. Operative risk in patients with severe obstructive pulmonary disease. Arch Intern Med. 1992;152:967–971 3. NHS Outcome Framework - https://www.gov.uk/government/publications/nhs-outcomes-framework-2015-to-2016Fig. 14 (abstract A433). Incidence of PPC in major elective surgery patient A434 The use of inhaled nitric oxide (INO) in neonatal, pediatric and adult intensive care units, a franco belgian multicenter prospective survey from the positive study group C. Barbanti1, J. Amour2, P. Gaudard3, B. Rozec4, P. Mauriat5, M. M'rini6, P.L. Leger7, G. Cambonie8, J.M. Liet9, C. Girard10, S. Laroche11, P. Damas12, Z. Assaf13, G. Loron14, L. Lecourt15, P. Pouard16 1Hopital Universitaire Necker Enfants Malades, Pediatric Cardiac Intensice Care Unit, Paris, France; 2Pitie-Salpetriere Hospital and Pierre and Marie Curie University, Reanimation Chirurgicale Cardio Vasculaire et Thoracique, Paris, France; 3Arnaud de Villeneuve, Anesthésie Réanimation, Montpellier, France; 4Hopital Laennec CHU de Nantes, Anesthesie Réanimation, Nantes, France; 5Maison du Haut leveque CHU Bordeaux, Congenital Cardiac Surgery Unit, Bordeaux, France; 6Clinic Pasteur, Service Anesthesie Réanimation, Toulouse, France; 7Hopital Trousseau, Réanimation Néonatale et Polyvalente, Paris, France; 8Hopital Arnaud de Villeneuve, Réanimation Pédiatrique et Neonatale, Montpellier, France; 9CHU de Nantes, Pédiatric Intensive Care Unit, Nantes, France; 10CHU Bocage, Réanimation Cardio Vasculaire et Polyvalente, Dijon, France; 11Air Liquide Santé International, Reserach Center Paris Saclay, Paris, France; 12Hopital Start-Tilman, Services de Soins Intensifs Généraux, Liege, Belgium, 13Hopital Necker, Réanimation Néonatale, Paris, France; 14American Memorial Hospital, Réanilmation Néonatale et Polyvalente, Reims, France; 15Air Liquide Santé International, Gentilly, France; 16Hopital Necker Enfants Malades, Pediatric Cardiac Intensive Care, Paris, France Correspondence: C. Barbanti - Hopital Universitaire Necker Enfants Malades, Pediatric Cardiac Intensice Care Unit, Paris, France Introduction: Inhaled NO is a well-known selective pulmonary vasodilator in Europe since 20 years nevertheless few study really described the daily ICU practice. The objective of this study was to determine the gap between guidelines and real life.

r Enfants Malades, Pediatric Cardiac Intensice Care Unit, Paris, France Introduction: Inhaled NO is a well-known selective pulmonary vasodilator in Europe since 20 years nevertheless few study really described the daily ICU practice. The objective of this study was to determine the gap between guidelines and real life. Methods: This is a multicenter, prospective, observational study on iNO administered through an integrated delivery and monitoring device (EZ-Kinox) in 12 French and 1 Belgian centers for pulmonary arterial hypertension after cardiac surgery (PAH CS) and for persistent pulmonary hypertension of the newborn (PPHN). The following parameters were observed: dose, treatment duration, ventilation modes, monitoring procedures, weaning procedures and occurrence of a rebound effect. Concomitant pulmonary vasodilators treatments and safety data were also collected.

(PAH CS) and for persistent pulmonary hypertension of the newborn (PPHN). The following parameters were observed: dose, treatment duration, ventilation modes, monitoring procedures, weaning procedures and occurrence of a rebound effect. Concomitant pulmonary vasodilators treatments and safety data were also collected. Results: 236 patients were studied among 238 patients enrolled within one year period : 81 children and 117 adults with PAHCC and 38 neonates with PPHN. Echocardiography or pulmonary artery catheterization were performed to diagnose PAH respectively in 86 % and 31 % of the cardiac patients. In the neonatal group 97.4 % had an echocardiographic diagnosis. Inhaled NO was initiated before ICU admission in adult population (57 %) but rarely in cardiac pediatric patients (12.7 %), p < 0.01 and 38.9 % in neonates. The median initial dosage of iNO set was 20 ppm [18–20] in the cardiac pediatric group and 10 ppm [10–15] for adults and 16.7 ppm [11.2-20] for neonates with a median duration respectively 3.9 days [1.9-6.1], 3.8 days [1.8-6.8] and 3.07 days [1.04-5.74]. The NO therapy classically was delivered during controlled ventilation mode including high frequency oscillation and more and more via non invasive ventilation as high flow nasal cannula. The clinical effect of iNO was considered sufficient in 89 % of the cases and the dose was gradually decreased before definitive withdrawal in 86 % of the cases. Adverse effects (AE) occurred 75 times (80 % of the patients without AE) including rebound effect in 1.2 % of children, 3.4 % of adults and 2.6 % of neonates, methaemoglobinemia was observed on 7.9 % of neonates and 1.7 % of adults. The NO2 generated by contact between NO and O2 was above 0.5 ppm value in 17 % of the pediatric cases, 1 % of the adult cases and never observed on neonates population. All these AE related to iNO recovered without sequelae. Pulmonary vasodilators (levosimendan, sildenafil, milrinone) were associated in 95 % of the cases in pediatrics, 23 % in adults (p < 0.01) and 23.7 % in neonates. ICU stay was 8 days [6–15] for children, 10 days [6–16] for adults and 3.7 days [1–5.7] for neonates.

All these AE related to iNO recovered without sequelae. Pulmonary vasodilators (levosimendan, sildenafil, milrinone) were associated in 95 % of the cases in pediatrics, 23 % in adults (p < 0.01) and 23.7 % in neonates. ICU stay was 8 days [6–15] for children, 10 days [6–16] for adults and 3.7 days [1–5.7] for neonates. Discussion. This survey confirms the good efficacy and safety of NO therapy in the three populations. Occurrence of rebound effect is particularly rare with a large application of a gradually withdrawal of iNO and pulmonary vasodilators use. The usage of last generation of NO devices subject to prior training allows good compliance with recommendations. A435 Peri-operative use of plasmalyte in kidney transplant recipients improves biochemical profile and may result in faster renal recovery D. Randall1, A. Adwaney1, M. Blunden1, J.R. Prowle1,2, C.J. Kirwan1,2 1Barts Health NHS Trust, Renal and Transplant Medicine, London, UK; 2Barts Health NHS Trust, Adult Critical Care, London, UK Correspondence: C.J. Kirwan - Barts Health NHS Trust, Renal and Transplant Medicine, London, UK Introduction: Administration of normal saline has been associated with the development and exacerbation of acute kidney injury, [1] and this may be related to hyperchloraemia. Since recipients of kidney transplants often require large volumes of fluid replacement, we hypothesised improved post-operative biochemistry and quicker renal recovery following a switch from saline to Plasmalyte fluid therapy during the routine peri-operative period.

y, [1] and this may be related to hyperchloraemia. Since recipients of kidney transplants often require large volumes of fluid replacement, we hypothesised improved post-operative biochemistry and quicker renal recovery following a switch from saline to Plasmalyte fluid therapy during the routine peri-operative period. Methods: The peri-operative fluid management guidance was changed in our centre in March 2013. We performed a single centre retrospective observational study of consecutive renal transplant recipients before and after the policy change to assess its impact on transplant patients. We used the local renal and pathology databases, paper and computerised patient notes and clinic letters to collect data. We excluded patients with incomplete data of the peri-operative period or who had not received the standard fluid regimen of the time. Results: 47 patients received normal saline and 29 received plasmalyte. Demographics were similar between groups, except the saline group contained more patients receiving a kidney donated after cardiac death (36 % vs 17 %, p < 0.001). In both groups, fluid administration rates were matched to urine output and overall fluid balance was equal between the two groups. Electrolyte data for the two groups are shown in Fig. 15 and demonstrate significantly higher potassium and chloride and lower bicarbonate levels in the normal saline group (P < 0.0001).

. In both groups, fluid administration rates were matched to urine output and overall fluid balance was equal between the two groups. Electrolyte data for the two groups are shown in Fig. 15 and demonstrate significantly higher potassium and chloride and lower bicarbonate levels in the normal saline group (P < 0.0001). There was a higher proportion of hyperkalaemia (serum K+ >6.0 mmol/L) immediately post-op (21 % vs 0 %), over the first 24 hours (32 % vs 21 %) and over the second 24 hours post-operatively (15 % vs 7 %) in the patients who received normal saline (P < 0.001). The saline group had a longer length of stay (11 days vs 7 days, p < 0.0001), and lower eGFR at 3 months post-op (43 vs 51 ml/min; P = 0.016), although by 1-year post-transplant, mean renal function was identical at 51 ml/min in both groups (Fig. 16). Conclusions: The sole use of Plasmalyte fluid replacement in the peri-operative period is associated with improved potassium, chloride and acid base homeostasis after kidney transplantation when compared to normal saline. The use of Plasmalyte is safe (despite the concern that it contains potassium), and may be associated with improved early renal outcomes. Whilst evidence of benefit for balanced replacement fluids in the general population is conflicting, after transplantation the effects may be more pronounced as the volumes of fluid administered are greater. References [1] Yunos NM et al. JAMA. 2012 Oct 17;308(15):1566–72Fig. 15 (abstract A435). ᅟ Fig. 16 (abstract A435). ᅟ

Conclusions: The sole use of Plasmalyte fluid replacement in the peri-operative period is associated with improved potassium, chloride and acid base homeostasis after kidney transplantation when compared to normal saline. The use of Plasmalyte is safe (despite the concern that it contains potassium), and may be associated with improved early renal outcomes. Whilst evidence of benefit for balanced replacement fluids in the general population is conflicting, after transplantation the effects may be more pronounced as the volumes of fluid administered are greater. References [1] Yunos NM et al. JAMA. 2012 Oct 17;308(15):1566–72Fig. 15 (abstract A435). ᅟ Fig. 16 (abstract A435). ᅟ A436 Albumin, type of surgery, and apache II score predict return home after surgery in critically ill patients aged more than 85 years running title: predictors for home return of patients aged >85 years S.H. Kim, S. Na, J. Kim Yonsei University College of Medicine, Anesthesia and Pain Medicine, Seoul, Republic of Korea Correspondence: S.H. Kim - Yonsei University College of Medicine, Anesthesia and Pain Medicine, Seoul, Republic of Korea Introduction: Owing to an aging society, both the number of operations for patients aged >85 years and the average age of patients admitted to the intensive care unit (ICU) are rapidly increasing. However, mortality is not an appropriate outcome measurement in patients aged >85 years; a more important outcome is home return (HR), because the quality of life is valuable to these patients.

for patients aged >85 years and the average age of patients admitted to the intensive care unit (ICU) are rapidly increasing. However, mortality is not an appropriate outcome measurement in patients aged >85 years; a more important outcome is home return (HR), because the quality of life is valuable to these patients. Objectives: To identify predictors for HR of patients aged >85 years admitted to the ICU after surgery. Study design and settings: Retrospective analysis of medical records at a university hospital. Subjects: Patients aged >85 years admitted to the ICU after surgery (n = 187). Methods: Patients were divided into a HR group (patients who returned home after discharge) and non-HR group (deceased or transferred to nursing facilities). Perioperative data and outcome were assessed and compared. Multivariate logistic regression analysis was conducted to identify independent predictors. Results: The average age of patients was 88 years. HR occurred in 61 % of patients, and mortality was 9 %. The HR group showed higher preoperative albumin level than the non-HR group. More patients in the non-HR group experienced hip surgery than in the HR group (51 % vs. 12 %, P < 0.001). APACHE II score was higher (P < 0.001) in the non-HR group. In multivariate analysis, preoperative albumin, hip surgery, and APACHE II score were identified as independent predictors of HR. Conclusion: Predictors of HR of surgical critically ill elderly patients included preoperative albumin level, hip surgery, and APACHE II score on ICU admission. References

Results: The average age of patients was 88 years. HR occurred in 61 % of patients, and mortality was 9 %. The HR group showed higher preoperative albumin level than the non-HR group. More patients in the non-HR group experienced hip surgery than in the HR group (51 % vs. 12 %, P < 0.001). APACHE II score was higher (P < 0.001) in the non-HR group. In multivariate analysis, preoperative albumin, hip surgery, and APACHE II score were identified as independent predictors of HR. Conclusion: Predictors of HR of surgical critically ill elderly patients included preoperative albumin level, hip surgery, and APACHE II score on ICU admission. References 1. Hennessy D, Juzwishin K, Yergens D, Noseworthy T, Doig C. Outcomes of elderly survivors of intensive care: a review of the literature. Chest 2005; 127: 1764–74. 2. Nguyen YL, Angus DC, Boumendil A, Guidet B. The challenge of admitting the very elderly to intensive care. Ann Intensive Care 2011; 1: 29. 3. Bagshaw SM, Webb SA, Delaney A et al. Very old patients admitted to intensive care in Australia and New Zealand: a multi-centre cohort analysis. Crit Care 2009; 13: R45.

1. Hennessy D, Juzwishin K, Yergens D, Noseworthy T, Doig C. Outcomes of elderly survivors of intensive care: a review of the literature. Chest 2005; 127: 1764–74. 2. Nguyen YL, Angus DC, Boumendil A, Guidet B. The challenge of admitting the very elderly to intensive care. Ann Intensive Care 2011; 1: 29. 3. Bagshaw SM, Webb SA, Delaney A et al. Very old patients admitted to intensive care in Australia and New Zealand: a multi-centre cohort analysis. Crit Care 2009; 13: R45. A437 Postoperative serum chloride levels and acute kidney injury in the ICU after liver transplantation S.-Y. Oh1, C.W. Jung2, S.-H. Yoo2, S.-H. Min2, E.-J. Chung2, H. Lee2, N.J. Lee1, K.W. Lee1, K.-S. Suh1, H.G. Ryu2 1Seoul National University College of Medicine, Department of Surgery, Seoul, Republic of Korea; 2Seoul National University College of Medicine, Department of Anesthesiology, Seoul, Republic of Korea Correspondence: S.-Y. Oh - Seoul National University College of Medicine, Department of Surgery, Seoul, Republic of Korea Introduction: Immediate postoperative liver transplant recipients are at risk for developing acute kidney injury (AKI) with a reported incidence between 17 to 95 %. Recent findings suggest that the chloride-rich resuscitation fluids may increase the incidence of acute kidney injury in critically ill and surgical patients. Hyperchloremic fluids have been shown to induce renal vasoconstriction, attenuate renal artery flow velocity, and reduce renal cortical tissue perfusion.

o 95 %. Recent findings suggest that the chloride-rich resuscitation fluids may increase the incidence of acute kidney injury in critically ill and surgical patients. Hyperchloremic fluids have been shown to induce renal vasoconstriction, attenuate renal artery flow velocity, and reduce renal cortical tissue perfusion. Objectives: The association between postoperative serum chloride levels and the incidence of immediate postoperative acute kidney injury in the ICU in liver transplant recipients were evaluated. Methods: Adult patients (≥18 years old) who underwent liver transplantation at Seoul National University Hospital between July 2010 and December 2012. Preoperative serum chloride levels were measured within 24 hours before the operation and postoperative serum chloride levels were routinely measured every 8 hours during the initial 72 hours after ICU admission (10 measurements). The average serum chloride level was defined as the average of the 10 serum chloride levels in patients who did not develop AKI or developed AKI 72 hours after liver transplantation and the average of the preoperative serum chloride level and the serum chloride levels measured before the development of AKI in those who developed AKI within 72 hours. Patients were divided into 3 groups according to their average chloride level.; normochloremia group (96–106 mEq/L), hypochloremia group (<96 mEq/L), or hyperchloremia group (>106 mEq/L). The primary outcome was postoperative AKI according to the RIFLE criteria. Secondary outcomes included in-hospital mortality and surgical site infection.

into 3 groups according to their average chloride level.; normochloremia group (96–106 mEq/L), hypochloremia group (<96 mEq/L), or hyperchloremia group (>106 mEq/L). The primary outcome was postoperative AKI according to the RIFLE criteria. Secondary outcomes included in-hospital mortality and surgical site infection. Results: AKI developed in 46.6 % (158/339) of the patients. The overall in-hospital mortality was 4.7 %. Compared to normochloremia, AKI was more frequent in patients with hyperchloremia (adjusted OR 1.826 [95 % CI 1.133-2.944], P = 0.013) and hypochloremia (adjusted OR 8.794 [95 % CI 1.820-42.498], P = 0.007). There was no difference in preoperative chloride levels between patients who developed AKI and those who did not (104.54 ± 8.41 vs 103.18 ± 6.14 mEq/L, P = 0.089). Conclusions: Hyperchloremia and hypochloremia were both associated with an increased risk of developing AKI in the immediate postoperative period after liver transplantation in the ICU. References 1. Hoste EA, Clermont G, Kersten A, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Crit Care 2006; 10: R73 2. Barri YM, Sanchez EQ, Jennings LW, et al. Acute kidney injury following liver transplantation: definition and outcome. Liver transplantation : official publication of the American Association for the Study of Liver Diseases and the International Liver Transplantation Society 2009; 15: 475–83

1. Hoste EA, Clermont G, Kersten A, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Crit Care 2006; 10: R73 2. Barri YM, Sanchez EQ, Jennings LW, et al. Acute kidney injury following liver transplantation: definition and outcome. Liver transplantation : official publication of the American Association for the Study of Liver Diseases and the International Liver Transplantation Society 2009; 15: 475–83 A438 The association between serum sodium fluctuations and mortality in surgical patients requiring intensive care D.C. Marshall, R.J. Goodson, J.D. Salciccioli, J. Shalhoub Imperial College London, London, UK Correspondence: J.D. Salciccioli - Imperial College London, London, UK Introduction: Electrolyte abnormalities are common in critically ill populations and are associated with mortality. Previous reports have demonstrated a relationship between serum sodium fluctuations and mortality in a cohort of surgical critically ill patients (1–3). Objectives: Our primary aim was to assess the association between serum sodium fluctuations and 28-day mortality in a cohort of adult surgical patients requiring intensive care. Our secondary aim was to assess the association between dysnatraemia and 28-day mortality in surgical patients requiring intensive care.

A438 The association between serum sodium fluctuations and mortality in surgical patients requiring intensive care D.C. Marshall, R.J. Goodson, J.D. Salciccioli, J. Shalhoub Imperial College London, London, UK Correspondence: J.D. Salciccioli - Imperial College London, London, UK Introduction: Electrolyte abnormalities are common in critically ill populations and are associated with mortality. Previous reports have demonstrated a relationship between serum sodium fluctuations and mortality in a cohort of surgical critically ill patients (1–3). Objectives: Our primary aim was to assess the association between serum sodium fluctuations and 28-day mortality in a cohort of adult surgical patients requiring intensive care. Our secondary aim was to assess the association between dysnatraemia and 28-day mortality in surgical patients requiring intensive care. Methods: We performed a retrospective analysis of the Multi-Parameter Intelligent Monitoring in Intensive Care (MIMIC-II) database. Subjects were categorized by severity of dysnatraemia and sodium fluctuations during the index ICU admission. Dysnatraemia was defined as a sodium concentration upon ICU admission outside physiologic range (135–145 mmol/L). Univariate and multivariable logistic regression was used to test primary and secondary aims.

e. Subjects were categorized by severity of dysnatraemia and sodium fluctuations during the index ICU admission. Dysnatraemia was defined as a sodium concentration upon ICU admission outside physiologic range (135–145 mmol/L). Univariate and multivariable logistic regression was used to test primary and secondary aims. Results: We identified 8600 subjects, 39 % female with a median age of 66 years for analysis; 28-day mortality was 8 %. Fluctuations in serum sodium were found to be associated with 28-day mortality (adjusted odds ratio (OR) per 1 mmol/L change: 1.10 (95%CI 1.08-1.13)) in dysnatraemia. In subjects who remained normotraemic, there was an association between fluctuation in serum sodium and 28-day mortality (adjusted OR 1.13, 95%CI 1.10 - 1.16; p < 0.001). Subjects with dysnatraemia were more likely to be dead at 28-days (17 % vs 7 %; p < 0.001). Severity of dysnatraemia was associated with 28-day mortality (adjusted ORs (95 % CI): severe hyponatraemia (<130 mmol/L) 2.68 (1.72-4.17); mild hyponatraemia (130–134 mmol/L) 1.59 (1.24-2.05); mild hypernatraemia (146–150 mmol/L) 2.46 (1.76-3.43); severe hypernatraemia (>150 mmol/L) 4.73 (2.54-8.80)). Conclusions: Fluctuations of serum sodium, including in patients whose sodium remained within the normal range, were independently associated with an increase in 28-day mortality. Severity of dysnatraemia was associated with 28-day mortality. These findings are consistent with previous reports and suggest that minor fluctuations in sodium may be associated with mortality. References

Conclusions: Fluctuations of serum sodium, including in patients whose sodium remained within the normal range, were independently associated with an increase in 28-day mortality. Severity of dysnatraemia was associated with 28-day mortality. These findings are consistent with previous reports and suggest that minor fluctuations in sodium may be associated with mortality. References 1. Funk G-C, et al. Incidence and prognosis of dysnatremias present on ICU admission. Intensive Care Medicine 2009;36(2):304–11. 2. Gucyetmez B, et al. Dysnatremia on intensive care unit admission is a stronger risk factor when associated with organ dysfunction. Minerva Anestesiology 2014;80(10):1096–104 3. Sakr Y, et al. Fluctuations in Serum Sodium Level Are Associated With an Increased Risk of Death in Surgical ICU Patients: Critical Care Medicine 2013;41(1):133–42.

2. Gucyetmez B, et al. Dysnatremia on intensive care unit admission is a stronger risk factor when associated with organ dysfunction. Minerva Anestesiology 2014;80(10):1096–104 3. Sakr Y, et al. Fluctuations in Serum Sodium Level Are Associated With an Increased Risk of Death in Surgical ICU Patients: Critical Care Medicine 2013;41(1):133–42. A439 Venous-arterial CO2 to arterial-venous o2 difference ratio (CV-ACO2/DA-VO2) - the intra-operative anaerobic threshold? An assessment during liver resection surgery E.K. Potter1,2, J. Kirk-Bayley1,3, N.D. Karanjia1,3, L.G. Forni1,3, B.C. Creagh-Brown1,3 1Royal Surrey County Hospital, ICU and SPACeR Research Group, Guildford, UK; 2University of Surey, Guildford, UK; 3University of Surrey, Guildford, UK Correspondence: E.K. Potter - Royal Surrey County Hospital, ICU and SPACeR Research Group, Guildford, UK Introduction: Cv-aCO2/Da-vO2 has been suggested as a marker of adequacy of resuscitation in shock states. A ratio >1 reflects the metabolic tipping point at which CO2 production exceeds oxygen consumption, akin to the anaerobic threshold in cardiopulmonary exercise testing, but with the advantage of measurement from paired arterial and central venous blood gas analysis. It is of prognostic value in critical care, and in sepsis when combined with lactate[1, 2]. We investigate the role of Cv-aCO2/Da-vO2 in liver resection surgery. Our anaesthetic technique uses fluid restriction, diuresis and venodilation to decrease CVP, which improves outcome[3]. On completion of resection, goal-directed fluid therapy (GDFT) is initiated. This technique provides an ideal model of rapid haemodynamic changes in which to investigate Cv-aCO2/Da-vO2.

ction surgery. Our anaesthetic technique uses fluid restriction, diuresis and venodilation to decrease CVP, which improves outcome[3]. On completion of resection, goal-directed fluid therapy (GDFT) is initiated. This technique provides an ideal model of rapid haemodynamic changes in which to investigate Cv-aCO2/Da-vO2. Objectives: To establish: 1) Is there a relationship between peri-operative Cv-aCO2/Da-vO2 and outcome? 2) If Cv-aCO2/Da-vO2, inverse CO2 gap or serum lactate correlate with cardiac output (CO) Methods: This is a subset of data from the prospective observational study MicroHepFlow, investigating microcirculation in liver resection surgery. All patients received general and thoracic epidural anaesthesia, fluid restriction, GTN, remifentanil and furosemide. Oesophageal Doppler (OD) was used at the anaesthetists' discretion. Prospective data was collected at pre-defined time points. The Clavien-Dindo (CD) Classification was used for post-operative complications. Statistical analysis was with linear regression and Fischers exact test. Results: In 31 patients, 17 with CO data, a Cv-aCO2/Da-vO2 > 1 and lactate ≥2 was common (55 % at maximal desiccation and 52 % following GDFT). There was no consistent relationship between this and the incidence of complications. C-D complications ≥ Grade III were rare (6 %). Neither Cv-aCO2/Da-vO2, inverse CO2 gap nor serum lactate concentration significantly correlated with CO.

and lactate ≥2 was common (55 % at maximal desiccation and 52 % following GDFT). There was no consistent relationship between this and the incidence of complications. C-D complications ≥ Grade III were rare (6 %). Neither Cv-aCO2/Da-vO2, inverse CO2 gap nor serum lactate concentration significantly correlated with CO. Conclusions: The expected positive association between Cv-aCO2/Da-vO2 > 1 and lactate ≥2 with post-operative complications was not observed, potentially due to a low incidence of serious complications. The lack of significant correlation between Cv-aCO2 / Da-vO2 and inverse CO2 gap with cardiac output is informative and suggests that these variables are poor surrogates of cardiac output in this setting. References 1. Mekontso-Dessap A, Castelain V, Anguel N, et al., (2002) Combination of venoarterial PCO2 difference with arteriovenous O2 content difference to detect anaerobic metabolism in patients. Intensive Care Med. 28: 272–277 2. Ospina-Tascon GA, Umana M, Bermudez W, et al., (2015) Combination of arterial lactate levels and venous-arterial CO2 to arterial-venous O 2 content difference ratio as markers of resuscitation in patients with septic shock. Intensive Care Med 41: 796–805 3. Jones RM, Moulton CE, Hardy KJ, (1998) Central venous pressure and its effect on blood loss during liver resection. Br J Surg. 85: 1058–1060

2. Ospina-Tascon GA, Umana M, Bermudez W, et al., (2015) Combination of arterial lactate levels and venous-arterial CO2 to arterial-venous O 2 content difference ratio as markers of resuscitation in patients with septic shock. Intensive Care Med 41: 796–805 3. Jones RM, Moulton CE, Hardy KJ, (1998) Central venous pressure and its effect on blood loss during liver resection. Br J Surg. 85: 1058–1060 A440 Predicting persistent requirement for moderate vasopressor support (vasoplegia) following major gynae-oncology debulking surgery (GODS) - vasogods, a single-centre retrospective ICU study M. Bossy1, M. Nyman2, A. Tailor3, B. Creagh-Brown4, SPACeR group (Surrey Peri-operative, Anaesthesia and Critical Care Collaborative Research Group) 1Royal Surrey County Hospital, ICU and SPACeR Research Group, Intensive Care, Guildford, UK; 2University of Southampton, Medical School, Southampton, UK; 3Royal Surrey County Hospital, Department of Gynaecological Oncology, Guildford, UK; 4Royal Surrey County Hospital, Intensive Care, SPACeR Research Group, Guildford, UK Correspondence: M. Bossy - Royal Surrey County Hospital, ICU and SPACeR Research Group, Intensive Care, Guildford, UK Introduction: Despite goal-directed fluid therapy, vasopressor support may be necessary in the early post-operative period to maintain adequate systemic blood pressure. Persistent requirement for moderate doses of vasopressor support (vasoplegia) has not been reported in patients recovering from following major gynae-oncology debulking surgery (GODS). Risk factors for, and complications of, vasoplegia have not been described.

rative period to maintain adequate systemic blood pressure. Persistent requirement for moderate doses of vasopressor support (vasoplegia) has not been reported in patients recovering from following major gynae-oncology debulking surgery (GODS). Risk factors for, and complications of, vasoplegia have not been described. Objectives: Describe the incidence of vasoplegia following GODS, and the risk factors for, and complications of, vasoplegia. Methods: Patients were identified from our ICU patient database (WardWatcher software) and additional clinical details acquired from electronic databases including the IntelliSpace Critical Care and anesthesia (Philips) and a local bespoke system (The Guildford Gynae-oncology database). Statistical analysis was performed using binary logistic regression using log transformed data for all non-parametric variables (SPSS).

linical details acquired from electronic databases including the IntelliSpace Critical Care and anesthesia (Philips) and a local bespoke system (The Guildford Gynae-oncology database). Statistical analysis was performed using binary logistic regression using log transformed data for all non-parametric variables (SPSS). Results: Between February 2014 and September 2015 there were 176 elective admissions to our ICU for care of patients following GODS. The median age was 67 years (IQR range 56–76) and median weight was 70 kg. 75 % had ovarian or endometrial cancer. The median estimated blood loss (EBL) was 900mls. 67.1 % of this group had an epidural inserted for post-operative analgesia and the mean rate of a standard infusion was 10mls/h. 52.3 % of patients received a vasopressor for at least one hour during their post-operative stay on ICU (either phenylephrine or noradrenaline). 16.5 % of the group remained on a dose of noradrenaline ≥0.1mcg/kg/min at 08:00 on the first post operative day - considered to reflect vasoplegia. Median ICU length of stay (LOS) was 2.0 days (IQ range 1.1-3.0), and median hospital LOS 5.7 days (IQR 4.3-7.1). ICU mortality was < 1 % (1/176) and 6 month mortality 4 % (7/176). The only pre-operative variable associated with requiring vasopressors was haemoglobin; median was 119 g/L in those who required vs. 128 g/L in those who did not (p = 0.02). Epidural use was associated with requiring vasopressors (OR 3.74,CI 1.91-7.30, p < 0.005). Cumulative fluid balance was associated with vasoplegia (OR of log 26.22, CI 4.75-144.59, p < 0.005) and a peak lactate greater than median, 3.1 mmol/L (OR of log 4.99, CI 1.63-15.33, p = 0.005). The presence of vasoplegia was association with a prolonged hospital stay (odds ratio of 6.40, CI 2.31-17.70, p < 0.0005).

. Cumulative fluid balance was associated with vasoplegia (OR of log 26.22, CI 4.75-144.59, p < 0.005) and a peak lactate greater than median, 3.1 mmol/L (OR of log 4.99, CI 1.63-15.33, p = 0.005). The presence of vasoplegia was association with a prolonged hospital stay (odds ratio of 6.40, CI 2.31-17.70, p < 0.0005). Hospital LOS in those who didn't require any vasopressors was 4.9d (3.9-6.1), in those who required some vasopressors was 5.8d (5.0-7.9), and in those with had vasoplegia was 7.0d (6.0-9.8), p < 0.0001 KWT. Conclusions: Lower pre-operative haemoglobin concentration was associated with requirement for vasopressors, as was epidural use and cumulative fluid balance. Vasoplegia was uncommon, associated with cumulative fluid balance and elevated peak lactate, and a prolonged duration of hospital stay. Mortality was low. A441 Withdrawn A442 Noninvasive cardiac output and pressure measurements with finger cuff during lateral decubitus and one lung ventilation in thoracic surgery: preliminary results D. D'Antini1, S. Spadaro2, F. Valentino1, F. Sollitto3, G. Cinnella1, L. Mirabella1 1University of Foggia, Anesthesiology and Intensive Care, Foggia, Italy; 2University of Ferrara, Anesthesiology and Intensive Care, Ferrara, Italy; 3University of Foggia, Thoracic Surgery, Foggia, Italy Correspondence: D. D'Antini - University of Foggia, Anesthesiology and Intensive Care, Foggia, Italy Introduction: Patients undergoing thoracic surgery commonly undergo a radial catheter placement to fulfill the need for continuous blood pressure monitoring, other than blood gas analyses 1.

A441 Withdrawn A442 Noninvasive cardiac output and pressure measurements with finger cuff during lateral decubitus and one lung ventilation in thoracic surgery: preliminary results D. D'Antini1, S. Spadaro2, F. Valentino1, F. Sollitto3, G. Cinnella1, L. Mirabella1 1University of Foggia, Anesthesiology and Intensive Care, Foggia, Italy; 2University of Ferrara, Anesthesiology and Intensive Care, Ferrara, Italy; 3University of Foggia, Thoracic Surgery, Foggia, Italy Correspondence: D. D'Antini - University of Foggia, Anesthesiology and Intensive Care, Foggia, Italy Introduction: Patients undergoing thoracic surgery commonly undergo a radial catheter placement to fulfill the need for continuous blood pressure monitoring, other than blood gas analyses 1. Objectives: In this prospective observational study, we recorded cardiac output and pressure measurements obtained by an uncalibrated pulse contour method in a totally non invasive manner (ClearSight system, Edwards Lifesciences) to assess hemodynamic variations during normal and one lung ventilation (OLV) in supine and lateral decubitus in thoracic surgery. Methods: After Ethical Committee approval was obtained, we enrolled patients with age > 18 years, scheduled for thoracic surgery and OLV with double-lumen tube in lateral decubitus. Exclusion criteria: ASA > III, severe baseline hemodynamic impairment which required invasive monitoring and amine infusion.

Objectives: In this prospective observational study, we recorded cardiac output and pressure measurements obtained by an uncalibrated pulse contour method in a totally non invasive manner (ClearSight system, Edwards Lifesciences) to assess hemodynamic variations during normal and one lung ventilation (OLV) in supine and lateral decubitus in thoracic surgery. Methods: After Ethical Committee approval was obtained, we enrolled patients with age > 18 years, scheduled for thoracic surgery and OLV with double-lumen tube in lateral decubitus. Exclusion criteria: ASA > III, severe baseline hemodynamic impairment which required invasive monitoring and amine infusion. Key hemodynamic parameters, including Stroke Volume Index (SVI), Stroke Volume Variation (SVV), Cardiac Index (CI) and Continuous Blood Pressure (cBP) were recorded at three time points: two lung ventilation in supine position (T1), two lung ventilation in lateral decubitus (T2) and one lung ventilation in lateral decubitus (T3). Results: 11 consecutive patients scheduled for thoracic surgery were included. Overall, our patients showed no hemodynamic impairment throughout the passage to lateral decubitus and OLV (CI 3,1 ± 0,9 l/min/m2 at T1, 2.5 ± 0.6 l/min/m2 at T2, 2.5 ± 0.7 l/min/m2 at T3; p = 0.09 T2 and T3 vs T1). Only in 3 patients a decrease in CI was detected after the position change (−37 %, −42 % and −43 %, respectively, T2 vs T1), which required prompt intervention.

hroughout the passage to lateral decubitus and OLV (CI 3,1 ± 0,9 l/min/m2 at T1, 2.5 ± 0.6 l/min/m2 at T2, 2.5 ± 0.7 l/min/m2 at T3; p = 0.09 T2 and T3 vs T1). Only in 3 patients a decrease in CI was detected after the position change (−37 %, −42 % and −43 %, respectively, T2 vs T1), which required prompt intervention. Conclusions: The lack of invasiveness and the simplicity of the system allowed satisfactory monitoring and, subsequently, appropriate therapy, without the need of arterial catheter placement for invasive monitoring, while an automatic, intermittent, non-invasive blood pressure measurement would have not readily detected the observed hemodynamic variations. Even if most studies underline the not-interchangeability of the finger cuff method with the currently used invasive devices2, in our low to moderate risk thoracic surgery patients not receiving an arterial line, ClearSight was able to track hemodynamic changes during the different phases of the surgical procedure, in particular those related to position changes and to the OLV. References 1. Haas S, Kiefmann R, Eichhorn V, et al. Anaesthesia. 2009;58(11):1085–96. 2. Ameloot K, Palmers PJ, Malbrain ML. Curr Opin Crit Care. 2015;21(3):232–9.

Even if most studies underline the not-interchangeability of the finger cuff method with the currently used invasive devices2, in our low to moderate risk thoracic surgery patients not receiving an arterial line, ClearSight was able to track hemodynamic changes during the different phases of the surgical procedure, in particular those related to position changes and to the OLV. References 1. Haas S, Kiefmann R, Eichhorn V, et al. Anaesthesia. 2009;58(11):1085–96. 2. Ameloot K, Palmers PJ, Malbrain ML. Curr Opin Crit Care. 2015;21(3):232–9. A443 The global end-diastolic volume (GEDI) could be more appropiate to reduce the intraoperative bleeding than central venous pressure (PVC) during mayor liver resections (metastasized colon carcinoma) F.J. Redondo Calvo1, N. Bejarano2, D. Padilla3, V. Baladron4, P. Villajero3, R. Villazala4, J. Redondo4, A.S. Yuste4 1Facultad Medicina Ciudad Real, Hospital General Universitario de Ciudad Real. Anestesiologia y Reanimacion, Ciudad Real, Spain; 2Facultad Medicina Ciudad Real, Hospital General Universitario de Ciudad Real, Cuidados Criticos Pediatricos, Ciudad Real, Spain; 3Facultad Medicina Ciudad Real, Hospital General Universitario de Ciudad Real, Cirugía Hepatobiliar, Ciudad Real, Spain; 4Hospital General Universitario de Ciudad Real, Anestesiologia y Reanimación, Ciudad Real, Spain Correspondence: F.J. Redondo Calvo - Facultad Medicina Ciudad Real, Hospital General Universitario de Ciudad Real. Anestesiologia y Reanimacion, Ciudad Real, Spain Goal of study: The aim of our study was to evaluate the predictive value of CVP with regard GEDI, and correlate these parameters to cardiac Index (CI) and extravascular lung water index (EVLWI).

acultad Medicina Ciudad Real, Hospital General Universitario de Ciudad Real. Anestesiologia y Reanimacion, Ciudad Real, Spain Goal of study: The aim of our study was to evaluate the predictive value of CVP with regard GEDI, and correlate these parameters to cardiac Index (CI) and extravascular lung water index (EVLWI). Methods: Prospective study. Surgical intensive care unit, university hospital. Patients and interventions: 129 hemodynamic measurements using the PiCCO (Pulsion Medical System, Germany) were performed in 28 patiens during major liver resection. Results: Mean CVP (8,44 +/− 4,22 mmHg) was normal, whereas mean GEDI (625,3 +/− 111,34 mL/m2) was decreased. Forty-two CVP measurements were elevated despite simultaneous GEDI levels indicating a normal or decreased preload. Sensitivity, specificity, positive predictive value, and negative predictive value of CVP with regard to volume depletion (GEDI < 650) were 8,25 (0–14,66. CI 95 %), 100 (99,89-100, CI 95 %), 100 (50–100, CI 95 %), 29, 4 (21,99-26,96, CI 95 %) respectively. CVP did not correlate to GEDI (r = −0,045, p = 0,42), CI (r = 0,13, p = 0,216) and EVLWI (extravascular lung water index) (r = −0,04, p = 0,38). GEDI significantly correlated to CI (r = −0,23, p < 0,01) and VVS (r = −0,32, p < 0,01).

99,89-100, CI 95 %), 100 (50–100, CI 95 %), 29, 4 (21,99-26,96, CI 95 %) respectively. CVP did not correlate to GEDI (r = −0,045, p = 0,42), CI (r = 0,13, p = 0,216) and EVLWI (extravascular lung water index) (r = −0,04, p = 0,38). GEDI significantly correlated to CI (r = −0,23, p < 0,01) and VVS (r = −0,32, p < 0,01). Conclusions: Volume depletion according to GEDI was found in more than half the patiens. The predictive values of CVP with regard to volume depletion were low GEDI and its changes significantly correlated to CI and its changes, which was not observed for CVP. Therefore, GEDI appears to be more appropiate for volume management during mayor liver resections with the aim to avoid intraoperative bleeding and transfusión. Note: This abstract has been previously published and is available at [4]. It is included here as a complete record of the abstracts from the conference. References 1. Smyrniotis V, Kostopanagiotou G, Theodoraki K, et al. The role of central venous pressure and type of vascular control in blood loss during mayor liver resections. Am J Surg. 2004; 187:398–402 2. De la Rocca G, Costa MG, Coccia C, et al. Preload and hemodynamic assesment during liver transplantation. A comparison between pulmonary artery catheter and transpulmonar indicator dilution technique. Eur J Anesth 2002; 19:868–875

1. Smyrniotis V, Kostopanagiotou G, Theodoraki K, et al. The role of central venous pressure and type of vascular control in blood loss during mayor liver resections. Am J Surg. 2004; 187:398–402 2. De la Rocca G, Costa MG, Coccia C, et al. Preload and hemodynamic assesment during liver transplantation. A comparison between pulmonary artery catheter and transpulmonar indicator dilution technique. Eur J Anesth 2002; 19:868–875 3. Huber W, Umgelter A, Reindl W, et al. Volume assessment in patients with necrotizing pancreatitis: a comparison of intrathoracic blood volumen index, central venous pressure, and hematocrit, and their correlation to cardiac index and extravascular lung wáter. Crit Care Med 2008; 36:2348–5. 4. Redondo Calvo FJ, Bejarano N, Padilla D, Villarejo P, Baladron V, Villazalo R, Yuste AS, Arenas P (2015) A comparison of global end-diastolic volume (GEDI) and central venous pressure (CVP) as parameter for volume assessment I patients during major liver resections. Intensive Care Medicine Experimental 3(Suppl 1): A216 Grant acknowledgement We express our gratitude to Mutua Madrileña Fundation (Madrid, Spain) for its grant collaboration by without which this work could not have been completed.

4. Redondo Calvo FJ, Bejarano N, Padilla D, Villarejo P, Baladron V, Villazalo R, Yuste AS, Arenas P (2015) A comparison of global end-diastolic volume (GEDI) and central venous pressure (CVP) as parameter for volume assessment I patients during major liver resections. Intensive Care Medicine Experimental 3(Suppl 1): A216 Grant acknowledgement We express our gratitude to Mutua Madrileña Fundation (Madrid, Spain) for its grant collaboration by without which this work could not have been completed. A444 Incidence, risk factors of weaning-induced pulmonary oedema and effects of fluid removal J. Liu1,2, F. Shen1,3, J.-L. Teboul1, N. Anguel1, A. Beurton1, N. Bezaz1, C. Richard1, X. Monnet1 1Hôpital de Bicêtre, Hôpitaux universitaires Paris-Sud, Université Paris-Sud, Service de réanimation médicale, Inserm UMR_S999, Le Kremlin-Bicêtre, France; 2The First Affiliated Hospital of Chongqing Medical University, Department of Emergency Medicine and Critical Care Medicine, Chongqing, China; 3Affiliated Hospital of Guizhou Medical University, Department of Critical Care Medicine, Guiyang, China Correspondence: X. Monnet - Hôpital de Bicêtre, Hôpitaux universitaires Paris-Sud, Université Paris-Sud, Service de réanimation médicale, Inserm UMR_S999, Le Kremlin-Bicêtre, France Introduction: Weaning-induced pulmonary oedema (WiPO) is a well-recognised cause of weaning failure, but its incidence and risk factors have not been established and the effects of fluid removal have not been systematically described. Objectives: To describe the epidemiology of WiPO and the effects of diuretics.

A444 Incidence, risk factors of weaning-induced pulmonary oedema and effects of fluid removal J. Liu1,2, F. Shen1,3, J.-L. Teboul1, N. Anguel1, A. Beurton1, N. Bezaz1, C. Richard1, X. Monnet1 1Hôpital de Bicêtre, Hôpitaux universitaires Paris-Sud, Université Paris-Sud, Service de réanimation médicale, Inserm UMR_S999, Le Kremlin-Bicêtre, France; 2The First Affiliated Hospital of Chongqing Medical University, Department of Emergency Medicine and Critical Care Medicine, Chongqing, China; 3Affiliated Hospital of Guizhou Medical University, Department of Critical Care Medicine, Guiyang, China Correspondence: X. Monnet - Hôpital de Bicêtre, Hôpitaux universitaires Paris-Sud, Université Paris-Sud, Service de réanimation médicale, Inserm UMR_S999, Le Kremlin-Bicêtre, France Introduction: Weaning-induced pulmonary oedema (WiPO) is a well-recognised cause of weaning failure, but its incidence and risk factors have not been established and the effects of fluid removal have not been systematically described. Objectives: To describe the epidemiology of WiPO and the effects of diuretics. Methods: We included 81 patients who performed 283 spontaneous breathing trials (SBT).Three experts established the diagnosis of WiPO based on various patient characteristics. Demographic data were collected. A passive leg raising (PLR) test was performed before SBT in 81 patients. We also observed the effects of diuretics when this treatment was decided.

performed 283 spontaneous breathing trials (SBT).Three experts established the diagnosis of WiPO based on various patient characteristics. Demographic data were collected. A passive leg raising (PLR) test was performed before SBT in 81 patients. We also observed the effects of diuretics when this treatment was decided. Results: SBT failed in 129 cases (46 % of all SBT). WiPO occurred in 59 % of these failing cases. Compared to patients without WiPO (n = 52), patients with at least one WiPO (n = 29) had a significantly higher prevalence of obesity (45 % vs. 17 %, resp), COPD (38 % vs. 12 %, resp), dilated cardiopathy (48 % vs. 17 %, resp) and low left ventricular ejection fraction (55 % vs. 21 %, resp). In 16 cases with WiPO and a negative PLR at baseline, diuretics were administered. In 9 of these cases, the PLR remained negative before the following SBT. A new episode of WiPO occurred in 7 of these instances while the two other ones were extubated. In 7 other cases, the PLR became positive before the following SBT. WiPO did not occur anymore in 6 of these 7 patients who were extubated, while the remaining one was not. Conclusions: WiPO was responsible for 59 % of weaning failures. Obesity, previous cardiopathy and COPD were its main risk factors. When fluid removal had changed preload-independence to preload-dependence, the following SBT was very likely to succeed.

Results: SBT failed in 129 cases (46 % of all SBT). WiPO occurred in 59 % of these failing cases. Compared to patients without WiPO (n = 52), patients with at least one WiPO (n = 29) had a significantly higher prevalence of obesity (45 % vs. 17 %, resp), COPD (38 % vs. 12 %, resp), dilated cardiopathy (48 % vs. 17 %, resp) and low left ventricular ejection fraction (55 % vs. 21 %, resp). In 16 cases with WiPO and a negative PLR at baseline, diuretics were administered. In 9 of these cases, the PLR remained negative before the following SBT. A new episode of WiPO occurred in 7 of these instances while the two other ones were extubated. In 7 other cases, the PLR became positive before the following SBT. WiPO did not occur anymore in 6 of these 7 patients who were extubated, while the remaining one was not. Conclusions: WiPO was responsible for 59 % of weaning failures. Obesity, previous cardiopathy and COPD were its main risk factors. When fluid removal had changed preload-independence to preload-dependence, the following SBT was very likely to succeed. A445 Heart lung interactions measured by means of phase synchronization analysis under three ventilatory modes: pressure control, pressure support and NAVA T. Fossali1, R. Colombo1, D. Ottolina1, M. Rossetti1, C. Mazzucco2, A. Marchi2, A. Porta3, E. Catena1 1Ospedale Luigi Sacco - Milano, Department of Anaesthesia and Intensive Care, Milano, Italy; 2Politecnico di Milano, Department of Electronics information and Bioengineering, Milano, Italy; 3Università degli studi di Milano, Department of Biomedical Science of Health, Milano, Italy Correspondence: T. Fossali - Ospedale Luigi Sacco - Milano, Department of Anaesthesia and Intensive Care, Milano, Italy Introduction: Pulmonary and cardiac systems are deeply connected in physiologic conditions by autonomic nervous system. In particular Respiratory Sinus Arrhythmia (RSA) modifies heart rate in correspondence of inspiration to permit a better ventilation-perfusion matching. Cardio-respiratory phase synchronization analysis (CRPS) defined as a repetitive occurrence of m heartbeats at the same relative phases within n consecutive breathing cycles, also provides indexes relevant to nonlinear interactions between cardiac and respiratory systems complementary to RSA (1). These mechanisms are poorly studied during mechanical ventilation.

lysis (CRPS) defined as a repetitive occurrence of m heartbeats at the same relative phases within n consecutive breathing cycles, also provides indexes relevant to nonlinear interactions between cardiac and respiratory systems complementary to RSA (1). These mechanisms are poorly studied during mechanical ventilation. Objectives: Aim of the study was to compare cardio-ventilatory interactions in different mechanical ventilation modes by means of CRPS and RSA analysis. Methods: The study was conducted in the Intensive Care Unit of Luigi Sacco Hospital, Milan, Italy. The experimental protocol consisted of three randomized sessions of 30 minutes corresponding to different ventilation modes: Pressure Control Ventilation (PCV), Pressure Support Ventilation (PSV) and Neurally Adjusted Ventilatory Assist (NAVA). Electrocardiographic (ECG) and airway volume (AWV) signals were acquired at 250 Hz sampling rate and continuously registered (Philips IntelliVue MX800, iXtrend software). Signals were analyzed offline: the Hilbert transform was applied to AWV signal to estimate the respiration phase. In the synchrogram the phase values of the AWV observed in correspondence of the R-wave peak on the ECG was drawn as function of the cardiac beat number. The quantification of cardio-respiratory coupling was assessed by the synchronization index (SYNC), i.e. the percentage of beats in which it could be observed the most frequent PLR. RSA amplitude, the difference between the longest and the shortest R-R interval on the ECG, was estimate for each respiratory cycle.

mber. The quantification of cardio-respiratory coupling was assessed by the synchronization index (SYNC), i.e. the percentage of beats in which it could be observed the most frequent PLR. RSA amplitude, the difference between the longest and the shortest R-R interval on the ECG, was estimate for each respiratory cycle. Results: 20 mechanically ventilated patients were included. Mechanical ventilation was set to guarantee constant minute ventilation (MV 10 ± 4 vs 10 ± 3 vs 10 ± 2 liters/minute for PCV, PSV and NAVA respectively). CRPS analysis showed a significant variation of SYNC index between the different modes (Fig. 3), while RSA was not able to detect meaningful differences (23 ± 26 ms, 24 ± 24 ms and 22 ± 22 ms, p = 0.6, for PCV, PSV and NAVA respectively). Conclusions: Analysis of CRPS is possible in ventilated patients. Our results endorse the hypothesis that under ventilation with a fixed frequency, the interactions between cardiac and respiratory fluctuations were stronger than in the case of patient driven modes of mechanical ventilation, as PSV and NAVA (2). The lack of variation of RSA may be due to sedation and mechanical ventilation. References 1. Schäfer C et al., Heartbeat synchronized with ventilation. Nature. 1998 Mar 19;392(6673):239–40. 2. Larsen PD et al., Cardioventilatory coupling: effects of IPPV. British Journal of Anaesthesia. 1999 82; 546–550.

Conclusions: Analysis of CRPS is possible in ventilated patients. Our results endorse the hypothesis that under ventilation with a fixed frequency, the interactions between cardiac and respiratory fluctuations were stronger than in the case of patient driven modes of mechanical ventilation, as PSV and NAVA (2). The lack of variation of RSA may be due to sedation and mechanical ventilation. References 1. Schäfer C et al., Heartbeat synchronized with ventilation. Nature. 1998 Mar 19;392(6673):239–40. 2. Larsen PD et al., Cardioventilatory coupling: effects of IPPV. British Journal of Anaesthesia. 1999 82; 546–550. Drugs and poisons in the ICU A446 Substance abuse as predictor of in-hospital mortality in patients with cardiac arrest K.H. Tollisen1, G.Ø. Andersen2, F. Heyerdahl3, D. Jacobsen4 1Oslo University Hospital, Acute Medicine, Oslo, Norway; 2Oslo University Hospital, Departement of Cardiology, Oslo, Norway; 3Oslo University Hospital, Departement of Anesthesiology, Oslo, Norway; 4Oslo University Hospital, Departement of Acute Medicin, Oslo, Norway Correspondence: K.H. Tollisen - Oslo University Hospital, Acute Medicine, Oslo, Norway Introduction: Initial cardiac rhythm (ICR) is associated with mortality in cardiac arrest (CA) patients with return of spontaneous circulation (ROSC) [1]. However, knowledge regarding a possible association between substance abuse and mortality is limited.

llisen - Oslo University Hospital, Acute Medicine, Oslo, Norway Introduction: Initial cardiac rhythm (ICR) is associated with mortality in cardiac arrest (CA) patients with return of spontaneous circulation (ROSC) [1]. However, knowledge regarding a possible association between substance abuse and mortality is limited. Objectives: (1)To determine whether substance abuse is associated with in-hospital mortality in patients with CA admitted to ICU`s (intensive care units), and (2) to study the association between substance abuse and ICR. Methods: CA patients with ROSC-- admitted to the ICU`s at Oslo University Hospital Ullevål from Feb.3, 2014 to Feb.2,2015 -- were included as part of a multicenter prospective observational study of ICU patients in Oslo (n = 900). Informed consent was given by the next of kin, and if possible, later by the patient. Illicit drug/alcohol abuse (acute or chronic) was assessed via standard questionnaires, biochemical analysis (drug screen) and medical records. The patients were divided into groups based on ICR: 1) non-shockable rhythm (asystole/pulseless electric activity) and 2) shockable rhythm (VF/VT). Chi-square test was used for statistical analysis, p value < 0.05 was considered statistically significant.

nnaires, biochemical analysis (drug screen) and medical records. The patients were divided into groups based on ICR: 1) non-shockable rhythm (asystole/pulseless electric activity) and 2) shockable rhythm (VF/VT). Chi-square test was used for statistical analysis, p value < 0.05 was considered statistically significant. Results: 145 patients (116 (80 %) males)) were included, reflecting a participation rate of 95 %. 72(50 %) had non-shockable rhythm, and 73(50 %) had shockable rhythm (mean age 62 vs 63). There were significantly more females in the non-shockable group (21vs.8, p < 0.01), corresponding with earlier studies. Acute drug intoxication was considered triggering cause in 7 patients (5 %), all males (mean age 43vs 64). Main agents were: opioids (4/7), alcohol (2/7,both with blood levels > 4.2 g/L (92 mmol/L)) and benzodiazepines (1/7). All these patients were in the non-shockable group. However, the sample size was too small for further analysis to be pursued. Among the 31 (21 %) classified with chronic substance abuse (27(87 %)males, mean age 59 vs. 64) alcohol was the predominant substance (75 %), followed by illicit substances (21 %) and prescription drugs (5 %). The proportion of patients with chronic abuse was significantly higher among patients with non-shockable rhythm (32 % vs 13 % p = 0.005). In-hospital mortality was significantly higher among patients with non-shockable rhythm (83 % vs 42 %, p < 0.001). Chronic abuse, however, was not significantly associated with in-hospital mortality (p = 0.43).

ith chronic abuse was significantly higher among patients with non-shockable rhythm (32 % vs 13 % p = 0.005). In-hospital mortality was significantly higher among patients with non-shockable rhythm (83 % vs 42 %, p < 0.001). Chronic abuse, however, was not significantly associated with in-hospital mortality (p = 0.43). Conclusions: Chronic substance abuse is common among patients with CA and is significantly associated with non-shockable ICR. As expected, in-hospitality mortality was significantly higher in the non-shockable group. Although chronic abuse was associated with non-shockable initial rhythm, it was not associated with in-hospital mortality. References 1. Sasson, C., et al., Predictors of survival from out-of-hospital cardiac arrest: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes, 2010. 3(1): p. 63–81.

Conclusions: Chronic substance abuse is common among patients with CA and is significantly associated with non-shockable ICR. As expected, in-hospitality mortality was significantly higher in the non-shockable group. Although chronic abuse was associated with non-shockable initial rhythm, it was not associated with in-hospital mortality. References 1. Sasson, C., et al., Predictors of survival from out-of-hospital cardiac arrest: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes, 2010. 3(1): p. 63–81. A447 Volatile or intravenous anesthetics in a mouse model of cardiac arrest? M.C. de Waard, A.R.J. Girbes VU University Medical Center Amsterdam, Amsterdam, Netherlands Correspondence: M.C. de Waard - VU University Medical Center Amsterdam, Amsterdam, Netherlands Introduction: Cardiac arrest (CA) is a major health problem in western countries. Despite improvements in quality of treatment, it still has a high rate of morbidity and mortality in patients. CA results in a harmful inflammatory response and overwhelming amounts of reactive oxygen species are produced, spreading out across the body, causing leakage of vascular walls and damage to organs. The mouse is an increasingly used preferred model in preclinical CA studies due to the knowledge concerning the mouse genome and the general availability of genetically modified mice. Volatile anesthetics are often used in these models as they are easy to apply and short-acting. However, volatile anesthetics are known to protect against I/R injury, reduce myocardial infarct size by preconditioning and may protect vital organs by postconditioning.

he general availability of genetically modified mice. Volatile anesthetics are often used in these models as they are easy to apply and short-acting. However, volatile anesthetics are known to protect against I/R injury, reduce myocardial infarct size by preconditioning and may protect vital organs by postconditioning. Objectives: This pilot study investigates the use of i.v. or volatile anesthetics in a mouse model for CA.

he general availability of genetically modified mice. Volatile anesthetics are often used in these models as they are easy to apply and short-acting. However, volatile anesthetics are known to protect against I/R injury, reduce myocardial infarct size by preconditioning and may protect vital organs by postconditioning. Objectives: This pilot study investigates the use of i.v. or volatile anesthetics in a mouse model for CA. Methods: Experiments were approved by the ethical committee of VU Medical Centre Amsterdam. Mice randomly assigned to one of the two groups. Thereafter they were weighed, sedated with 4 % sevoflurane and buprenorphine (0.05 mg/kg s.c.) or a mixture of fentanyl/midazolam/acepromazine (FMA: 0.47 mg/kg, 9.38 mg/kg, 9.38 mg/kg; respectively i.p.), intubated and placed on a heating mat. The mice were pressure-controlled ventilated with 21 % O2 (90–100 breaths/min, PEEP 4cmH2O, peak 16cmH2O). Anaesthesia was maintained with 2.5-3 % sevoflurane or with FMA (0.08 mg/kg, 1.56 mg/kg, 1.56 mg/kg i.v.). 21 mice were sedated and anesthetized with sevoflurane and 6 with FMA. Electrocardiogram (ECG) was recorded continuously. For measurement of aortic pressure, a fluid filled polyethylene catheter was inserted in the right carotid artery, advanced into the aortic arch and connected to a pressure transducer. Via a fluid filled polyethylene catheter, inserted in the jugular vein, potassium chloride (KCl: 0.08 mg/kg) was administered and subsequently CA induced. CA was defined as a mean arterial pressure (MAP) of less than 20 mmHg and no ECG activity and lasted for 8 min. Thereafter the mice were resuscitated by manual chest compression (~300/min) and a bolus 0.1 ml epinephrine (0.1 mg/ml i.v.). Resuscitation ended after 15 min or upon ROSC, defined as sinus rhythm on ECG and mean arterial pressure >40 mmHg lasting at least 5 min.

s than 20 mmHg and no ECG activity and lasted for 8 min. Thereafter the mice were resuscitated by manual chest compression (~300/min) and a bolus 0.1 ml epinephrine (0.1 mg/ml i.v.). Resuscitation ended after 15 min or upon ROSC, defined as sinus rhythm on ECG and mean arterial pressure >40 mmHg lasting at least 5 min. Results: 17 out of 21 (81 %) mice in the sevoflurane group achieved ROSC within 90 seconds after the start of resuscitation. In contrast, none of the 6 mice sedated and anesthetized with FMA achieved ROSC within 15 minutes after CA. No significant differences in aortic pressure, heart rate or body temperature was observed before CA. Conclusion: Sevoflurane is the best choice for anesthetics in a mouse CA model with successful resuscitation. However, our data suggest a profound postconditioning effect of sevoflurane which should be investigated in further detail. A448 Hydrochlorothiazide in ICU-acquired hypernatremia - a randomized controlled trial M.C.O. van IJzendoorn1,2, H. Buter1, W.P. Kingma1, G.J. Navis2, E.C. Boerma1 1Medical Center Leeuwarden, Intensive Care, Leeuwarden, Netherlands; 2University Medical Center Groningen, Internal Medicine / Nephrology, Groningen, Netherlands Correspondence: M.C.O. van IJzendoorn - Medical Center Leeuwarden, Intensive Care, Leeuwarden, Netherlands Introduction: Impaired renal sodium excretion may contribute to ICU-acquired hypernatremia (IAH). Thiazides are suggested as an effective treatment by interfering in renal sodium reabsorption [1].

gen, Netherlands Correspondence: M.C.O. van IJzendoorn - Medical Center Leeuwarden, Intensive Care, Leeuwarden, Netherlands Introduction: Impaired renal sodium excretion may contribute to ICU-acquired hypernatremia (IAH). Thiazides are suggested as an effective treatment by interfering in renal sodium reabsorption [1]. Objectives: Primary aim of the study was to reduce serum sodium concentration (sNa) in patients with IAH, defined as sNa ≥ 143 mmol/l, with hydrochlorothiazide (HCT) in comparison to placebo. Secondary endpoints were a difference in urine sodium concentration (uNa) and the duration of severe IAH, defined as sNa ≥ 145 mmol/l. Methods: A monocentric, double-blind placebo-controlled trial was conducted in 50 patients with IAH and an impaired renal cation excretion, defined as (urine potassium + uNa) < sNa in a spot urine sample. Patients were randomized for enteral HCT 25 mg (n = 25) or placebo (n = 25) 1qd for a maximum of 7 days. Patients on renal replacement therapy, on medications inducing diabetes insipidus or with recent use of diuretics were excluded. The trial received no funding and was registered at clinicaltrials.gov (NTC01974739).

Patients were randomized for enteral HCT 25 mg (n = 25) or placebo (n = 25) 1qd for a maximum of 7 days. Patients on renal replacement therapy, on medications inducing diabetes insipidus or with recent use of diuretics were excluded. The trial received no funding and was registered at clinicaltrials.gov (NTC01974739). Results: Baseline characteristics in both groups were comparable. At baseline median sNa was 146 mmol/l in both groups ([145–148] in the HCT-group, [144–150] in the placebo group, p = 0.4). Over time sNa decreased significantly with 4 mmol/l in both groups, with no significant difference between groups. At baseline uNa was 25 [10–62]mmol/l in the HCT-group and 20 [10–66]mmol/l in the placebo group. Over time uNa increased significantly in both groups: 46 [16–86]mmol/l in the HCT-group and 36 [9–78]mmol/l in the placebo group, with no difference between groups (Fig. 17, Table 7). Median duration of sNa ≥ 145 mmol/l was 3 days in both groups (p = 0.91). No adverse advents or side effects were registered. Conclusions: HCT 25 1qd did not significantly affect sNa nor uNa in patients with IAH. References 1. Overgaard-Steensen C, Ring T. (2013) Clinical review: Practical approach to hyponatraemia and hypernatraemia in critically ill patients. Crit Care. 17(1):206.Fig. 17 (abstract A448). Decrease sNa and increase uNa at study end Table 7 (abstract A448). Results at last day of study inclusion

Conclusions: HCT 25 1qd did not significantly affect sNa nor uNa in patients with IAH. References 1. Overgaard-Steensen C, Ring T. (2013) Clinical review: Practical approach to hyponatraemia and hypernatraemia in critically ill patients. Crit Care. 17(1):206.Fig. 17 (abstract A448). Decrease sNa and increase uNa at study end Table 7 (abstract A448). Results at last day of study inclusion HCT (n = 25) Placebo (n = 25) P-value Serum [Na+], mmol/l 141 [137–147] 144 [139–146] 0.32 Urine [Na+], mmol/l 110 [70–124] 84 [52–126] 0.34 Decrease in serum [Na+], mmol 4 [1–9] 4 [2–6] 0.47 Increase in urine [Na+], mmol 46 [26–86] 36 [9–78] 0.31 A449 Mass methanol poisoning in the czech republic 2012/2013. Hospital costs and clinical outcome J. Rulisek1, M. Balik2,3, S. Zacharov4 11st Medical Faculty, Charles University, Anaesthesia and Intensive Care, Prague, Czech Republic; 2Vseobecna Fakutni Nemocnice, KARIM, Prague, Czech Republic; 3General University Hospital, 1st Faculty of Medicine, Charles University in Prague, Prague, Czech Republic; 41st Medical Faculty, Charles University, Prague, Czech Republic Correspondence: J. Rulisek - 1st Medical Faculty, Charles University, Anaesthesia and Intensive Care, Prague, Czech Republic Introduction: Most of mass methanol outbreaks are reported in underdeveloped countries were financial resources are scarce. In The Czech Republic, in 2012–2013 the mass methanol outbreak was reported with 138 victims. Finding optimal cost-effective treatment is cornerstone for every medical system.

ague, Czech Republic Introduction: Most of mass methanol outbreaks are reported in underdeveloped countries were financial resources are scarce. In The Czech Republic, in 2012–2013 the mass methanol outbreak was reported with 138 victims. Finding optimal cost-effective treatment is cornerstone for every medical system. Objectives: Primary objective of our study was to define total hospital cost, find major parameters which influence total hospital cost and its relation to clinical outcome. Choice of antidote (ethanol or fomepizol) and choice of elimination technique (continuous or intermittent hemodialysis), its impact to hospital costs and clinical outcome were analyzed. Methods: In cooperation with Czech Toxicology Information Centre (TIC) and Czech medical insurance companies reimbursed medical costs were defined. 54 survivors fulfilled follow up investigation. Hospital costs and its correlation to admission parameters was provided. Effectiveness of treatment modalities and consequent medical costs were analyzed. Results: The median hospital costs were 2,422 (interquartile range, IQR: 1,364-4,748) euros. The median one-year medical costs were 633(IQR 439–1,228) euros. Anion gap, pH, HCO3-, lactate, serum creatinine, methanol level, decreased GCS strongly correlated with total hospital costs. All non-survivors were comatose at admission. Fomepizol treatment increase adjusted hospital costs 2,8 times without difference in clinical outcome. Choice of intermittent hemodialysis led to faster residual methanol clearance, formic acid clearance and showed trend to decreased hospital costs.

th total hospital costs. All non-survivors were comatose at admission. Fomepizol treatment increase adjusted hospital costs 2,8 times without difference in clinical outcome. Choice of intermittent hemodialysis led to faster residual methanol clearance, formic acid clearance and showed trend to decreased hospital costs. Conclusions: We were able to define major admission parameters that have impact to total hospital costs. Fomepizol antidote significantly increases costs without impact to clinical outcome. Intermittent hemodialysis shortens secondary elimination time and tend to lower costs. A450 Risk factor of linezolid-induced thrombocytopenia in the intensive care unit H.S. Kim1, S.J. Jeon1, H. Namgung1, E. Lee1, E. Lee2, Y.-J. Cho3, Y.J. Lee3 1Seoul National University Bundang Hospital, Department of Pharmacy, Seongnam-si, Republic of Korea; 2Seoul National University, College of Pharmacy, Seoul, Republic of Korea; 3Seoul National University Bundang Hospital, Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seongnam-si, Republic of Korea Correspondence: H.S. Kim - Seoul National University Bundang Hospital, Department of Pharmacy, Seongnam-si, Republic of Korea Introduction: Few studies have evaluated the prognostic importance of linezolid-induced thrombocytopenia in intensive care unit (ICU) patients because there are many confounding conditions such as sepsis, disseminated intravascular coagulation, and low thrombocytopenia level when admitted to the ICU.

Republic of Korea Introduction: Few studies have evaluated the prognostic importance of linezolid-induced thrombocytopenia in intensive care unit (ICU) patients because there are many confounding conditions such as sepsis, disseminated intravascular coagulation, and low thrombocytopenia level when admitted to the ICU. Objectives: We determined the prevalence and risk factors of linezolid-induced thrombocytopenia in the ICU. Methods: Patients treated with linezolid from January 2005 to December 2015 in the ICU were reviewed retrospectively. To differentiate other causes of thrombocytopenia, the Naranjo algorithm was used to determine the degree of probability for every linezolid-induced thrombocytopenia.

Objectives: We determined the prevalence and risk factors of linezolid-induced thrombocytopenia in the ICU. Methods: Patients treated with linezolid from January 2005 to December 2015 in the ICU were reviewed retrospectively. To differentiate other causes of thrombocytopenia, the Naranjo algorithm was used to determine the degree of probability for every linezolid-induced thrombocytopenia. Results: Sixty patients were included in this study. The mean (SD) age of the patients was 69.8 (±11.9) years. Twenty-nine patients (48.3 %) presented linezolid-induced thrombocytopenia. Patients with thrombocytopenia had higher incidence of any malignancy (41.4 % vs. 9.7 % p = 0.007), elevated baseline creatinine (1.7 [0.9-2.5] vs. 0.9 [0.6-1.3], median [IQR] mg/dL, p = 0.042) and lowered platelet on the first day of linezolid administration (160 [128–230] vs. 194 [118–285] *103/mm3 median [IQR], p = 0.296). Patients who developed thrombocytopenia received more transfusions (34.5 % vs. 6.5 %, p = 0.009) and showed increased ICU mortality (62.1 % vs. 32.3 %, p = 0.037). By means of a logistic regression model, higher incidence of any malignancy (odds ratio, 9.060, 95 % confidence interval, 1.942-42.271) and elevated serum creatinine level after linezolid administration (odds ratio, 1.655, 95 % confidence interval, 1.027-2.669) were significant risk factors of linezolid-induced thrombocytopenia in the ICU. Conclusions: Patients with any malignancy or elevated baseline creatinine who were treated with linezolid in the ICU are more likely to develop thrombocytopenia. References

Results: Sixty patients were included in this study. The mean (SD) age of the patients was 69.8 (±11.9) years. Twenty-nine patients (48.3 %) presented linezolid-induced thrombocytopenia. Patients with thrombocytopenia had higher incidence of any malignancy (41.4 % vs. 9.7 % p = 0.007), elevated baseline creatinine (1.7 [0.9-2.5] vs. 0.9 [0.6-1.3], median [IQR] mg/dL, p = 0.042) and lowered platelet on the first day of linezolid administration (160 [128–230] vs. 194 [118–285] *103/mm3 median [IQR], p = 0.296). Patients who developed thrombocytopenia received more transfusions (34.5 % vs. 6.5 %, p = 0.009) and showed increased ICU mortality (62.1 % vs. 32.3 %, p = 0.037). By means of a logistic regression model, higher incidence of any malignancy (odds ratio, 9.060, 95 % confidence interval, 1.942-42.271) and elevated serum creatinine level after linezolid administration (odds ratio, 1.655, 95 % confidence interval, 1.027-2.669) were significant risk factors of linezolid-induced thrombocytopenia in the ICU. Conclusions: Patients with any malignancy or elevated baseline creatinine who were treated with linezolid in the ICU are more likely to develop thrombocytopenia. References 1. Zhang Z, Liang Z, Li H, Chen L, She D (2013) Comparative evaluation of thrombocytopenia in adult patients receiving linezolid or glycopeptides in a respiratory intensive care unit. Exp Ther Med 7(2):501–507. doi: 10.3892/etm.2013.1437 2. James N George (2015) Drug-induced thrombocytopenia UptodateWeb. http://www.uptodate.com/contents/drug-induced-thrombocytopenia. Accessed 20 February 2015.

1. Zhang Z, Liang Z, Li H, Chen L, She D (2013) Comparative evaluation of thrombocytopenia in adult patients receiving linezolid or glycopeptides in a respiratory intensive care unit. Exp Ther Med 7(2):501–507. doi: 10.3892/etm.2013.1437 2. James N George (2015) Drug-induced thrombocytopenia UptodateWeb. http://www.uptodate.com/contents/drug-induced-thrombocytopenia. Accessed 20 February 2015. 3. John Papadopoulos (2012) Drug-Induced Complications in the Critically Ill Patient: A Guide for Recognition and Treatment, New York 4. Naranjo CA, Busto U, Sellers EM, Sandor P, Ruiz I, Roberts EA, Janecek E, Domecq C, Greenblatt DJ (1981) A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther. 30(2):239–245. Grant acknowledgement I wish to express my sincere thanks to Eunsook Lee, the chief of department of pharmacy, for providing me with all the necessary facilities. I take this opportunity to record my sincere thanks to all the faculty members of the ICU multidisciplinary team for their help and encouragement.

4. Naranjo CA, Busto U, Sellers EM, Sandor P, Ruiz I, Roberts EA, Janecek E, Domecq C, Greenblatt DJ (1981) A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther. 30(2):239–245. Grant acknowledgement I wish to express my sincere thanks to Eunsook Lee, the chief of department of pharmacy, for providing me with all the necessary facilities. I take this opportunity to record my sincere thanks to all the faculty members of the ICU multidisciplinary team for their help and encouragement. A451 Physiological and biochemical effects of an intravenous furosemide bolus in critically ill patients A. Huang1,2, L. Cioccari1,3, N. Luethi1, J. Mårtensson1,4, R. Bellomo1,5 1Austin Health, University of Melbourne, Department of Intensive Care, Heidelberg, Australia; 2Changi General Hospital, Department of Anaesthesia and Surgical Intensive Care, Singapore, Singapore; 3Lucerne Cantonal Hospital, Department of Intensive Care, Lucerne, Switzerland; 4Karolinska Institutet, Section of Anaesthesia and Intensive Care Medicine, Department of Physiology and Pharmacology, Stockholm, Sweden; 5The University of Melbourne, Parkville, Melbourne, Australia Correspondence: L. Cioccari - Austin Health, University of Melbourne, Department of Intensive Care, Heidelberg, Australia Introduction: Intravenous boluses of furosemide are commonly given to critically ill patients to modulate fluid balance1–3. However, their effect on urinary output (UO), fluid balance, electrolyte losses and hemodynamics has not been systematically studied. Therefore, we sought to compare UO, fluid balance, electrolyte losses, biochemical, and hemodynamic effects in the six hours before and after the administration of a 40 mg IV bolus of furosemide.

their effect on urinary output (UO), fluid balance, electrolyte losses and hemodynamics has not been systematically studied. Therefore, we sought to compare UO, fluid balance, electrolyte losses, biochemical, and hemodynamic effects in the six hours before and after the administration of a 40 mg IV bolus of furosemide. Methods: We conducted a prospective observational study in critically ill patients where the attending clinician had decided to administer a 40 mg bolus of IV furosemide. We collected information on UO, fluid balance, urinary electrolyte losses, serum and urinary creatinine, serum biochemistry, and hemodynamics and compared the 6-hour pre-bolus period with the 6-hour post-bolus period. We performed linear regression analysis to identify predictors of the extent of urinary output response.

collected information on UO, fluid balance, urinary electrolyte losses, serum and urinary creatinine, serum biochemistry, and hemodynamics and compared the 6-hour pre-bolus period with the 6-hour post-bolus period. We performed linear regression analysis to identify predictors of the extent of urinary output response. Results: We studied 24 patients (13 females) with a median (IQR) age of 72 (64, 81) years. The intravenous bolus induced a > 1000 ml increase in urine output (UO) in 6 (25 %) patients, a 500–1000 ml increase in 7 (29.2 %) patients, a < 500 ml increase in 9 (37.5 %) patients, and no increase in 2 (8.3 %) patients. The median UO increased in 23 (96 %) patients and failed to increase in 1 (4 %) patient for a median difference in UO of 561 (250, 1046) ml. The 6-hour fluid balance became negative in 16 (72.7 %) patients and positive in 8 (36.3 %) patients, with a median change of −520 (−1206, 98) ml. However, we observed no difference in free water clearance. Furosemide significantly increased urinary sodium, potassium and chloride losses by 67 (31, 129) mmol, 9.4 (1.8, 18) mmol and 61 (44, 138) mmol respectively, and decreased blood chloride levels by 1.5 (0.5, 3.0) mmol/l. There were no detectable changes in hemodynamics. On linear regression analysis, UO increased by 37 ml (95 % CI 19 to 56 ml) / 6 hour for every 1 mmHg increase in mean arterial pressure and decreased by 35 ml (95 % CI 9 to 61 ml) / 6 hour for every 1 g/L increase in serum albumin.

oride levels by 1.5 (0.5, 3.0) mmol/l. There were no detectable changes in hemodynamics. On linear regression analysis, UO increased by 37 ml (95 % CI 19 to 56 ml) / 6 hour for every 1 mmHg increase in mean arterial pressure and decreased by 35 ml (95 % CI 9 to 61 ml) / 6 hour for every 1 g/L increase in serum albumin. Conclusions: We defined the overall median effect of furosemide on urine output, urinary sodium, potassium and chloride losses and fluid balance in critically ill patients and provided information on how such effects were modulated by mean arterial pressure and albumin levels. Our findings are likely to assist clinicians in their therapeutic expectations and choices. References 1. Bagshaw SM, Delaney A, Jones D, et al. Diuretics in the management of acute kidney injury: a multinational survey. Contributions to nephrology 2007;156:236–249. 2. Mehta RL, Pascual MT, Soroko S, et al. Diuretics, mortality, and nonrecovery of renal function in acute renal failure. Jama 2002;288(20):2547–2553. 3. Uchino S, Doig GS, Bellomo R, et al. Diuretics and mortality in acute renal failure. Critical care medicine 2004;32(8):1669–1677. Grant acknowledgement None.

1. Bagshaw SM, Delaney A, Jones D, et al. Diuretics in the management of acute kidney injury: a multinational survey. Contributions to nephrology 2007;156:236–249. 2. Mehta RL, Pascual MT, Soroko S, et al. Diuretics, mortality, and nonrecovery of renal function in acute renal failure. Jama 2002;288(20):2547–2553. 3. Uchino S, Doig GS, Bellomo R, et al. Diuretics and mortality in acute renal failure. Critical care medicine 2004;32(8):1669–1677. Grant acknowledgement None. A452 No support for the use of intravenous lipid emulsion in oral poisoning -a systematic review and analysis of 169 published cases M. Forsberg1, G. Edman1, J. Höjer2,3, S. Forsberg3,4 1TioHundra AB, Norrtälje, Sweden; 2Swedish Poisons Information Centre, Stockholm, Sweden; 3Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden; 4TioHundra AB, Anaesthesia and Intensive Care, Norrtälje, Sweden Correspondence: M. Forsberg - TioHundra AB, Norrtälje, Sweden Introduction: In 1998, Weinberg and colleagues reported results from a study that pointed towards reduced bupivacaine toxicity by intravenous lipid emulsion (ILE) in rats[1]. The use of ILE as an antidote has increased dramatically and associations in several countries have launched guidelines regarding the use of ILE in local anaesthetic systemic toxicity (LAST). Recently, two systematic reviews from the international expert team 'Lipid Emulsion Therapy Workgroup' were published[2,3]. The review on ILE for LAST concluded that treatment appears to be effective in some cases of LAST, but there is no evidence showing that ILE is more effective than vasopressors. The review on ILE for non-local anaesthetic toxicity concluded that the quality of evidence for ILE being an effective antidote remains low to very low.

on ILE for LAST concluded that treatment appears to be effective in some cases of LAST, but there is no evidence showing that ILE is more effective than vasopressors. The review on ILE for non-local anaesthetic toxicity concluded that the quality of evidence for ILE being an effective antidote remains low to very low. Objectives: We present a systematic review including all published reports regarding humans treated with lipid rescue from 1998 until 2016. Methods: Web of science and PubMed were used for the literature search. To limit the search we used predefined criteria. This resulted in 195 papers, which were independently reviewed and categorized by using a modified version of WHO-UMC causality categories[4]. This is a 6-graded assessment scale that combines clinical and pharmacological aspects with the quality of documentation. Grade 1 implies a certain causal connection, 2 probable, 3 possible and 4 unlikely. Grade 5 implies none or negative effect of ILE and 6 unassessable. Log p-values were identified for all involved toxins. Further, the reviewing experts evaluated which main symptom reasonable had led to the decision of giving ILE.

tion. Grade 1 implies a certain causal connection, 2 probable, 3 possible and 4 unlikely. Grade 5 implies none or negative effect of ILE and 6 unassessable. Log p-values were identified for all involved toxins. Further, the reviewing experts evaluated which main symptom reasonable had led to the decision of giving ILE. Results: In total, 135 case reports received grade 1–4 and 34 grade 5. Forty of the 135 case reports that had reported a positive effect of ILE were parenteral intoxications while 95 were oral poisonings. Three of 135 cases were classified as a certain correlation between giving ILE and improvement of symptoms, 34 cases as probable correlation, 71 possible and 27 unlikely. Mean classification grade in parenteral intoxications was 2.3. Mean classification grade among the oral poisonings was 3.1, which was a highly significant difference. None of the oral poisonings received the classification grade 'certain'. The parenteral intoxications had a mean log p-value of 4.2 which was significantly higher than that for oral poisonings which was 3.2. Indications for ILE in parenteral intoxications were cardiac arrest and neurological symptoms and for oral poisonings hypotension and arrhythmia. Conclusions: The evidence for ILE being an effective antidote is sparse. Thirty-four out of 135 published cases indicates no or negative effect, despite the immense impact of publication bias. It seems reasonable not to use ILE in oral poisonings. References 1. Weinberg Anesthesiology 1998;88:1071–5. 2. Hoegberg Clin Toxicol 2016;epubl 3. Levine Clin Toxicol 2016;epubl

Conclusions: The evidence for ILE being an effective antidote is sparse. Thirty-four out of 135 published cases indicates no or negative effect, despite the immense impact of publication bias. It seems reasonable not to use ILE in oral poisonings. References 1. Weinberg Anesthesiology 1998;88:1071–5. 2. Hoegberg Clin Toxicol 2016;epubl 3. Levine Clin Toxicol 2016;epubl 4. WHO-UMC http://who-umc.org/Graphics/24734.pdf. A453 Fondaparinux treatment in patients with acute Heparin-induced thrombocytopenia and acute renal failure with continuous renal replacement therapy M.T. Chiquito Freile1, F.N. Hidalgo1, J.A. Martinez Molina1, R. Lecumberri2, A. Figuerola Rosselló1, P. Medrano Travieso1 1Clinica Universidad de Navarra, Anesthesiology, Pamplona, Spain; 2Clinica Universidad de Navarra, Hematology, Pamplona, Spain Correspondence: M.T. Chiquito Freile - Clinica Universidad de Navarra, Anesthesiology, Pamplona, Spain Introduction: Inmune Heparin-induced thrombocytopenia (HIT) is a prothrombotic disorder caused by an immune-mediated complication that results from a platelet activating inmune response triggered by the interaction of heparin with a specific platelet protein, platelet factor 4 (PFA). Fondaparinux is a safety treatment of HIT, even though is not recommended in acute renal failure (ARF) because of it renal elimination and significant risk of accumulation. Objectives: Report the safety profile and efficacy of fondaparinux (thrombotic and major bleeding complications) in patients with HIT and ARF during continuous renal replacement therapy (CRRT).

A453 Fondaparinux treatment in patients with acute Heparin-induced thrombocytopenia and acute renal failure with continuous renal replacement therapy M.T. Chiquito Freile1, F.N. Hidalgo1, J.A. Martinez Molina1, R. Lecumberri2, A. Figuerola Rosselló1, P. Medrano Travieso1 1Clinica Universidad de Navarra, Anesthesiology, Pamplona, Spain; 2Clinica Universidad de Navarra, Hematology, Pamplona, Spain Correspondence: M.T. Chiquito Freile - Clinica Universidad de Navarra, Anesthesiology, Pamplona, Spain Introduction: Inmune Heparin-induced thrombocytopenia (HIT) is a prothrombotic disorder caused by an immune-mediated complication that results from a platelet activating inmune response triggered by the interaction of heparin with a specific platelet protein, platelet factor 4 (PFA). Fondaparinux is a safety treatment of HIT, even though is not recommended in acute renal failure (ARF) because of it renal elimination and significant risk of accumulation. Objectives: Report the safety profile and efficacy of fondaparinux (thrombotic and major bleeding complications) in patients with HIT and ARF during continuous renal replacement therapy (CRRT). Methods: We reviewed consecutive elegible patients with HIT and ARF in one medical university center (Clínica Universidad de Navarra) over a 5 year period, who received fondaparinux for anticoagulation at the time of the HIT diagnosis. Results:

Objectives: Report the safety profile and efficacy of fondaparinux (thrombotic and major bleeding complications) in patients with HIT and ARF during continuous renal replacement therapy (CRRT). Methods: We reviewed consecutive elegible patients with HIT and ARF in one medical university center (Clínica Universidad de Navarra) over a 5 year period, who received fondaparinux for anticoagulation at the time of the HIT diagnosis. Results: - Of the 12 patients recorded that received fondaparinux as anticoagulation therapy, 5 of them shared the criteria of HIT and ARF with CRRT. We analyzed four of the five patients, excluding one because of loss of information. - The mean age of the patients was 67 yo, three men and one woman. All of them present acute renal failure with a mean basal creatinine of 2.57, meeting the criteria of CRRT. - All patients had cardiac surgery with heparin administration last more than 24 hrs with a mean of 4.5 days. - Platelets at admission were documented and in all cases they dropped more than 50 % of the original value. - After the administration of Fondaparinux, all the patients had a recovery of the platelet count until the basal value, with a mean of 28.25 days. - None of the patients developed new, recurrent or progressive thrombosis. No major bleeding complications were documented. Conclusions: Fondaparinux is a safety alternative of anticoagulation therapy in patients with HIT and ARF during continuos renal replacement therapy. References - Yasser Sakr, Heparin-induced thrombocytopenia in the ICU: an overview. Crit Care.2011; 15:211

- None of the patients developed new, recurrent or progressive thrombosis. No major bleeding complications were documented. Conclusions: Fondaparinux is a safety alternative of anticoagulation therapy in patients with HIT and ARF during continuos renal replacement therapy. References - Yasser Sakr, Heparin-induced thrombocytopenia in the ICU: an overview. Crit Care.2011; 15:211 - Honore et al. Fondaparinux: another potential treatment for heparin-induced thrombocytopenia type II. Crit Care. 2016; 20:14 - T.E. Warkentin et al., Fondaparinux treatment of acute heparin-induced thrombocytopenia confirmed by the serotonin-release assay: a 30-month, 16 patient case series. Journal of Thrombosis and Haemostasis, 9: 2389–2396 A454 Hyperthermia and multiple organ dysfunction syndrome associated with methamphetamine derivatives and 3,4-methylenedioxymethamphetamine ("ecstasy") G. Tuero Leon1, J. Gonzalez Sanchez1, L. Sahuquillo Frias2, D. Balsells Rosello2, J.A. Garcia Verdejo1, J.A. Noria Serrano1 1Hospital Can Misses, Intensive Care Unit, Ibiza, Spain; 2Hospital Can Misses, Clinical Analysis Service, Ibiza, Spain Correspondence: G. Tuero Leon - Hospital Can Misses, Intensive Care Unit, Ibiza, Spain Introduction: Hyperthermia with rhabdomyolysis and Multiple Organ Dysfunction Syndrome (MODS) is a rare, life-threatening complication of methamphetamine derivatives (MA) and 3,4-methylenedioxymethamphetamine (MDMA) consumption. The association between hyerthermia and MODS was established from case reports subject to publication bias.

perthermia with rhabdomyolysis and Multiple Organ Dysfunction Syndrome (MODS) is a rare, life-threatening complication of methamphetamine derivatives (MA) and 3,4-methylenedioxymethamphetamine (MDMA) consumption. The association between hyerthermia and MODS was established from case reports subject to publication bias. Objectives: To explore the association between hyperthermia and MODS development in patients requiring ICU admission due to MA/MDMA intoxication. Methods: Patients with a positive urine test for MA (Triage TOX Drug Screen. Biosite. USA) who required admission to ICU at Can Misses Hospital between the 1st of January 2004 and the 31th of December 2014 were identified. Patients admitted for major trauma, near-drowning or acute coronary syndrome were excluded. Vital signs, clinical data, toxicological results, time prior to ICU admission and mortality at discharge from ICU were recorded. Only tympanic temperatures were considered. SAPS-3 and daily SOFA scores were calculated. Hyperthermia was set as a tympanic temperature above 38,9 °C. MODS was defined by the presence of a score of 3 or 4 in at least two items of the SOFA scale, excluding the neurological at 48 hour of admission or at the time of death.

ly tympanic temperatures were considered. SAPS-3 and daily SOFA scores were calculated. Hyperthermia was set as a tympanic temperature above 38,9 °C. MODS was defined by the presence of a score of 3 or 4 in at least two items of the SOFA scale, excluding the neurological at 48 hour of admission or at the time of death. Results: Forty patients were identified after excluding three according to pre-established criteria. They were predominantly young (26 years (19–41)) male (33 patients) with no previous medical history (33 patients). The co-use of other substances was common. The main cause of admission was altered level of consciousness (35 patients), 25 (62.5 %) were in coma. Four cases were admitted due to an out-of-hospital cardiac arrest. Seventeen patients (42.5 %) presented hyperthermia, of which eleven had MODS (4 died within the first 24 hours). No patient with a maximum temperature below 39 °C developed MODS. Six patients with hyperthermia (35.2 %) had a favorable outcome without MODS. Among them, the maximum temperature (40.62 ± 0,7 vs 41.46 ± 0.7 °C, p 0,03), the maximum value of CPK (3.967 ± 1.283 vs 40.387 ± 1.027 U/L, p 0.01), the SAPS-3 (45,5 ± 4 vs 80,7 ± 3, p < 0,001) and SOFA scores (6 (0–9) vs 10 (6–19), p 0,001 at 24 hours; 2 (0–8) vs 15 (13–17), p 0,003 at 48 hours) were significantly lower. No differences in origin (e.g. club), co-use of other substances or temperature control measures were observed among this group. There was a trend toward a greater delay in admission to ICU (176 ± 77 vs 103 ± 9 min, p 0.76) that was not statistically significant among this group. All patients admitted due to out-of-hospital cardiac arrest presented with hyperthermia (41,8 ± 0,5 °C) and MODS (3 died within the first 24 hours).

d among this group. There was a trend toward a greater delay in admission to ICU (176 ± 77 vs 103 ± 9 min, p 0.76) that was not statistically significant among this group. All patients admitted due to out-of-hospital cardiac arrest presented with hyperthermia (41,8 ± 0,5 °C) and MODS (3 died within the first 24 hours). Conclusions: 1. Hyperthermia appears to be strongly associated with the presence of MODS in patients with MA/MDMA intoxications.2. There is a subset of patients with hyperthermia who do not develop MODS. This is not fully explainable by differences in peak temperature or in hyperthermia management. A455 Double-check of IV medication at the bedside; does it bring safety or risks? D. Winterwerp1, T. van Galen2 1RN ICU Nurse, VU University Medical Centre, Amsterdam, Netherlands; 2ICU Nursing staff manager, VU University Medical Center, Amsterdam, Netherlands Correspondence: D. Winterwerp - RN ICU Nurse, VU University Medical Centre, Amsterdam, Netherlands Introduction: The Dutch Healthcare Inspectorate and Quality Institutes mandated hospitals to perform a double-check during IV medication preparation and administering at the bedside. Double check during IV med. preparation was already secured into the workflow. The bedside check is an addition to the nursing workload. The ICU nursing staff of VU University medical center (VUmc) is concerned that the bedside check will not be as effective as expected and that implementing this double check will generate disturbances in other ICU processes.

as already secured into the workflow. The bedside check is an addition to the nursing workload. The ICU nursing staff of VU University medical center (VUmc) is concerned that the bedside check will not be as effective as expected and that implementing this double check will generate disturbances in other ICU processes. Objectives: Perform a workload measurement during 4 ICU shifts. Perform a literature study and survey other ICU's. Interview with the dedicated ICU pharmacist. Methods: The workload measurement was performed to clarify how much extra time the bedside check added to the normal nursing task performance. We measured the frequency and time of the bedside check. The literature study was performed to find evidence of experiences and potential problems with the single check versus double check. Interviewing the pharmacist was to inquire the concerning double checks as well as supporting advice. Results: The workload measurement has been performed during four 8-hour shifts. There were 9 to 12 patients admitted at the ICU during these shifts. The N/P ratio was 1:1.5 up to 2.0. The additional check takes 1,5 min. per patient, this is added to the 2 minutes that the IV med. preparation check already takes. The bedside check added 27 interruptions to the 15 interruptions for IV med. preparation because the preparation check included multiple medications simultaneously and the bedside check 1 of 2 IV meds per moment. There are more quantitative studies then qualitative studies [1]. Most studies conclude that more studies need to be conducted [2].

Results: The workload measurement has been performed during four 8-hour shifts. There were 9 to 12 patients admitted at the ICU during these shifts. The N/P ratio was 1:1.5 up to 2.0. The additional check takes 1,5 min. per patient, this is added to the 2 minutes that the IV med. preparation check already takes. The bedside check added 27 interruptions to the 15 interruptions for IV med. preparation because the preparation check included multiple medications simultaneously and the bedside check 1 of 2 IV meds per moment. There are more quantitative studies then qualitative studies [1]. Most studies conclude that more studies need to be conducted [2]. Other Dutch hospitals struggle with the same problem. Some hospitals enlarged pharmacy service to prepare more medication to decrease nurse's workload.

Results: The workload measurement has been performed during four 8-hour shifts. There were 9 to 12 patients admitted at the ICU during these shifts. The N/P ratio was 1:1.5 up to 2.0. The additional check takes 1,5 min. per patient, this is added to the 2 minutes that the IV med. preparation check already takes. The bedside check added 27 interruptions to the 15 interruptions for IV med. preparation because the preparation check included multiple medications simultaneously and the bedside check 1 of 2 IV meds per moment. There are more quantitative studies then qualitative studies [1]. Most studies conclude that more studies need to be conducted [2]. Other Dutch hospitals struggle with the same problem. Some hospitals enlarged pharmacy service to prepare more medication to decrease nurse's workload. Discussion: Evaluating double check workload, it was less time consuming then expected although the frequency of the check brings more interruptions. Every interruption brings a disturbance to nurse's concentration and work, this could potentially increase patient risk. Therefore the amount of interruptions should be reduced to a minimum. Meanwhile, technology makes giant leaps. This has the potential to replace the double check. BCMA is making progress and is more efficient and foul proof than human checks. Because of the wide spread variety in results its hard to draw further Conclusions: It is even doubtful whether further studies make sense, given the current advances in technology. If available, BCMA technique (combined with CPOE) should be implemented as soon as possible. If not available, than hospitals should focus on a select part of medication checking based on specific risks and workflow [3].

: It is even doubtful whether further studies make sense, given the current advances in technology. If available, BCMA technique (combined with CPOE) should be implemented as soon as possible. If not available, than hospitals should focus on a select part of medication checking based on specific risks and workflow [3]. References 1. What is the Evidence on Double Checks? 2012 [2][3]Does a double check for IV medication improve patient safety? Rachel Cox A456 Rational colistin use in the intensive care units of a university affiliated hospital in Iran based on a standard treatment guideline A. Vazin1, I. Karimzade2, A. Zand2 1Shiraz University of Medical Sciences, Clinical Pharmacy Department, Shiraz, Islamic Republic of Iran; 2Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran Correspondence: A. Vazin - Shiraz University of Medical Sciences, Clinical Pharmacy Department, Shiraz, Islamic Republic of Iran Introduction: Inadequate antimicrobial treatment of nosocomial infections is an independent determinant of mortality in the critically ill patient. Multi-drug resistant Gram-negative bacilli, mainly Acinetobacter species, are responsible for a significant proportion of nosocomial infections. Acinetobacter has been found to be resistant to many currently available antimicrobial agents. Recently there has been renewed interest in colistin, which had earlier been abandoned for serious adverse effects. Objectives: To assess rational colistin use in a prospective study at the ICU wards of a university affiliated hospital in Iran

A456 Rational colistin use in the intensive care units of a university affiliated hospital in Iran based on a standard treatment guideline A. Vazin1, I. Karimzade2, A. Zand2 1Shiraz University of Medical Sciences, Clinical Pharmacy Department, Shiraz, Islamic Republic of Iran; 2Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran Correspondence: A. Vazin - Shiraz University of Medical Sciences, Clinical Pharmacy Department, Shiraz, Islamic Republic of Iran Introduction: Inadequate antimicrobial treatment of nosocomial infections is an independent determinant of mortality in the critically ill patient. Multi-drug resistant Gram-negative bacilli, mainly Acinetobacter species, are responsible for a significant proportion of nosocomial infections. Acinetobacter has been found to be resistant to many currently available antimicrobial agents. Recently there has been renewed interest in colistin, which had earlier been abandoned for serious adverse effects. Objectives: To assess rational colistin use in a prospective study at the ICU wards of a university affiliated hospital in Iran Methods: Standard treatment Guideline for colistin use were provided and approved by pharmacy and therapeutics committee and Study criteria were developed to assess the several parameters involved in colistin therapy. These parameters include the appropriateness of indication, dosing regimen, duration of therapy and monitoring for toxicity. Clinical and paraclinical parameters such as Glomerular Filtration Rate (GFR), microbial culture, antibacterial sensitivity, WBC count and fever were collected and recorded for analysis.

therapy. These parameters include the appropriateness of indication, dosing regimen, duration of therapy and monitoring for toxicity. Clinical and paraclinical parameters such as Glomerular Filtration Rate (GFR), microbial culture, antibacterial sensitivity, WBC count and fever were collected and recorded for analysis. Results: One hundred patients were enrolled in this study, including 64 male and 36 female patients. In most of patients(97 %), the origin of infection was hospital acquired.In this study colistin therapy was 87 % specific, in the remaining 13 %, the initial therapy with colistin was empirical. The most frequent infection was ventilator associated pneumonia (69 %). None of the patients received loading dose. The maintenance dose of colistin in 76 %, dosing interval in 71 %, and duration of treatment in 84 %, was appropriate. In this study, 68.12 % of colistin use was compatible with standard treatment guideline. Conclusions: Based on the results of this study,the main observed drawbacks were inappropriate dosing regimen,especially not considering loading dose in life threatening infections, inappropriate maintenance dose, and continuation of colistin, which has to be revised in order to achieve an effective treatment. References 1. Shahbazi F, Dashti-khavidaki S. Colistin: efficacy and safety in different populations,Expert Rev. Clin. Pharmacol.2015:8(4):423–448. 2. Napier BA, Burd EM, Satola SW, et al. Clinical use of colistin induces cross-resistance to host antimicrobials in Acinetobacter baumannii. mBio.2013: 4(3):e00021-13.

Conclusions: Based on the results of this study,the main observed drawbacks were inappropriate dosing regimen,especially not considering loading dose in life threatening infections, inappropriate maintenance dose, and continuation of colistin, which has to be revised in order to achieve an effective treatment. References 1. Shahbazi F, Dashti-khavidaki S. Colistin: efficacy and safety in different populations,Expert Rev. Clin. Pharmacol.2015:8(4):423–448. 2. Napier BA, Burd EM, Satola SW, et al. Clinical use of colistin induces cross-resistance to host antimicrobials in Acinetobacter baumannii. mBio.2013: 4(3):e00021-13. 3.Berlana D, Llop JM, Fort E, et al. Use of colistin in the treatment of multiple-drug-resistant gram-negative infections. Am J Health Syst Pharm 2005; 62:39. 4. Cantón R, Horcajada JP, Oliver A, Garbajosa PR, Vila J. Inappropriate use of antibiotics in hospitals: The complex relationship between antibiotic use and antimicrobial resistance. Enfermedades Infecciosas y Microbiología Clínica. 2013;31:3–11 Grant acknowledgement We would also like to thank Shiraz University of medical sciences for financial support of this study.

bsolute eosinophil count of ≤10/ul or 10-50/ul have an adjusted OR of discharge to a care facility of 1.34 (95%CI, 1.28-1.41; P < 0.001) or 1.30 (95%CI, 1.24-1.36; P < 0.001) relative to patients with an absolute eosinophil count of 50-350/ul. An absolute eosinophil count >350/ul was not associated with study outcomes. Conclusions: In critical illness survivors, eosinopenia at ICU admission is a robust predictor of mortality following discharge, hospital readmission, and discharge to a care facility. Eosinopenia is a marker for ICU survivors at an especially high risk for adverse outcomes. Thus, patients with eosinopenia may benefit from closer post discharge follow-up and higher intensity rehabilitation.

4. Cantón R, Horcajada JP, Oliver A, Garbajosa PR, Vila J. Inappropriate use of antibiotics in hospitals: The complex relationship between antibiotic use and antimicrobial resistance. Enfermedades Infecciosas y Microbiología Clínica. 2013;31:3–11 Grant acknowledgement We would also like to thank Shiraz University of medical sciences for financial support of this study. A457 Pregnancy and intoxication: a retrospective analysis of 18 patients E. Ozen, S. Ekemen, A. Akcan, E. Sen, B. Buyukkidan Yelken 1Eskisehir Osmangazi University Faculty of Medicine, Department of Anesthesiology and Reanimation, Division of Intensive Care, Eskisehir, Turkey Correspondence: E. Ozen - Eskisehir Osmangazi University Faculty of Medicine, Department of Anesthesiology and Reanimation, Division of Intensive Care, Eskisehir, Turkey Objectives: The aim of this study is to evaluate our treatment results in pregnant women presented with intoxication. Methods: The patients admitted to intensive care unit (ICU) with intoxication between April 2010-May 2015 were evaluated retrospectively and 18 pregnant intoxication cases were achieved.

A457 Pregnancy and intoxication: a retrospective analysis of 18 patients E. Ozen, S. Ekemen, A. Akcan, E. Sen, B. Buyukkidan Yelken 1Eskisehir Osmangazi University Faculty of Medicine, Department of Anesthesiology and Reanimation, Division of Intensive Care, Eskisehir, Turkey Correspondence: E. Ozen - Eskisehir Osmangazi University Faculty of Medicine, Department of Anesthesiology and Reanimation, Division of Intensive Care, Eskisehir, Turkey Objectives: The aim of this study is to evaluate our treatment results in pregnant women presented with intoxication. Methods: The patients admitted to intensive care unit (ICU) with intoxication between April 2010-May 2015 were evaluated retrospectively and 18 pregnant intoxication cases were achieved. Results: Median age was 29 (22–40). Median pregnancy week was 15.5 (5–32). Eleven (61.1 %) patients were multi drug intoxication, 5 (27.8 %) patients were carbon monoxide (CO) intoxication and 2 patients were mushroom intoxication. The drugs that were taken in order to commit suicide were mostly antidepressants, antibiotics and analgesics. Median duration of stay in ICU was 1.5 (1–5) days. Median Glasgow Coma Score (GCS) was 15 (14–15). Additional comorbid diseases were positive in 4 (22.2 %) patients. Hyperbaric oxygen therapy (HBOT) was applied to all patients with CO intoxication. Median application number of HBOT was 1 (1–3). Activated carbon, proton pump inhibitors for gastric protection and medical treatment were given to all patients with multi drug intoxication. The 2 patients with mushroom intoxication received only symptomatic treatment.

) was applied to all patients with CO intoxication. Median application number of HBOT was 1 (1–3). Activated carbon, proton pump inhibitors for gastric protection and medical treatment were given to all patients with multi drug intoxication. The 2 patients with mushroom intoxication received only symptomatic treatment. When we compare the pre and post treatment AST, ALT and CK values of the patients with multi drug intoxication there were significant decrease. But, there were no differences between troponin values. In the patients with CO intoxication there were no significant differences between the pre and post treatment carboxyhemoglobin (COHb) values (Table 8). All patients were discharged with healing. Seventeen of the patients had no problem related with their pregnancy and they gave birth at term (37–38 weeks). All babies were healthy at birth and they did not have any problem associated with intoxication. One of the patients ended her pregnancy with her own decision in her early pregnancy. Conclusions: In conclusion, we did not observe any severe complication in clinically stable intoxication patients with normal laboratory findings in their pregnancy.Table 8 (abstract A457). Comparison between pre and post treatment Pretreatment Posttreatment p Mean (Range) Mean (Range) AST 14.78 (11.00–20.20) 14.45 (9.00–20.00) <0.001 Drug intoxication ALT 9.76 (5.98–20.62) 9.00 (6.00–13.00) <0.001 CK 53.80 (33.00–85.00) 52.09 (30.00–120.00) <0.001 Troponin 0.03 (0.01–0.07) 0.016 (0.00–0.03) NS Carbon monoxide intoxication COHb 14.25 (0.60–27.90) 2.70 (0.60–4.60) NS

atment p Mean (Range) Mean (Range) AST 14.78 (11.00–20.20) 14.45 (9.00–20.00) <0.001 Drug intoxication ALT 9.76 (5.98–20.62) 9.00 (6.00–13.00) <0.001 CK 53.80 (33.00–85.00) 52.09 (30.00–120.00) <0.001 Troponin 0.03 (0.01–0.07) 0.016 (0.00–0.03) NS Carbon monoxide intoxication COHb 14.25 (0.60–27.90) 2.70 (0.60–4.60) NS A458 Alcohol-related major traumatic brain injury: a province-wide retrospective analysis N. Kureshi1,2, L. Fenerty1, G. Thibault-Halman1, M. Erdogan3, S. Walling1, R.S. Green2,3, D.B. Clarke1 1Dalhousie University, Queen Elizabeth II Health Sciences Centre, Division of Neurosurgery, Halifax, Canada; 2Dalhousie University, Critical Care, Halifax, Canada; 3Trauma Nova Scotia, Halifax, Canada Correspondence: R.S. Green - Dalhousie University, Critical Care, Halifax, Canada Introduction: Traumatic brain injury (TBI) is a leading cause of death and disability. In the Canadian province of Nova Scotia, TBI occurs in approximately 50 % of major trauma seen annually. Although alcohol use increases the risk of experiencing TBI, it remains unclear whether outcomes in alcohol-impaired TBI patients are different from those of unimpaired patients. Objectives: The objective of this study was to describe the characteristics and patterns of major TBI seen over a 14-year period at the provincial level in Canada. We evaluated the effect of alcohol on length of stay (LOS) and mortality in patients with major TBI.

A458 Alcohol-related major traumatic brain injury: a province-wide retrospective analysis N. Kureshi1,2, L. Fenerty1, G. Thibault-Halman1, M. Erdogan3, S. Walling1, R.S. Green2,3, D.B. Clarke1 1Dalhousie University, Queen Elizabeth II Health Sciences Centre, Division of Neurosurgery, Halifax, Canada; 2Dalhousie University, Critical Care, Halifax, Canada; 3Trauma Nova Scotia, Halifax, Canada Correspondence: R.S. Green - Dalhousie University, Critical Care, Halifax, Canada Introduction: Traumatic brain injury (TBI) is a leading cause of death and disability. In the Canadian province of Nova Scotia, TBI occurs in approximately 50 % of major trauma seen annually. Although alcohol use increases the risk of experiencing TBI, it remains unclear whether outcomes in alcohol-impaired TBI patients are different from those of unimpaired patients. Objectives: The objective of this study was to describe the characteristics and patterns of major TBI seen over a 14-year period at the provincial level in Canada. We evaluated the effect of alcohol on length of stay (LOS) and mortality in patients with major TBI. Methods: This was a retrospective case series. Data were obtained from the Nova Scotia Trauma Registry for all patients presenting with major TBI (abbreviated injury score [AIS] head ≥3) in Nova Scotia hospitals between 2002 and 2015. Injury rates were calculated on the basis of 100,000 population using population estimates from Statistics Canada. Patients were compared by blood alcohol concentration (BAC) at time of injury: negative (0–1.9 mmol/L), low (2–21 mmol/L), and moderate/high (≥22 mmol/L). A logistic regression model was constructed to test for outcomes and adjusted for the effects of age, gender, location, injury severity score (ISS), maximum AIS head, and BAC level.

mpared by blood alcohol concentration (BAC) at time of injury: negative (0–1.9 mmol/L), low (2–21 mmol/L), and moderate/high (≥22 mmol/L). A logistic regression model was constructed to test for outcomes and adjusted for the effects of age, gender, location, injury severity score (ISS), maximum AIS head, and BAC level. Results: Overall, 4518 major TBI patients were seen in provincial hospitals during the study period. The mean age of TBI patients was 51 ± 25 years; 72 % were male. The majority of injuries were the result of blunt trauma (93 %), with relatively few major TBIs resulting from penetrating trauma (7 %). The most common mechanisms of injury were falls (46 %) and motor vehicle crashes (27 %). Analysis of census-based subpopulations demonstrated that injury rates varied significantly by geography (from 25 to 65 TBIs per 100,000 population). Testing for alcohol was performed in 43 % of cases. Patients who were tested for alcohol tended to be male (79 %) and middle-aged (mean age 44 ± 20 years). A positive BAC (i.e., ≥2 mmol/L) was found in 47 % of patients who were tested. Mean acute LOS was similar for all three BAC groups. Mortality was independently predicted by increasing age (odds ratio [OR] = 1.01; p < 0.001), male gender (OR = 1.30; p = 0.049), an ISS between 16 and 75 (OR = 2.67; p < 0.001), a maximum AIS head of 5 or 6 (OR = 3.21; p < 0.001), injuries occurring outside of the capital city of Halifax (OR = 1.70; p < 0.001), and having a lower BAC level (OR = 0.99; p < 0.001).

easing age (odds ratio [OR] = 1.01; p < 0.001), male gender (OR = 1.30; p = 0.049), an ISS between 16 and 75 (OR = 2.67; p < 0.001), a maximum AIS head of 5 or 6 (OR = 3.21; p < 0.001), injuries occurring outside of the capital city of Halifax (OR = 1.70; p < 0.001), and having a lower BAC level (OR = 0.99; p < 0.001). Conclusions: The results of this study show significant regional variation in major TBI rates in the province of Nova Scotia. Our findings suggest that low BAC levels are associated with increased mortality in major TBI patients. There are ongoing needs for prevention and intervention efforts that focus on unintentional falls and motor vehicle crashes, especially in older adults. Further study is warranted to elucidate the mechanisms underlying the effects of alcohol on outcomes in patients with major TBI.

eased mortality in major TBI patients. There are ongoing needs for prevention and intervention efforts that focus on unintentional falls and motor vehicle crashes, especially in older adults. Further study is warranted to elucidate the mechanisms underlying the effects of alcohol on outcomes in patients with major TBI. A459 N-acetylcysteine ameliorates liver injury in a rat model of intestinal ischemia-reperfusion P. Briassoulis1, K. Kalimeris1, A. Ntzouvani2, T. Nomikos2, K. Papaparaskeva3, E. Politi4, G. Kostopanagiotou1 1Attikon University Hospital, 2nd Department of Anaesthesiology, Athens, Greece; 2Harokopeio University, Department of Nutrition and Dietetics, School of Health Science and Education, Athens, Greece; 3'Agia Olga' Hospital, Laboratory of Pathology, Athens, Greece; 4Aretaieion University Hospital, Laboratory of Cytology, Athens, Greece Correspondence: P. Briassoulis - Attikon University Hospital, 2nd Department of Anaesthesiology, Athens, Greece Introduction: N-acetylcysteine (NAC) is an antioxidant with direct and indirect antioxidant actions used in the clinical setting. Oxidative stress is known to play a pivotal role in the intestinal ischemia-reperfusion (IIR). Objectives: Therefore we studied the effect of different pretreatment regimens with NAC on the IIR injury in rats.

A459 N-acetylcysteine ameliorates liver injury in a rat model of intestinal ischemia-reperfusion P. Briassoulis1, K. Kalimeris1, A. Ntzouvani2, T. Nomikos2, K. Papaparaskeva3, E. Politi4, G. Kostopanagiotou1 1Attikon University Hospital, 2nd Department of Anaesthesiology, Athens, Greece; 2Harokopeio University, Department of Nutrition and Dietetics, School of Health Science and Education, Athens, Greece; 3'Agia Olga' Hospital, Laboratory of Pathology, Athens, Greece; 4Aretaieion University Hospital, Laboratory of Cytology, Athens, Greece Correspondence: P. Briassoulis - Attikon University Hospital, 2nd Department of Anaesthesiology, Athens, Greece Introduction: N-acetylcysteine (NAC) is an antioxidant with direct and indirect antioxidant actions used in the clinical setting. Oxidative stress is known to play a pivotal role in the intestinal ischemia-reperfusion (IIR). Objectives: Therefore we studied the effect of different pretreatment regimens with NAC on the IIR injury in rats. Methods: Thirty-five male Wistar rats were randomly assigned to 5 groups. In group SHAM only laparotomy was performed. Group CONTROL underwent IIR without NAC. In the other groups, NAC was administered intraperitoneally with different regimens: 150 mg/kg before ischemia (NAC150), 300 mg/kg before ischemia (NAC300) and 150 mg/kg before ischemia plus 150 mg/kg 5 min before reperfusion (NAC150 + 150). Measurements in tissues and blood were conducted at 4 hours of reperfusion following exsanguination.

AC was administered intraperitoneally with different regimens: 150 mg/kg before ischemia (NAC150), 300 mg/kg before ischemia (NAC300) and 150 mg/kg before ischemia plus 150 mg/kg 5 min before reperfusion (NAC150 + 150). Measurements in tissues and blood were conducted at 4 hours of reperfusion following exsanguination. Results: Histological score of the liver was significantly improved in NAC300 compared with control (1.7 ± 0.5 vs. 2.9 ± 1.1 respectively, p = 0.05). In addition, NAC treatment significantly reduced liver transaminases in all groups of treatment, mostly in group NAC300. Plasma malondialdehyde levels were lower with NAC treatment, although not statistically significantly. Lung glutathione peroxidase was significantly increased in group NAC300 (p = 0.04), while the other oxidation biomarkers showed no significant differences. Conclusions: NAC exerts a significant protective role in liver injury following IIR, which seems to be independent of an intestinal protective effect. Additional administration of NAC before reperfusion was of no further benefit. The most effective regimen among the compared regimens was that of 300 mg/kg before ischemia. Keywords: intestinal ischemia-reperfusion, n-acetylcysteine, oxidative stress, liver injury, rat model

Conclusions: NAC exerts a significant protective role in liver injury following IIR, which seems to be independent of an intestinal protective effect. Additional administration of NAC before reperfusion was of no further benefit. The most effective regimen among the compared regimens was that of 300 mg/kg before ischemia. Keywords: intestinal ischemia-reperfusion, n-acetylcysteine, oxidative stress, liver injury, rat model Oral Sessions: Tuesday, 4 October 2016 Abstract award winning session A460 Pre-hospital emergency anaesthesia in awake hypotensive trauma patients: life saving or detrimental? K. Crewdson, M. Rehn, A. Weaver, K. Brohi, D. Lockey London's Air Ambulance, Pre-hospital Emergency Medicine, London, UK Correspondence: K. Crewdson - London's Air Ambulance, Pre-hospital Emergency Medicine, London, UK Introduction: The benefits of pre-hospital emergency anaesthesia (PHEA) are controversial. The timing of this intervention can be a difficult decision, particularly for conscious trauma patients with significant hypotension secondary to presumed hypovolaemia. Patients who are hypovolaemic prior to induction of anaesthesia are at risk of severe cardiovascular instability post induction, which may be difficult to manage. Objectives: The aim of this study was to compare the mortality rate for severely injured hypovolaemic trauma patients (without evidence of major neurological injury) undergoing PHEA with a patient cohort with the same physiology who were transported to hospital without PHEA.

Oral Sessions: Tuesday, 4 October 2016 Abstract award winning session A460 Pre-hospital emergency anaesthesia in awake hypotensive trauma patients: life saving or detrimental? K. Crewdson, M. Rehn, A. Weaver, K. Brohi, D. Lockey London's Air Ambulance, Pre-hospital Emergency Medicine, London, UK Correspondence: K. Crewdson - London's Air Ambulance, Pre-hospital Emergency Medicine, London, UK Introduction: The benefits of pre-hospital emergency anaesthesia (PHEA) are controversial. The timing of this intervention can be a difficult decision, particularly for conscious trauma patients with significant hypotension secondary to presumed hypovolaemia. Patients who are hypovolaemic prior to induction of anaesthesia are at risk of severe cardiovascular instability post induction, which may be difficult to manage. Objectives: The aim of this study was to compare the mortality rate for severely injured hypovolaemic trauma patients (without evidence of major neurological injury) undergoing PHEA with a patient cohort with the same physiology who were transported to hospital without PHEA. Methods: A five year retrospective database review of a physician-led pre-hospital trauma service was performed to identify trauma patients who were hypotensive (systolic blood pressure < 90 mmHg) on scene with a GCS between 13 and 15 between 01/09/2009 and 31/08/2014. Pre-hospital records were consulted independently by two pre-hospital clinicians to determine the likelihood of hypovolaemia based on clinical factors.

performed to identify trauma patients who were hypotensive (systolic blood pressure < 90 mmHg) on scene with a GCS between 13 and 15 between 01/09/2009 and 31/08/2014. Pre-hospital records were consulted independently by two pre-hospital clinicians to determine the likelihood of hypovolaemia based on clinical factors. The primary outcome measure was mortality, defined as death before hospital discharge. The secondary outcome measure was mortality for patients hypotensive on scene secondary to presumed hypovolaemia.

performed to identify trauma patients who were hypotensive (systolic blood pressure < 90 mmHg) on scene with a GCS between 13 and 15 between 01/09/2009 and 31/08/2014. Pre-hospital records were consulted independently by two pre-hospital clinicians to determine the likelihood of hypovolaemia based on clinical factors. The primary outcome measure was mortality, defined as death before hospital discharge. The secondary outcome measure was mortality for patients hypotensive on scene secondary to presumed hypovolaemia. Results: In total, 9480 patients were attended in the study period; 236 patients were included in the final analysis. Of these, 101 patients received PHEA on scene and 135 did not. Fifteen patients who underwent PHEA on scene died (14.9 %) compared with six patients (4.4 %) in the group of patients who were not intubated on scene, p = 0.01. The unadjusted odds ratio for death using the exact logistic regression model was 3.763 (1.30-12.21) p = 0.01. This increased to 5.96 (95 % CI 1.76-20.17) p = 0.004 after adjustment for age, mechanism of injury and initial heart rate. Subgroup analysis identified a statistically significant increase in mortality for hypovolaemic trauma patients who underwent PHEA. In the group of 101 patients anaesthetised on scene, 58 patients (57.4 %) were considered to be hypovolaemic prior to induction of anaesthesia, 14 died (24 %). In the group of 43 patients (42.6 %) intubated on scene not meeting criteria for hypovolaemia, one patient died (2 %), p = 0.003. After adjustment for age, mechanism of injury and initial heart rate the odds ratio for mortality was 9.99 (95 % CI 1.69-58.98) p = 0.01.

c prior to induction of anaesthesia, 14 died (24 %). In the group of 43 patients (42.6 %) intubated on scene not meeting criteria for hypovolaemia, one patient died (2 %), p = 0.003. After adjustment for age, mechanism of injury and initial heart rate the odds ratio for mortality was 9.99 (95 % CI 1.69-58.98) p = 0.01. Conclusions: Our results show an association between PHEA and mortality in awake hypotensive trauma patients, which is increased when hypotension is confirmed to be due to hypovolaemia. This study supports the hypothesis that where patients are hypovolaemic and awake on scene it might, where possible, be appropriate to delay induction of anaesthesia until after hospital arrival.

mortality in awake hypotensive trauma patients, which is increased when hypotension is confirmed to be due to hypovolaemia. This study supports the hypothesis that where patients are hypovolaemic and awake on scene it might, where possible, be appropriate to delay induction of anaesthesia until after hospital arrival. A461 The extra physiotherapy in critical care (EPICC) multi-centre randomised controlled trial S. Wright1, K. Thomas1, C. Baker1, L. Mansfield1, V. Stafford1, C. Wade1, G. Watson2, A. Bryant3, T. Chadwick3, J. Shen3, J. Wilkinson2, J. Furneval4, A. Henderson4, K. Hugill5, P. Howard5, A. Roy4, S. Bonner5, S. Baudouin1 1Newcastle upon Tyne Hospitals NHS Foundation Trust, Perioperative and Critical Care, Newcastle upon Tyne, UK; 2Newcastle University, Clinical Trials Unit, Newcastle upon Tyne, UK; 3Newcastle University, Institute of Health & Society, Newcastle upon Tyne, UK; 4City Hospitals Sunderland NHS Foundation Trust, Anaesthesia, Sunderland, UK; 5South Tees Hospitals NHS Foundation Trust, Anaesthesia, Middlesborough, UK Correspondence: S. Baudouin - Newcastle upon Tyne Hospitals NHS Foundation Trust, Perioperative and Critical Care, Newcastle upon Tyne, UK Introduction: Early Mobilisation Therapy (EMT) may improve outcome in the critically ill but the evidence is very limited 1. We have conducted a multi-centre randomised controlled trial (RCT) comparing two levels of EMT. Objectives: To test the hypothesis that more intensive EMT will improve longer term physical outcome in the critically ill.

A461 The extra physiotherapy in critical care (EPICC) multi-centre randomised controlled trial S. Wright1, K. Thomas1, C. Baker1, L. Mansfield1, V. Stafford1, C. Wade1, G. Watson2, A. Bryant3, T. Chadwick3, J. Shen3, J. Wilkinson2, J. Furneval4, A. Henderson4, K. Hugill5, P. Howard5, A. Roy4, S. Bonner5, S. Baudouin1 1Newcastle upon Tyne Hospitals NHS Foundation Trust, Perioperative and Critical Care, Newcastle upon Tyne, UK; 2Newcastle University, Clinical Trials Unit, Newcastle upon Tyne, UK; 3Newcastle University, Institute of Health & Society, Newcastle upon Tyne, UK; 4City Hospitals Sunderland NHS Foundation Trust, Anaesthesia, Sunderland, UK; 5South Tees Hospitals NHS Foundation Trust, Anaesthesia, Middlesborough, UK Correspondence: S. Baudouin - Newcastle upon Tyne Hospitals NHS Foundation Trust, Perioperative and Critical Care, Newcastle upon Tyne, UK Introduction: Early Mobilisation Therapy (EMT) may improve outcome in the critically ill but the evidence is very limited 1. We have conducted a multi-centre randomised controlled trial (RCT) comparing two levels of EMT. Objectives: To test the hypothesis that more intensive EMT will improve longer term physical outcome in the critically ill. Methods: A full trial protocol has been published2. We conducted a 3 centre RCT between January 2012 to July 2015. Eligible patients had received at least 48 hours of invasive ventilation before recruitment. The Intervention arm received up to 90 min per day of EMT Monday to Friday whilst the control arm received up to 30 min per day Monday to Friday. The primary trial outcome was the physical component of the SF-36 quality of life measure at 6 months after recruitment.

48 hours of invasive ventilation before recruitment. The Intervention arm received up to 90 min per day of EMT Monday to Friday whilst the control arm received up to 30 min per day Monday to Friday. The primary trial outcome was the physical component of the SF-36 quality of life measure at 6 months after recruitment. Results: 308 patients were recruited (150 assigned to the intervention arm and 158 to control). The groups were well matched in terms of age, gender, pre-morbid function, illness severity, type of admission and BMI. Between enrollment and final follow-up at 6 months 94 (30.5 %) participants died. Of the remaining 214 participants 116 completed and returned the SF-36 at 6 months. There was no significant difference in baseline demographics between survivors who completed the SF-36 and those who failed to respond. Full survival data were available for all participants via a national database. The total number of EMT session delivered to the Intervention group was higher than the standard group (2061 v 1326) and the time delivered was correspondingly higher (456.2 v 255.8 hours in total). Physical health as measured by the SF-36 Physical Component score improved in both groups over the 6 month follow up period with no significant difference at 6 months (37.0 Intervention v 37.1 Control; difference in means and 95 % CI 0.3 [−6.1 to 6.7]). There was no significant difference in any of the secondary measures or in 6 month mortality (Intervention 28.7 % standard 34.4 % Cox proportional hazard ratio 1:00 [95 % CI 0.60-1.68]).

iod with no significant difference at 6 months (37.0 Intervention v 37.1 Control; difference in means and 95 % CI 0.3 [−6.1 to 6.7]). There was no significant difference in any of the secondary measures or in 6 month mortality (Intervention 28.7 % standard 34.4 % Cox proportional hazard ratio 1:00 [95 % CI 0.60-1.68]). Conclusions: Intensive EMT did not improve the recovery or outcome at 6 months of critically ill patients ventilated for at least 48 hours compared to standard EMT. References 1. Hashem MD et al. Early Mobilization and Rehabilitation of the Critically Ill Patient. Chest 2016: S0012-3692(16)41641-7 2. Thomas K et al. Extra Physiotherapy in Critical Care (EPICC) Trial Protocol: a randomised controlled trial of intensive versus standard physical rehabilitation therapy in the critically ill. BMJ Open 2015;5(5):e008035. Grant acknowledgement This abstract presents independent research funded by the National Institute for Health Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference Number PB-PG-0909-20021). The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.

abstract presents independent research funded by the National Institute for Health Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference Number PB-PG-0909-20021). The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. A462 Colonization, nosocomial infection and antibiotic consumption after four years application of selective digestive decontamination in an intensive care unit in a university hospital C. Sánchez Ramírez1, S. Hípola Escalada1, M.A. Hernández Viera1, M. Cabrera Santana1, L. Caipe Balcázar1, N. Sangil Monroy2, F. Artiles Campelo3, C.F. Lübbe Vázquez1, P. Saavedra Santana4, S. Ruiz Santana1 1University Hospital of Gran Canaria Dr. Negrín, Intensive Care Unit, Las Palmas de Gran Canaria, Spain; 2University Hospital of Gran Canaria Dr. Negrín, Pharmacy Department, Las Palmas de Gran Canaria, Spain; 3University Hospital of Gran Canaria Dr. Negrín, Microbiology Department, Las Palmas de Gran Canaria, Spain; 4University of Las Palmas de Gran Canaria, Mathematics and Informatics Deparment, Las Palmas de Gran Canaria, Spain Correspondence: C. Sánchez Ramírez - University Hospital of Gran Canaria Dr. Negrín, Intensive Care Unit, Las Palmas de Gran Canaria, Spain Objectives: To prospectively evaluate the impact of Selective Digestive Decontamination (SDD) application on nosocomial infections and colonization rates after 4 years in an ICU.

n Correspondence: C. Sánchez Ramírez - University Hospital of Gran Canaria Dr. Negrín, Intensive Care Unit, Las Palmas de Gran Canaria, Spain Objectives: To prospectively evaluate the impact of Selective Digestive Decontamination (SDD) application on nosocomial infections and colonization rates after 4 years in an ICU. Methods: This study was conducted in a 30-bed-medical-surgical ICU. All consecutive patients admitted to the ICU from October 1, 2011 to September 30, 2015 expected to require tracheal intubation > 48 hours were given SDD (SDD study group) with a 4-day course of intravenous cefotaxime, plus enteral colistin, tobramycin, nystatin in an oropharyngeal paste and in a digestive solution. Oropharyngeal and rectal swabs were obtained on admission and once weekly. We used ENVIN nosocomial infection criteria. We compared all patients admitted to ICU who acquired nosocomial ICU infections from October 1, 2010 to September 30, 2011 (non-SDD group) to the SDD study group. In both groups, categorical variables were summarized as frequencies and percentages and the continuous ones as means and standard deviations (SD) when the data followed the normal distribution or medians and interquartile ranges (IQR) when they did not. The percentages were compared using the test of chi-square test or Fisher exact test, means with the t-test and medians with the Wilcoxon test for independent samples. Those variables that showed statistical significance in the univariate analysis were introduced in a multivariate logistic regression analysis. For each one of the acquired infections (catheter-related and other secondary bacteremias, pneumonia and urinary infections and antibiotic resistant bacteria (ARB) infections) the incidences per 1000 days of exposure in each cohort and the corresponding relative risks were obtained using the Poisson regression. Statistical significance was set at p ≤ 0.05. We analyzed colistin and tobramycin resistant colonization and also antibiotic consumption (Defined antibiotics Daily Doses (DDD).

tions) the incidences per 1000 days of exposure in each cohort and the corresponding relative risks were obtained using the Poisson regression. Statistical significance was set at p ≤ 0.05. We analyzed colistin and tobramycin resistant colonization and also antibiotic consumption (Defined antibiotics Daily Doses (DDD). Results: Results are shown in Figs. 18, 19 and 20. There were no statistical significant differences between both groups in type of ICU admission or demographic data. Patients with SDD had significantly less Extended Spectrum Betalactamase (ESBL), Gram Negative Bacteria Multirresistant (GNB MR) and Acinetobacter spp infections. We had also a significant reduction in nosocomial pneumonias and other secondary bacteremias and ARB rates in SDD group versus non SDD. There was no infection by Clostridium difficile. The exogenous infections were 76,08 %. Colistin resistant colonization was 14,1 % and tobramycin resistant colonization was 20,7 % of samples. There was a decrease on the DDD/100 ICU stays after SDD. Conclusions: After 4 years applying SDD a significant reduction of infections by ESBL, GNB MR and Acinetobacter was observed. A significant decrease of nosocomial pneumonia, secondary bacteremias and ARB infections rates was also shown. An antibiotic consumption reduction was found after SDD. Low rates of colistin and tobramycin resistant colonization bacteria have been observed.Fig. 18 (abstract A462). Univariate analysis Fig. 19 (abstract A462). Multivariate logistic regression analysis Fig. 20 (abstract A462). Nosocomial infection rates

Conclusions: After 4 years applying SDD a significant reduction of infections by ESBL, GNB MR and Acinetobacter was observed. A significant decrease of nosocomial pneumonia, secondary bacteremias and ARB infections rates was also shown. An antibiotic consumption reduction was found after SDD. Low rates of colistin and tobramycin resistant colonization bacteria have been observed.Fig. 18 (abstract A462). Univariate analysis Fig. 19 (abstract A462). Multivariate logistic regression analysis Fig. 20 (abstract A462). Nosocomial infection rates A463 Hypertonic lactate improves cerebral energy metabolism and brain perfusion in critically ill patients with aneurysmal subarachnoid hemorrhage L. Carteron1, C. Patet1, H. Quintard2, D. Solari1, P. Bouzat3, M. Oddo1 1CHUV, Department of Intensive Care Medicine, Neuroscience Critical Care Research Group, Lausanne, Switzerland; 2University Hospital, Department of Anesthesia and Intensive Care Medicine, Nice, France; 3University Hospital, Department of Anesthesia and Intensive Care Medicine, Grenoble, France Correspondence: L. Carteron - CHUV, Department of Intensive Care Medicine, Neuroscience Critical Care Research Group, Lausanne, Switzerland Introduction: Lactate supports brain energy metabolism and promotes cerebral vasodilation. Hypertonic lactate (HL) solutions are emerging as a therapeutic alternative after acute brain injury however their effect following aneurysmal subarachnoid hemorrhage (SAH) is unknown.

l Care Research Group, Lausanne, Switzerland Introduction: Lactate supports brain energy metabolism and promotes cerebral vasodilation. Hypertonic lactate (HL) solutions are emerging as a therapeutic alternative after acute brain injury however their effect following aneurysmal subarachnoid hemorrhage (SAH) is unknown. Methods: We prospectively studied critically ill SAH patients (n = 7, age 64 ± 6 years) monitored with cerebral microdialysis (CMD) and transcranial Doppler (TCD) who were resuscitated according to current guidelines and had no intracranial hypertension (clinicaltrial.gov NCT01573507). Intervention consisted of a 3-h infusion of HL (30 μmol/kg/min, administered within 48 h from SAH, to achieve blood arterial lactate ≈ 4–5 mmol/L). Endpoints were the effect of HL on cerebral energy metabolism (using hourly CMD concentrations of lactate, pyruvate and glucose) and brain perfusion (using hourly TCD measurement of middle cerebral artery cerebral blood flow velocities, MCBFV). Results: Treatment with HL was associated with a significant increase of cerebral extracellular lactate (3.6 ± 1.6 vs. 5.5 ± 1.5 mmol/L), pyruvate (118.5 ± 51.1 vs. 166.4 ± 47.5 μmol/L) and glucose (1.1 ± 0.2 vs. 1.6 ± 0.3 mmol/L; all p < 0.01, paired t test for comparisons between baseline and peak values during HL). HL therapy was also associated with improved brain perfusion (MCBFV 61.8 ± 28.5 vs. 79.7 ± 35.3 cm/sec; p < 0.05), while cerebral perfusion pressure (79 ± 16 vs. 83 ± 13 mmHg) and PbtO2 (20 ± 7 vs. 20 ± 7 mmHg) did not change significantly.

.01, paired t test for comparisons between baseline and peak values during HL). HL therapy was also associated with improved brain perfusion (MCBFV 61.8 ± 28.5 vs. 79.7 ± 35.3 cm/sec; p < 0.05), while cerebral perfusion pressure (79 ± 16 vs. 83 ± 13 mmHg) and PbtO2 (20 ± 7 vs. 20 ± 7 mmHg) did not change significantly. Conclusions: Our findings show that HL administered during the early post-injury phase following SAH improves brain energy metabolism by exerting a beneficial cerebral glucose sparing effect. HL also increases brain perfusion, which seems independent from cerebral perfusion pressure and PbtO2, and appears related to cerebral extracellular lactate supplementation. These preliminary data support larger studies to examine the value of HL solutions in patients with SAH and other forms of acute cerebrovascular diseases. Grants Supported by research grants from the Swiss National Science Foundation (SNSF), the Novartis Foundation for Biomedical Research, the Société Française d'Anesthésie et de Réanimation (SFAR) and the “Fondation des Gueules Cassées”.

Conclusions: Our findings show that HL administered during the early post-injury phase following SAH improves brain energy metabolism by exerting a beneficial cerebral glucose sparing effect. HL also increases brain perfusion, which seems independent from cerebral perfusion pressure and PbtO2, and appears related to cerebral extracellular lactate supplementation. These preliminary data support larger studies to examine the value of HL solutions in patients with SAH and other forms of acute cerebrovascular diseases. Grants Supported by research grants from the Swiss National Science Foundation (SNSF), the Novartis Foundation for Biomedical Research, the Société Française d'Anesthésie et de Réanimation (SFAR) and the “Fondation des Gueules Cassées”. A464 Randomized controlled trial using daily protocol based physiotherapy or protocol based physiotherapy with additional electrical muscle stimulation (EMS) in critically ill patients to prevent intensive care unit (ICU) acquired weakness (ICUAW) T. Wollersheim1,2, J. Malleike1, K. Haas1, N. Carbon1, J. Schneider1,2, C. Birchmeier3, J. Fielitz4,5, S. Spuler1,4, S. Weber-Carstens1,2 1Charité - Universitaetsmedizin Berlin, Berlin, Germany; 2Berlin Institute of Health (BIH), Berlin, Germany; 3Max Delbrück Center for Molecular Medicine (MDC), Berlin, Germany; 4ECRC - Experimental and Clinical Research Center, Berlin, Germany; 5Immanuel Hospital Bernau, Bernau, Germany Correspondence: T. Wollersheim - Charité - Universitaetsmedizin Berlin, Berlin, Germany Introduction: Intensive care unit acquired weakness (ICUAW) is a severe complication of critical illness. ICUAW is associated with muscle wasting, muscle weakness, respiratory failure as well as increased morbidity and mortality and features massive skeletal muscle atrophy by destruction of the contractile myosin elements. No specific treatment has been established.

ess (ICUAW) is a severe complication of critical illness. ICUAW is associated with muscle wasting, muscle weakness, respiratory failure as well as increased morbidity and mortality and features massive skeletal muscle atrophy by destruction of the contractile myosin elements. No specific treatment has been established. Objectives: Aim of the study was to investigate if protocol-based physiotherapy (pPT) or pPT with additional evoked muscle contraction by electrical muscle stimulation (EMS) are more effective to prevent skeletal muscle atrophy in critically ill patients when compared to standard physiotherapy (sPT). Methods: A randomized clinical trial was performed. Participants (inclusion criterion: SOFA score ≥ 9) were randomly split into 2 groups, either treated started at the day of admission with sole pPT or pPT with EMS daily. Data of patients treated with standard physiotherapy (sPT) were available from an earlier trial[1]. An open surgical muscle biopsy from the musculus vastus lateralis was obtained at median day 15 of critical illness. Histological examination of muscle biopsies and quantification of myocyte cross-sectional area (MCSA) for type I, IIa, and IIb fibers using ImageJ software (fiber count > 100 per patient) were performed after applying the toluidine blue ATPase Method: Non-parametric tests were performed. Ethic vote (Charité EA 2/041/10).

stological examination of muscle biopsies and quantification of myocyte cross-sectional area (MCSA) for type I, IIa, and IIb fibers using ImageJ software (fiber count > 100 per patient) were performed after applying the toluidine blue ATPase Method: Non-parametric tests were performed. Ethic vote (Charité EA 2/041/10). Results: 39 critical ill patients were enrolled in this final analysis (19sPT, 11pPT, 9 pPT + EMS). There were no significant differences in baseline characteristics and fiber type distribution between the groups. The MCSA for all fiber types of patients with pPT with or without additional EMS were larger than the MCSA of patients with sPT. Compared to pPT alone, additional evoked muscle activation by EMS increases MCSA for type I,IIa and IIb muscle fibers. Data are shown as frequency-distribution histograms. Conclusions: For the first time we show that pPT is associated with reduced skeletal muscle atrophy when compared to sPT. EMS resulted in an additional reduction of muscle atrophy. The histological findings showed reduced muscle atrophy through preventive physiotherapy procedures. Further studies are needed to solidify our findings and to analyze if this reduction in atrophic response can be confirmed on molecular (i.e. analysis of myosin heavy chain, protein homeostasis) and functional levels and improves the long-term outcome of critical ill patients. References 1. Wollersheim et al. (2014) Dynamics of myosin degradation in intensive care unit-acquired weakness during severe critical illness. Intensive Care Med. GRANT ACKNOWLEDGEMENT

Conclusions: For the first time we show that pPT is associated with reduced skeletal muscle atrophy when compared to sPT. EMS resulted in an additional reduction of muscle atrophy. The histological findings showed reduced muscle atrophy through preventive physiotherapy procedures. Further studies are needed to solidify our findings and to analyze if this reduction in atrophic response can be confirmed on molecular (i.e. analysis of myosin heavy chain, protein homeostasis) and functional levels and improves the long-term outcome of critical ill patients. References 1. Wollersheim et al. (2014) Dynamics of myosin degradation in intensive care unit-acquired weakness during severe critical illness. Intensive Care Med. GRANT ACKNOWLEDGEMENT Berlin Institute of Health (BIH), TRG3 and Clinical Scientist, DFG KFO 192/2 TP3.Fig. 21 (abstract A464). [MCSA]

Conclusions: For the first time we show that pPT is associated with reduced skeletal muscle atrophy when compared to sPT. EMS resulted in an additional reduction of muscle atrophy. The histological findings showed reduced muscle atrophy through preventive physiotherapy procedures. Further studies are needed to solidify our findings and to analyze if this reduction in atrophic response can be confirmed on molecular (i.e. analysis of myosin heavy chain, protein homeostasis) and functional levels and improves the long-term outcome of critical ill patients. References 1. Wollersheim et al. (2014) Dynamics of myosin degradation in intensive care unit-acquired weakness during severe critical illness. Intensive Care Med. GRANT ACKNOWLEDGEMENT Berlin Institute of Health (BIH), TRG3 and Clinical Scientist, DFG KFO 192/2 TP3.Fig. 21 (abstract A464). [MCSA] Haemodynamic monitoring A465 Central venous-to-arterial carbon dioxide difference and thehaldane effect: a limiting factor, or an additional marker of severity in shock? L. Enseñat1, A. Pérez-Madrigal1, P. Saludes1, L. Proença1,2, G. Gruartmoner1, C. Espinal1, J. Mesquida1 1Corporació Sanitària i Universitaria Parc Taulí, Universitat Autònoma de Barcelona, Critical Care Department, Sabadell, Spain; 2Hospital Prof. Dr. Fernando Fonseca, Serviço de Medicina I, Amadora, Portugal Correspondence: L. Enseñat - Corporació Sanitària i Universitaria Parc Taulí, Universitat Autònoma de Barcelona, Critical Care Department, Sabadell, Spain Objective: Central venous-to-arterial carbon dioxide difference (PcvaCO2) has demonstrated its prognostic value in critically ill patients suffering from shock, and current expert recommendations advocate for further resuscitation interventions when PcvaCO2 is elevated. PcvaCO2combination with arterial-venous oxygen content difference (PcvaCO2/CavO2) seems to enhance its performance when assessing anaerobic metabolism. However, the fact that PCO2 values might be altered by changes in blood O2content (the Haldane effect), has been presented as a limitation of PCO2-derived variables. The present study aimed at exploring the impact of the Haldane effect on PcvaCO2 and PcvaCO2/CavO2 during the early phase of shock.

anaerobic metabolism. However, the fact that PCO2 values might be altered by changes in blood O2content (the Haldane effect), has been presented as a limitation of PCO2-derived variables. The present study aimed at exploring the impact of the Haldane effect on PcvaCO2 and PcvaCO2/CavO2 during the early phase of shock. Methods: Prospective interventional study. Ventilated patients suffering from shock within the first 24 hours of ICU admission. Patients requiring FiO2 ≥ 0.5 were excluded. At inclusion, simultaneous arterial and central venous blood samples were collected. Patients underwent ahyperoxygenation test (5 minutes of FiO2 100 %), and arterial and central venous blood samples were repeated. Oxygenation and CO2 variables were calculated at both timepoints. Results: Twenty patients were studied. The main cause of shock was septic shock (70 %). The hyperoxygenation trial increased oxygenation parameters in arterial and venous blood, whereas PCO2 only changed at the venous site. Resulting PcvaCO2 and PcvaCO2/CavO2significantly increased (6.8 (4.9, 8.1) vs 7.6 (6.7, 8.5) mmHg, p 0.001; and 1.9 (1.4, 2.2) vs 2.3 (1.8, 3), p < 0.001, respectively). Baseline PcvaCO2, PcvaCO2/CavO2 and ScvO2 correlated with the magnitude of PO2 augmentation at the venous site within the trial(rho −0.46, p 0.04; rho 0.6, p < 0.01; andrho 0.7, p < 0.001, respectively). Increased PcvaCO2/CavO2values were associated with higher mortality in our sample (1.46 (1.21, 1.89) survivorsvs2.23 (1.86, 2.8) non-survivors, p < 0.01).

vO2 correlated with the magnitude of PO2 augmentation at the venous site within the trial(rho −0.46, p 0.04; rho 0.6, p < 0.01; andrho 0.7, p < 0.001, respectively). Increased PcvaCO2/CavO2values were associated with higher mortality in our sample (1.46 (1.21, 1.89) survivorsvs2.23 (1.86, 2.8) non-survivors, p < 0.01). Conclusions: PcvaCO2 and PcvaCO2/CavO2 are influenced by oxygenation changes not related to flow. Elevated PcvaCO2 and PcvaCO2/CavO2 values might not only derive from cardiac output inadequacy, but also from venous hyperoxia. Elevated PcvaCO2/CavO2 values were associated with higher PO2 transmission to the venous compartment, suggesting higher shunting phenomena. Rather than a limitation, the contribution of the Haldane effect to CO2-derived variables might represent an additional marker of severity of shock. References 1.- Mesquida J, et al. Central venous-to-arterial carbon dioxide difference combined with arterial-to-venous oxygen content difference is associated with lactate evolution in the hemodynamic resuscitation process in early septic shock. Crit Care 2015; 19:126 2.- Mekontso-Dessap et al. Combination of venoarterial PCO2 difference with arteriovenous O2 content difference to detect anaerobic metabolism in patients. Intensive Care Med 2002;28:272–277

1.- Mesquida J, et al. Central venous-to-arterial carbon dioxide difference combined with arterial-to-venous oxygen content difference is associated with lactate evolution in the hemodynamic resuscitation process in early septic shock. Crit Care 2015; 19:126 2.- Mekontso-Dessap et al. Combination of venoarterial PCO2 difference with arteriovenous O2 content difference to detect anaerobic metabolism in patients. Intensive Care Med 2002;28:272–277 A466 Pulmonary vascular permeability index and global ejection fraction: are the data appropriately corrected in case of femoral indicator injection for transpulmonary thermodilution? W. Huber, M. Eckmann, F. Elkmann, A. Gruber, T. Lahmer, U. Mayr, A. Herner, R. Schellnegger, J. Schneider, R.M. Schmid Technische Universität München, II. Medizinische Klinik, Munich, Germany Correspondence: W. Huber - Technische Universität München, II. Medizinische Klinik, Munich, Germany Introduction: Transpulmonary thermodilution (TPTD) has been established to measure cardiac output CO, global end-diastolic volume GEDV and extravascular lung water EVLW. To facilitate interpretation of these data several ratios have been developed, including pulmonary vascular permeability index (defined as EVLW/(0.25*GEDV)) and global ejection fraction ((4*stroke volume)/GEDV). PVPI and GEF have been associated to the aetiology of pulmonary oedema and to systolic cardiac function, respectively. However, several studies showed that the use of femoral venous access results in a marked overestimation of GEDV which also falsely reduces PVPI and GEF. One of these studies suggested a correction formula for femoral venous access that markedly reduces the bias for GEDV (1). Consequently, the last PiCCO-algorithm requires information about the CVC, and correction for femoral access has been shown. However, two recent studies demonstrated inconsistencies of this algorithm using incorrected GEDV for PVPI (2), but corrected GEDV for GEF. Nevertheless, these studies were based on mathematical analysis of data displayed by the PiCCO in a total of 15 patients equipped with only a femoral, but not with a jugular CVC.

. However, two recent studies demonstrated inconsistencies of this algorithm using incorrected GEDV for PVPI (2), but corrected GEDV for GEF. Nevertheless, these studies were based on mathematical analysis of data displayed by the PiCCO in a total of 15 patients equipped with only a femoral, but not with a jugular CVC. Objectives: To compare PVPI_fem and GEF_fem derived from femoral TPTD to values derived from jugular indicator injection in 25 patients with both jugular and femoral CVCs. Methods: 54 datasets in 25 patients CVC were recorded. Each dataset consisted of three triplicate TPTDs using the jugular venous access as the gold standard, while the the femoral access with (PVPI_fem_cor) and without (PVPI_fem_uncor) information about the femoral injection site was used to evaluate, if correction for femoral GEDV pertains to PVPI_fem and GEF_fem.

corded. Each dataset consisted of three triplicate TPTDs using the jugular venous access as the gold standard, while the the femoral access with (PVPI_fem_cor) and without (PVPI_fem_uncor) information about the femoral injection site was used to evaluate, if correction for femoral GEDV pertains to PVPI_fem and GEF_fem. Results: 15 male, 10 female patients; 60 ± 15 years; 79 ± 16 kg; 173 ± 8 cm. PVPI_fem_uncor was significantly lower than PVPI_jug (1.48 ± 0.47 vs. 1.84 ± 0.53; p < 0.001). Similarly, PVPI_fem_cor was significantly lower than PVPI_jug (1.49 ± 0.46 vs. 1.84 ± 0.53; p < 0.001). This is explained by the finding that PVPI_fem_uncor was not different to PVPI_fem_cor (1.48 ± 0.47 vs. 1.49 ± 0.46; n.s.). This clearly suggests that correction for femoral CVC does not pertain to PVPI. GEF_fem_uncor was significantly lower than GEF_jug (20.6 ± 5.1 % vs. 25.0 ± 6.1 %; p < 0.001). By contrast, GEF_fem_cor was not different to GEF_jug (25.6 ± 5.8 % vs. 25.0 ± 6.1 %; n.s.). Furthermore, GEF_fem_cor was significantly higher than GEF_fem_uncor (25.6 ± 5.8 % vs. 20.6 ± 5.1 %; p < 0.001). This emphasizes that an appropriate correction for femoral CVC is applied to GEF_fem_cor. Conclusions: Uncorrected femoral injection for TPTD results in significantly lower values for PVPI and GEF. While the last PiCCO algorithm appropriately corrects for GEF, the correction is not applied to PVPI. This results in substantial underestimation of PVPI. References 1. Saugel, Huber et al., Crit Care 2010; 14(3):R95; 2. Berbara, Huber et al. BMC Anesthesiol. 2014

Conclusions: Uncorrected femoral injection for TPTD results in significantly lower values for PVPI and GEF. While the last PiCCO algorithm appropriately corrects for GEF, the correction is not applied to PVPI. This results in substantial underestimation of PVPI. References 1. Saugel, Huber et al., Crit Care 2010; 14(3):R95; 2. Berbara, Huber et al. BMC Anesthesiol. 2014 A467 Echocardiographic assessment of fluid responsiveness in spontaneously breathing patients with hemodynamic instability: anything new? W. Ayoub1, W. Samy1, A. Esmat2, A. Battah3, S. Mukhtar3 1Cairo University Medical School, Intensive Care Department, Cairo, Egypt; 2Faculty of Medicine, Fayoum University, Intensive Care Department, Fayoum, Egypt; 3Cairo University Medical School, Intensive care Department, Cairo, Egypt Correspondence: W. Ayoub - Cairo University Medical School, Intensive Care Department, Cairo, Egypt Introduction: Prediction of fluid responsiveness in hemodynamically unstable patients with spontaneous breathing activity has been a clinical challenge. It has been best assessed by passive leg raising test. Pre-ejection period, the time from the onset of ventricular depolarization to the beginning of left ventricular ejection, is a systolic time interval found to decrease with greater preload1. The effect of passive leg raising test on the pre-ejection period has not been studied in this context.

passive leg raising test. Pre-ejection period, the time from the onset of ventricular depolarization to the beginning of left ventricular ejection, is a systolic time interval found to decrease with greater preload1. The effect of passive leg raising test on the pre-ejection period has not been studied in this context. Objectives: Our objective was to test whether fluid responsiveness could be predicted by the response of pre-ejection period to passive leg raising test. We also examined whether baseline end expiratory inferior vena cava diameter could predict fluid responsiveness in this category of patients. Methods: Thirty patients with spontaneous breathing activity considered for fluid loading were included. We used transthoracic echocardiography to measure stroke volume, and pre-ejection period before and during passive leg raising test as well as before and after fluid loading (500 ml saline 0.9 % over 15 minutes). An increase in stroke volume of 15 % or more after volume expansion defined fluid responders. We also measured baseline end expiratory inferior vena cava diameter obtained from the subcostal window.

fore and during passive leg raising test as well as before and after fluid loading (500 ml saline 0.9 % over 15 minutes). An increase in stroke volume of 15 % or more after volume expansion defined fluid responders. We also measured baseline end expiratory inferior vena cava diameter obtained from the subcostal window. Results: 19 patients were responders (63.3 %). Passive leg raising test induced-changes in stroke volume of ≥ 9.3 % predicted fluid responsiveness with a sensitivity of 100 % and specificity of 81.8 %, the area under receiver operating characteristic curve (AUC) was 0.96; 95 % confidence interval (CI) [0.91,1.0], meanwhile, passive leg raising test-induced changes in pre-ejection period of ≤ −5.0 % predicted fluid responsiveness with a sensitivity of 94.7 % and a specificity of 45.5 %, the AUC was 0.62; 95 % CI [0.4,0.85]. Baseline inferior vena cava diameter (in cm) failed to identify responders vs. non-responders (1.20 ± 0.37 vs 1.38 ± 0.51 respectively, p = 0.36). Conclusions: In hemodynamically unstable patients with spontaneous breathing activity, passive leg raising test-induced increase in stroke volume of ≥ 9.3 % accurately predicted fluid responsiveness, while passive leg raising test-induced decrease in pre-ejection period of ≤ −5.0 % was sensitive, but not specific in the prediction of fluid responsiveness. Baseline inferior vena cava diameter failed to identify fluid responders. References 1- Weissler AM. Current concepts in cardiology. Systolic-time intervals. N Engl J Med 296: 321–324, 1977

Conclusions: In hemodynamically unstable patients with spontaneous breathing activity, passive leg raising test-induced increase in stroke volume of ≥ 9.3 % accurately predicted fluid responsiveness, while passive leg raising test-induced decrease in pre-ejection period of ≤ −5.0 % was sensitive, but not specific in the prediction of fluid responsiveness. Baseline inferior vena cava diameter failed to identify fluid responders. References 1- Weissler AM. Current concepts in cardiology. Systolic-time intervals. N Engl J Med 296: 321–324, 1977 A468 Measurement of cutaneous blood flow using a laser doppler technique in patients with shock W. Mongkolpun, D. Orbegozo Cortés, C.P.R. Cordeiro, J.-L. Vincent, J. Creteur University Hospital Erasme, Université Libre de Bruxelles, Critical Care, Brussels, Belgium Correspondence: W. Mongkolpun - University Hospital Erasme, Université Libre de Bruxelles, Critical Care, Brussels, Belgium Introduction: Persistent signs of skin hypoperfusion after resuscitation in patients with shock are associated with worse outcomes, but assessment is rather subjective. Skin laser Doppler (SLD) techniques allow skin blood flow (SBF) to be measured more objectively and non-invasively. Objective: To measure SBF at the extremities using a SLD technique in patients with shock and to determine if they are correlated to patients outcomes.

A468 Measurement of cutaneous blood flow using a laser doppler technique in patients with shock W. Mongkolpun, D. Orbegozo Cortés, C.P.R. Cordeiro, J.-L. Vincent, J. Creteur University Hospital Erasme, Université Libre de Bruxelles, Critical Care, Brussels, Belgium Correspondence: W. Mongkolpun - University Hospital Erasme, Université Libre de Bruxelles, Critical Care, Brussels, Belgium Introduction: Persistent signs of skin hypoperfusion after resuscitation in patients with shock are associated with worse outcomes, but assessment is rather subjective. Skin laser Doppler (SLD) techniques allow skin blood flow (SBF) to be measured more objectively and non-invasively. Objective: To measure SBF at the extremities using a SLD technique in patients with shock and to determine if they are correlated to patients outcomes. Method: SBF was evaluated in the fingertip and toe (PU: perfusion units) using a SLD (PeriFlux System 5000, Perimed, Sweden) at basal skin temperature and after warming to 37 °C in 28 consecutive patients admitted with circulatory shock (need for vasoactive agents to maintain MAP > 60 mmHg) on ICU admission and 10 healthy volunteers. Organ dysfunction was assessed using the sequential organ failure assessment (SOFA) score. Statistical analysis was performed using STATA 13.0. The area under the receiver operating curve (AUC) was calculated to identify the predictive performance of SBF. Results are presented as median (p25-75) values.

thy volunteers. Organ dysfunction was assessed using the sequential organ failure assessment (SOFA) score. Statistical analysis was performed using STATA 13.0. The area under the receiver operating curve (AUC) was calculated to identify the predictive performance of SBF. Results are presented as median (p25-75) values. Result: Global ICU mortality was 32 %. Demographic data and SBF are shown in Table 9. The AUCs for prediction of mortality at basal temperature were 0.77 (0.58-0.97) and 0.65 (0.47-0.92) for the fingertip and toe respectively and 0.71 (0.51-0.93) for the SOFA score. Conclusion: Measurements of SLD at fingertip and toe may represent an interesting monitoring technique for shock patients. Their prognostic value is similar to SOFA score.Table 9 (abstract A468). Demograpic data Survivors(n=19) Non survivors (n=9) P All (n=28) Volunteer (n=10) P SOFA score 10(8–11) 12(9–13) 0.06 11(9–12) - - Age (years) 63 (55–57) 58 (41–68) 0.20 62 (54–72) 30(36–31) <0.01 Mean arterial pressure (mmHg) 69 (67–86) 67 (62–84) 0.40 69 (65–85) - - Cardiac index (L/min/m2) 2.7 (2.4–3.2) 2.4 (1.6–4.4) 0.60 2.6 (2.1–4.4) - - Noradrenaline (mcg/Kg/min) 0.16 (0.10–0.26) 0.50 (0.09–0.80) 0.16 0.18(0.10–0.45) - - Lactate (mmol/L) 1.4 (1.1–1.7) 3.8 (1.6–4.9) 0.01 1.6 (1.2–2.8) - - Shock Types •Septic, n 14 7 0.29 21 - - •Cardiogenic,n 5 1 - 6 - - •Haemorrhagic,n 0 1 - 1 - - Table 10 (abstract A468). SBF in Patients and Volunteers

(2.1–4.4) - - Noradrenaline (mcg/Kg/min) 0.16 (0.10–0.26) 0.50 (0.09–0.80) 0.16 0.18(0.10–0.45) - - Lactate (mmol/L) 1.4 (1.1–1.7) 3.8 (1.6–4.9) 0.01 1.6 (1.2–2.8) - - Shock Types •Septic, n 14 7 0.29 21 - - •Cardiogenic,n 5 1 - 6 - - •Haemorrhagic,n 0 1 - 1 - - Table 10 (abstract A468). SBF in Patients and Volunteers Survivors(n=19) Non survivors (n=9) P All (n=28) Volunteer (n=10) P Fingertip SBF at basal temperature (PU) 31.4 (11.1–49.4) 9.8 (7.8–12.3) 0.02 20.5(9.8–38.7) 175.7(196.5–218.0) <0.01 Fingertip SBF at 37°C(PU) 60.9(21.2–70.1) 17.2(13.9–49.0) 0.09 49.7(15.8–70.0) 274.8(201.2–320.1) <0.01 Toe SBF at basal temperature (PU) 32.5(8.7–44.7) 15.5(6.1–21.9) 0.20 16.3(8.2–35.8) 49.0(44.6–55.1) 0.02 Toe SBF at 37°C(PU) 39.2(19.5–72.9) 20.5(9.9–36.1) 0.04 30.7(17.3–52.7) 116.7(99.9–122.2) <0.01

.5–218.0) <0.01 Fingertip SBF at 37°C(PU) 60.9(21.2–70.1) 17.2(13.9–49.0) 0.09 49.7(15.8–70.0) 274.8(201.2–320.1) <0.01 Toe SBF at basal temperature (PU) 32.5(8.7–44.7) 15.5(6.1–21.9) 0.20 16.3(8.2–35.8) 49.0(44.6–55.1) 0.02 Toe SBF at 37°C(PU) 39.2(19.5–72.9) 20.5(9.9–36.1) 0.04 30.7(17.3–52.7) 116.7(99.9–122.2) <0.01 A469 Role of dynamic haemodynamic parameters for predicting fluid responsiveness in daily clinical routine in icu patients- results from a sub-analysis of the ICU-CardioMan trial S. Funcke1, H. Groesdonk2, B. Saugel1, G. Wagenpfeil3, S. Wagenpfeil3, D.A. Reuter1 1University Medical Center Hamburg-Eppendorf, Center of Anaesthesiology and Intensive Care Medicine, Hamburg, Germany; 2University Hospital of Homburg/Saar, Homburg/Saar, Germany; 3Saarland University, Campus Homburg, Homburg/Saar, Germany Correspondence: S. Funcke - University Medical Center Hamburg-Eppendorf, Center of Anaesthesiology and Intensive Care Medicine, Hamburg, Germany Introduction: Investigating the use of dynamic haemodynamic parameters, such as pulse pressure variation (PPV) and stroke volume variation (SVV), for predicting fluid responsiveness prior to volume expansion has led many scientific discussions lately. Also, according to current consensus statements their use is increasingly recommended in suitable patients [1]. Extended hemodynamic monitoring (EHD) beyond measurement of cardiac filling pressures (central venous pressure, pulmonary artery occlusion pressure) is still rare. A recent study revealed that only a minor part of critical ill patients are supervised with specific monitoring to assess fluid responsiveness [2]. It remains unclear if this is due to lacking availability of the specialised devices or the limitations of the use of automated functional parameters of preload.

till rare. A recent study revealed that only a minor part of critical ill patients are supervised with specific monitoring to assess fluid responsiveness [2]. It remains unclear if this is due to lacking availability of the specialised devices or the limitations of the use of automated functional parameters of preload. Objectives: This analysis was designed to assess the availability of monitoring, the clinical suitability of using PPV or SVV, and if and how they are actually used in German, Swiss, and Austrian ICU´s. Methods: This is a sub-analysis of the ICU-CardioMan Trial. For this multicentre, one day cross-sectional study data regarding patients' hemodynamic monitoring was collected from 161 participating ICUs from Germany, Switzerland and Austria by means of a web-based case report form. The ICUs gave detailed information on availability of EHD in general and provided data from 1789 patients concerning monitoring, as well as information on the patients' specific characteristics that have impact on treatment suitability of PPV and SVV to predict fluid responsiveness. Results: EHD was widely available throughout the examined ICUs. Only a minority of patients was reported to be monitored with EHD during the study period (see Table 11). In the group of the patients that needed catecholamines and/or vasopressors (39.2 %) the share of patients monitored with EHD was significantly higher. In this subgroup, 66.8 % of those patients had sinus or pacer rhythm. Of this portion another 30.9 % were mechanically ventilated in a controlled mode.

(see Table 11). In the group of the patients that needed catecholamines and/or vasopressors (39.2 %) the share of patients monitored with EHD was significantly higher. In this subgroup, 66.8 % of those patients had sinus or pacer rhythm. Of this portion another 30.9 % were mechanically ventilated in a controlled mode. Conclusions: Though widely available EHD to predict fluid responsiveness is still not comprehensively implemented in daily clinical routine. Over one third of patients that required catecholamines/ vasopressors fulfilled the criteria for assessment of PPV or SVV. The high share of patients not fulfilling the criteria for using PPV and SVV stresses the need for further development of functional parameters of preload operating independently from the presence of controlled mechanical ventilation. References 1. Cecconi et al. Intensive Care Med. 2014; 40: 1795–815 2. Preau et al. Anaesth Crit Care Pain Med. 2015 Nov 19Table 11 (abstract A469). Implemented extended monitoring Implemented Monitoring Total Vasoactive agents Mechanical ventilation n = 1789 n = 702 n = 874 Invasive pressure monitoring 87.0% 98.6% 96.1% Cardiac output monitoring 12.3% 24.2% 20.8% Cardiac filling pressures (CVP, PAOP) 55.4% 74.4% 69.5% Dynamic parameters (SVV, PPV) 8.7% 16.5% 14.9%

2. Preau et al. Anaesth Crit Care Pain Med. 2015 Nov 19Table 11 (abstract A469). Implemented extended monitoring Implemented Monitoring Total Vasoactive agents Mechanical ventilation n = 1789 n = 702 n = 874 Invasive pressure monitoring 87.0% 98.6% 96.1% Cardiac output monitoring 12.3% 24.2% 20.8% Cardiac filling pressures (CVP, PAOP) 55.4% 74.4% 69.5% Dynamic parameters (SVV, PPV) 8.7% 16.5% 14.9% Mechanical ventilation: clinical studies A470 Reconnection to mechanical ventilation for one hour after a successful spontaneous breathing trial reduces extubation failure and reintubation in critically ill patients: a multicenter randomized controlled trial M.M. Fernandez1, R. Fernandez2, M. Magret3, A. González-Castro4, M.T. Bouza5, M. Ibañez6, C. García7, B. Balerdi8, A. Mas9, V. Arauzo10, J.M. Añón11, F. Ruiz12, J. Ferreres13, R. Tomás14, M. Alabert15, A.I. Tizón16, S. Altaba17, N. Llamas18 1Hospital Universitari Mútua de Terrassa, ICU, Barcelona, Spain; 2Hospital Sant Joan de Dèu-Fundació Althaia Manresa, ICU, Barcelona, Spain; 3Hospital Juan XXIII, Tarragona, Spain; 4Hospital Universitario Marqués de Valdecilla, ICU, Santander, Spain; 5Hospital de A Coruña, ICU, A Coruña, Spain; 6Hospital Verge de la Cinta, ICU, Tortosa, Spain; 7Hospital Universitario de Canarias, Tenerife, Spain; 8Hospital Universitari Politècnic La Fe, Valencia, Spain; 9Consorci Sanitari Integral Moisés Broggi, Barcelona, Spain; 10Consorci Hospitalari de Terrassa, Barcelona, Spain; 11Hospital Virgen de la Luz-SESCAM, Cuenca, Spain; 12Hospital Medico-Quirúrgico de Jaén, Jaén, Spain; 13Hospital Clínico de Valencia, Valencia, Spain; 14Hospital General de Catalunya, Barcelona, Spain; 15Hospital General de Vic, Vic, Spain; 16Complexo Hospitalario Universitario de Ourense, Ourense, Spain; 17Hospital General Universitario de Castellón, Castellón, Spain; 18Hospital Universitario Rafael Méndez de Lorca, Murcia, Spain Correspondence: M.M. Fernandez - Hospital Universitari Mútua de Terrassa, ICU, Barcelona, Spain Introduction: Weaning from mechanical ventilation remains a challenge in intensive care units (ICU); respiratory muscle fatigue may be important in difficult weaning. We evaluated whether reconnection to mechanical ventilation for one hour after the effort of a spontaneous breathing trial could reduce extubation failure in critically ill patients.

chanical ventilation remains a challenge in intensive care units (ICU); respiratory muscle fatigue may be important in difficult weaning. We evaluated whether reconnection to mechanical ventilation for one hour after the effort of a spontaneous breathing trial could reduce extubation failure in critically ill patients. Objectives: reconnection to mechanical ventilation for one hour after the effort of a spontaneous breathing trial could reduce extubation failure in critically ill patients. Methods: In this prospective, randomized, multicenter trial, patients who successfully completed a spontaneous breathing trial were randomized to direct extubation (Control group) or reconnection to the ventilator for a one-hour rest before extubation (Rest group). Statistical analysis included multivariable logistic regression models for extubation failure.

al blood pressure, core temperature (Tc), and ICP and 2) a Tc ≥ 37.5 °C. Exclusion criteria were: 1) hypovolemia, 2) administration of drugs with hemodynamic effects during the study period, 3) administration of antipyretics in the 6 hours before the start of the study, 4) acute heart failure and 5) cerebral vasospasm. Results: 62 pts [male 35; mean age 52,2 (SD ± 14,2) years; median Glasgow Coma Scale (GCS) 6 (IQR 3–7) at ICU admission] were enrolled. Diagnosis at ICU admission were: 33 (53,25 %) subarachnoid hemorrhage (SAH), 16 (25,8 %) traumatic brain injury (TBI), 10 (16,1 %) intracerebral hemorrhage (ICH) and 3 (4,8 %) acute ischemic stroke (AIS). Paracetamol (1 gr intravenous) was administered to 32 pts and Diclofenac Sodium (12,5 mg intramuscular) to 30 pts. At t-0, 34 pts (54,8 %) showed an ICP ≤ 15 and 28 (45,2 %) an ICP > 15 mmHg. No statistically significant changes in ICP were observed in the overall population after antipyretic administration: 14.9 ± 5.1, 14.9 ± 4.6, 14.4 ± 4.5, 15.1 ± 4.3 mmHg at t-0, t-30, t-60 and t-120 respectively (P = 0.6). A statistically significant reduction in ICP (19.5 ± 3.2, 17.6 ± 4.6, 16.4 ± 4.9, 16.0 ± 4.4 mmHg at t-0, t-30, t-60 and t-120 respectively; P < 0.001) was observed in the group of pts with t-0 ICP > 15 mmHg. Conversely, a statistically significant increase in ICP (11.2 ± 2.7, 12.7 ± 3.4, 12.8 ± 3.5, 14.4 ± 4.1 mmHg at t-0, t-30, t-60 and t-120 respectively; P < 0.001) was observed in the group of pts with a t-0 ICP ≤ 15 mmHg. A statistically significant reduction in Tc, mean arterial pressure (MAP) and cerebral perfusion pressure (CPP) was observed in all groups of patients (overall population and pts with ICP > or ≤ 15 mmHg) after antipyretic therapy.

, multicenter trial, patients who successfully completed a spontaneous breathing trial were randomized to direct extubation (Control group) or reconnection to the ventilator for a one-hour rest before extubation (Rest group). Statistical analysis included multivariable logistic regression models for extubation failure. Results: We randomized 243 patients to the Control group and 227 patients to the Rest group. The most common spontaneous breathing trial method was T-tube, and the median time from intubation to spontaneous breathing trial was similar between groups (5.7 days in the Control group and 5.5 days in the Rest group; p = 0.2). Extubation failure within 48 hours after extubation was more common in the Control than in the Rest group (58 (23.9 %) vs. 24 (10.6 %) patients; p < 0.001). Rescue noninvasive ventilation was administered in 47 (57 %) patients, 16 (34 %) of whom required reintubation. Reintubation was more common in the Control than in the Rest group (35 (14 %) vs. 12 (5 %) patients; p < 0.001) mainly due to inability to manage secretions in both groups. Hospital and ICU length of stay and mortality did not differ between groups. Conclusion: One-hour rest after a successful SBT reduced the rates of extubation failure within 48 h after extubation and reintubation. References 1. Tobin MJ, Perez W, Guenther SM, et al. The pattern of breathing during successful and unsuccessful trials of weaning from mechanical ventilation. Am Rev Respir Dis 1986;134:111–8.

Conclusion: One-hour rest after a successful SBT reduced the rates of extubation failure within 48 h after extubation and reintubation. References 1. Tobin MJ, Perez W, Guenther SM, et al. The pattern of breathing during successful and unsuccessful trials of weaning from mechanical ventilation. Am Rev Respir Dis 1986;134:111–8. 2. Ely EW, Baker AM, Dunagan DP, et al. Effect on the duration of mechanical ventilation of identifying patients capable of breathing spontaneously. N Engl J Med 1996;335:1864–9. 3. Epstein S, Ciubotaru RL, Wong JB. Effect of failed extubation on the outcome of mechanical ventilation. Chest 1997;112:186–92. 4. Laghi F, D'Alfonso N, Tobin MJ. Pattern of recovery from diaphragmatic fatigue over 24 hours. J Appl Physiol 1995;79:539–46. 5. Laghi F, Cattapan S, Jubran A, et al. Is weaning failure caused by low-frequency fatigue of the diaphragm? Am J Respir Crit Care Med. 2003;167:120–7. 6. Goligher EC, Fan E, Herridge MS, et al. Evolution of diaphragm thickness during mechanical ventilation. Impact of inspiratory effort.Am J Respir Crit Care Med 2015;192:1080–8. (ClinicalTrials.gov number, NCT01915563)

5. Laghi F, Cattapan S, Jubran A, et al. Is weaning failure caused by low-frequency fatigue of the diaphragm? Am J Respir Crit Care Med. 2003;167:120–7. 6. Goligher EC, Fan E, Herridge MS, et al. Evolution of diaphragm thickness during mechanical ventilation. Impact of inspiratory effort.Am J Respir Crit Care Med 2015;192:1080–8. (ClinicalTrials.gov number, NCT01915563) A471 Driving pressure is a significant predictor of mortality in the Acurasys and Proseva randomized controlled trials in ARDS patients C. Guérin1,2, L. Papazian3, J. Reignier4, L. Ayzac5, A. Loundou3, J.-M. Forel3 1Hospices Civils de Lyon, Lyon, France; 2INSERM 955, IMRB, Créteil, France; 3APHM, CHU Nord, Marseille, France; 4CHU Nantes, Nantes, France; 5C-CLIN Sud-Est, Pierre Bénite, France Correspondence: C. Guérin - Hospices Civils de Lyon, Lyon, France Introduction: Driving pressure (ΔP) across the respiratory system has been shown to strongly predict hospital mortality in ARDS patients, with a threshold in the vicinity of 14–15 cm H2O above which the relative risk of mortality significantly increased (1). We wonder whether this result may be due to the wide range of tidal volume and PEEP used across the trials included (1). Objectives: Therefore, we investigated ΔP in two trials in which lung protective mechanical ventilation was applied to ARDS patients. Our working hypothesis is that ΔP is a risk factor just like compliance (Crs) or Plateau pressure (Pplat) of the respiratory system are.

A471 Driving pressure is a significant predictor of mortality in the Acurasys and Proseva randomized controlled trials in ARDS patients C. Guérin1,2, L. Papazian3, J. Reignier4, L. Ayzac5, A. Loundou3, J.-M. Forel3 1Hospices Civils de Lyon, Lyon, France; 2INSERM 955, IMRB, Créteil, France; 3APHM, CHU Nord, Marseille, France; 4CHU Nantes, Nantes, France; 5C-CLIN Sud-Est, Pierre Bénite, France Correspondence: C. Guérin - Hospices Civils de Lyon, Lyon, France Introduction: Driving pressure (ΔP) across the respiratory system has been shown to strongly predict hospital mortality in ARDS patients, with a threshold in the vicinity of 14–15 cm H2O above which the relative risk of mortality significantly increased (1). We wonder whether this result may be due to the wide range of tidal volume and PEEP used across the trials included (1). Objectives: Therefore, we investigated ΔP in two trials in which lung protective mechanical ventilation was applied to ARDS patients. Our working hypothesis is that ΔP is a risk factor just like compliance (Crs) or Plateau pressure (Pplat) of the respiratory system are. Methods: ARDS patients included in Acurasys (2) and Proseva (3) trials were used. The first trial compared the neuromuscular blocking agent (NMBA) cisatracurium vs. placebo and the second the prone vs. the supine position. Both had near inclusion criteria (notably PaO2/FIO2 < 150 mm Hg and PEEP ≥ 5 cm H2O) and similar lung protective mechanical ventilation (4) (in particular tidal volume 6 ml/kg predicted body weight and PEEP/FIO2 table). Both found survival benefit in the experimental group. SOFA, continuous NMBA infusion, prone position, combined use of NMBA and prone position, lactate, pH, PaCO2, PaO2/FIO2, lactate, breathing frequency, VT, PEEP, Pplat, Crs and ΔP were recorded at day 1 after inclusion together with gender, age and SAPSII at the time of admission and compared between survivors and nonsurvivors at day 90. Cox proportional hazard models were used with covariates significantly different between survivors and non survivors at the threshold of 0.20 and mortality at day 90 as dependent variable. We planned to test colinearity between ΔP, Crs and Pplat and, if it was verified, to use a specific Cox model for each of them.

t-120 respectively; P < 0.001) was observed in the group of pts with a t-0 ICP ≤ 15 mmHg. A statistically significant reduction in Tc, mean arterial pressure (MAP) and cerebral perfusion pressure (CPP) was observed in all groups of patients (overall population and pts with ICP > or ≤ 15 mmHg) after antipyretic therapy. Conclusions: ICP variations are influenced by ICP value before antipyretics administration. If these data are confirmed in future studies, the decision to start antipyretic therapy should take into account the baseline ICP value. References 1. Thompson HJ et Al. Hyperthermia following traumatic brain injury: a critical evaluation. Neurobiol Dis 2003; 12: 163–173. 2. Picetti E et Al. Intravenous paracetamol for fever control in acute brain injury patients: cerebral and hemodynamic effects. Acta Neurochir 2014; 156(10): 1953–9.

90. Cox proportional hazard models were used with covariates significantly different between survivors and non survivors at the threshold of 0.20 and mortality at day 90 as dependent variable. We planned to test colinearity between ΔP, Crs and Pplat and, if it was verified, to use a specific Cox model for each of them. Results: Both trials enrolled 805 patients of who 787 had data available at day 1 of who 533 survived and 254 did not. In the univariate analysis, ΔP averaged 13.7 ± 3.7 and 12.8 ± 3.7 cmH2O (P = 0.002) in nonsurvivors and survivors, respectively. Colinearity between ΔP, Crs and Pplat was statistically significant. The Cox model for ΔP alone is shown in Table 13. ΔP was associated with a 5 % increased risk of mortality at each 1 cmH2O-increment. Hazard ratios (HR) were 1.05 (1.02-1.08) (P = 0.004) and 0.985 (0.972-0.999) (P = 0.023) for Pplat and Crs, respectively, in the specific analyses. PEEP and VT were not significant risk factors in any model. Conclusions: When lung protective mechanical ventilation is applied ΔP is still a significant predictor of mortality. References 1. Amato MB. N Engl J Med 2015: 372: 747–755 2. Papazian L. New England journal of medicine 2010:363: 1107–1116 3. Guerin C. N Engl J Med 2013:368: 2159–2168 4. ARDSnet N Engl J Med 2000:342: 1301–1308Table 12 (abstract A471). ᅟ

Conclusions: When lung protective mechanical ventilation is applied ΔP is still a significant predictor of mortality. References 1. Amato MB. N Engl J Med 2015: 372: 747–755 2. Papazian L. New England journal of medicine 2010:363: 1107–1116 3. Guerin C. N Engl J Med 2013:368: 2159–2168 4. ARDSnet N Engl J Med 2000:342: 1301–1308Table 12 (abstract A471). ᅟ Variable HR HR 95% lower CI HR 95% upper CI P value Age, per year 1.04 1.03 1.05 <0.001 SOFA, per unit 1.06 1.02 1.10 0.001 Continuous NMBA, reference=yes 0.87 0.65 1.18 0.624 Prone position, reference=yes 0.69 0.50 0.94 0.021 Respiratory rate, per unit 1.01 0.99 1.01 0.376 Arterial pH, per unit 0.049 0.008 0.298 0.001 Lactate, per unit 15.84 1.22 205.18 0.034 Lactate x pH, per unit 0.69 0.48 0.98 0.039 ΔP, per cmH20 1.05 1.02 1.09 0.003

01 Continuous NMBA, reference=yes 0.87 0.65 1.18 0.624 Prone position, reference=yes 0.69 0.50 0.94 0.021 Respiratory rate, per unit 1.01 0.99 1.01 0.376 Arterial pH, per unit 0.049 0.008 0.298 0.001 Lactate, per unit 15.84 1.22 205.18 0.034 Lactate x pH, per unit 0.69 0.48 0.98 0.039 ΔP, per cmH20 1.05 1.02 1.09 0.003 A472 Practice of ventilation in critically ill patients without ARDS at start of mechanical ventilation (provent) - an international observational study A. Serpa Neto1,2, M. Gama de Abreu3, P. Pelosi4, M.J. Schultz2, for the PRoVENT investigators and the PROVE Network 1Hospital Israelita Albert Einstein, Critical Care Medicine, São Paulo, Brazil; 2Academic Medical Center, University of Amsterdam, Intensive Care, Amsterdam, Netherlands; 3University Hospital Carl Gustav Carus, Technische Universität Dresden, 21Pulmonary Engineering Group, Department of Anesthesiology and Intensive Care Medicine, Dresden, Germany; 4IRCCS San Martino IST, University of Genoa, Surgical Sciences and Integrated Diagnostics, Genoa, Italy Correspondence: A. Serpa Neto - Hospital Israelita Albert Einstein, Critical Care Medicine, São Paulo, Brazil INTRODUCTION: Evidence for harm from mechanical ventilation strategies using high tidal volumes in patients without acute respiratory distress syndrome (ARDS) at onset of ventilation is increasing, but it is highly uncertain how these patients are currently ventilated in intensive care units (ICUs) worldwide.

o, Brazil INTRODUCTION: Evidence for harm from mechanical ventilation strategies using high tidal volumes in patients without acute respiratory distress syndrome (ARDS) at onset of ventilation is increasing, but it is highly uncertain how these patients are currently ventilated in intensive care units (ICUs) worldwide. OBJECTIVES: To describe ventilation practice in ICU patients without ARDS at onset of mechanical ventilation, and to compare ventilation settings and outcomes between patients at low versus high risk for ARDS. METHODS: This was an international multicenter prospective cohort of patients without ARDS undergoing invasive ventilation, conducted during one week in 119 ICUs from 16 countries across four continents. The Lung Injury Prediction Score (LIPS) was used for risk stratification for ARDS (≤4 [low] vs. > 4 [high]). The primary outcome was to describe ventilation practice in ICU patients without ARDS at onset of mechanical ventilation. Secondary outcomes included comparison of ventilation settings and outcomes between patients at low versus high risk for ARDS.

S) was used for risk stratification for ARDS (≤4 [low] vs. > 4 [high]). The primary outcome was to describe ventilation practice in ICU patients without ARDS at onset of mechanical ventilation. Secondary outcomes included comparison of ventilation settings and outcomes between patients at low versus high risk for ARDS. RESULTS: Of 3,023 patients screened in participating ICUs, 935 fulfilled inclusion criteria (Fig. 23). Tidal volume sizes, and level of positive end-expiratory pressure (PEEP) were 7.9 [6.8-9.1] ml/kg predicted body weight (PBW), and 5 [5–7] cmH2O, respectively. Tidal volume sizes were not different between patients with low and high risk for ARDS; PEEP level was higher in patients with high risk though differences were minimal (6 [5–8] versus 5 [5–7] cm H2O; p < 0.001) (Fig. 24). Occurrence of pulmonary complications and hospital mortality was higher in patients with high risk for ARDS (35.4 % vs 23.4 %; p < 0.001 and 31.9 % vs 15.8 %; p < 0.001, respectively) (Fig. 25). CONCLUSIONS: A large proportion of patients without ARDS at onset of mechanical ventilation received high tidal volumes. ?A3B2 show $132#?>Patients with higher risk for ARDS had higher incidence of pulmonary complications and worse clinical outcomes. These findings indicate the potential for improvement in the management of patients without ARDS. References - Determann RM, et al. Crit Care 2010;14:R1. - Esteban A, et al. Am J Respir Crit Care Med 2013;188:220–30 - Serpa Neto A, et al. JAMA 2012;308:1651–1659. - Serpa Neto A, et al. Crit Care Med 2015;43:2155–63.

CONCLUSIONS: A large proportion of patients without ARDS at onset of mechanical ventilation received high tidal volumes. ?A3B2 show $132#?>Patients with higher risk for ARDS had higher incidence of pulmonary complications and worse clinical outcomes. These findings indicate the potential for improvement in the management of patients without ARDS. References - Determann RM, et al. Crit Care 2010;14:R1. - Esteban A, et al. Am J Respir Crit Care Med 2013;188:220–30 - Serpa Neto A, et al. JAMA 2012;308:1651–1659. - Serpa Neto A, et al. Crit Care Med 2015;43:2155–63. - Serpa Neto A, et al. Intensive Care Med 2014;40:950–7.Fig. 22 (abstract A472). Flow of Patient Screening and Enrollment Fig. 23 (abstract A472). Ventilation parameters in patientswithout ARDS Fig. 24 (abstract A472). Outcome from patients without ARDS

- Serpa Neto A, et al. JAMA 2012;308:1651–1659. - Serpa Neto A, et al. Crit Care Med 2015;43:2155–63. - Serpa Neto A, et al. Intensive Care Med 2014;40:950–7.Fig. 22 (abstract A472). Flow of Patient Screening and Enrollment Fig. 23 (abstract A472). Ventilation parameters in patientswithout ARDS Fig. 24 (abstract A472). Outcome from patients without ARDS A473 Are changes in diaphragm thickness during mechanical ventilation associated with clinical outcomes? A prospective multi-centre cohort study E.C. Goligher1, E. Fan1, M. Herridge1, S. Vorona1, M. Sklar1, M. Dres1,2, N. Rittayamai1, A. Lanys1, C. Urrea1, G. Tomlinson3, W.D. Reid4, G.D. Rubenfeld1, B.P. Kavanagh1, L.J. Brochard1, N.D. Ferguson1 1University of Toronto, Interdepartmental Division of Critical Care Medicine, Toronto, Canada; 2Sorbonne Paris Cité, UPMC Univ Paris 06 INSERM, UMRS1158, Groupe Hospitalier Pitié-Salpêtrière Charles Foix Service de Pneumologie et Réanimation Médicale, Paris, France; 3University of Toronto, Department of Medicine, Toronto, Canada; 4University of Toronto, Department of Physical Therapy, Toronto, Canada Correspondence: E.C. Goligher - University of Toronto, Interdepartmental Division of Critical Care Medicine, Toronto, Canada Introduction: Changes in diaphragm thickness (Tdi) are common during mechanical ventilation (MV) (1). However, the impact of changes in Tdi on clinical outcomes is unknown and it is uncertain whether targeting specific levels of inspiratory effort during MV would prevent changes in Tdi or accelerate liberation from MV.

Introduction: Changes in diaphragm thickness (Tdi) are common during mechanical ventilation (MV) (1). However, the impact of changes in Tdi on clinical outcomes is unknown and it is uncertain whether targeting specific levels of inspiratory effort during MV would prevent changes in Tdi or accelerate liberation from MV. Objectives: To determine whether changes in Tdi during the early course of MV are associated with impaired liberation from MV and to further establish the links between ventilator settings, patient inspiratory effort, changes in Tdi, and clinical outcomes. Methods: In 3 ICUs, Tdi and diaphragm thickening fraction (TF, a measure of inspiratory effort) were prospectively measured on a daily basis by ultrasound. Patients were classified according to the initial change in Tdi recorded on the first day that the change in Tdi exceeded 10 % up to MV day 7. Patient characteristics, ventilator settings, severity of illness scores and outcomes in hospital were recorded. Predicted relationships were analyzed by multivariable regression modelling and causal mediation analysis.

initial change in Tdi recorded on the first day that the change in Tdi exceeded 10 % up to MV day 7. Patient characteristics, ventilator settings, severity of illness scores and outcomes in hospital were recorded. Predicted relationships were analyzed by multivariable regression modelling and causal mediation analysis. Results: We enrolled 212 patients (outcomes available for 207). Initial changes in Tdi occurred early in the course of MV (median MV day 3, IQR 3–5). Consistent with our previous findings (1), the rate and direction of change in Tdi over time were strongly associated with TF in this cohort (p < 0.001). Controlled MV was associated with an accelerated decline in Tdi (p = 0.01) mediated by TF (proportion of effect mediated = 0.3, p < 0.01 for mediation effect). Both decreased and increased Tdi were associated with prolonged ventilator dependence (Table 12, Fig. 22, log rank p < 0.001), even after adjusting for age, severity of illness, sepsis, and comorbidities. Mean TF below 15 % or above 30 % over the first 3 days of MV was associated with prolonged ventilator dependence in survivors (adjusted p = 0.02) and a higher risk of complications of acute respiratory failure (adjusted p = 0.02). The associations between mean TF and these outcomes were mediated by changes in Tdi (proportion mediated 0.44, p = 0.05, and 0.29, p = 0.02, respectively).

ssociated with prolonged ventilator dependence in survivors (adjusted p = 0.02) and a higher risk of complications of acute respiratory failure (adjusted p = 0.02). The associations between mean TF and these outcomes were mediated by changes in Tdi (proportion mediated 0.44, p = 0.05, and 0.29, p = 0.02, respectively). Conclusions: Early changes in diaphragm thickness following initiation of MV are associated with marked differences in clinical outcomes. Both insufficient and excessive inspiratory effort levels are associated with prolonged ventilator dependence due in part to changes in Tdi. Titrating ventilatory support to maintain TF between 15-30 % might prevent changes in Tdi and accelerate liberation from MV. References 1. Goligher EC et al. Evolution of Diaphragm Thickness during Mechanical Ventilation. Impact of Inspiratory Effort. Am J Respir Crit Care Med 2015;192:1080–1088. Grant acknowledgement Supported by a Post-Doctoral Fellowship from the Canadian Institutes of Health Research.Table 13 (abstract A473). Outcomes associated with changes in Tdi during MV

1. Goligher EC et al. Evolution of Diaphragm Thickness during Mechanical Ventilation. Impact of Inspiratory Effort. Am J Respir Crit Care Med 2015;192:1080–1088. Grant acknowledgement Supported by a Post-Doctoral Fellowship from the Canadian Institutes of Health Research.Table 13 (abstract A473). Outcomes associated with changes in Tdi during MV Initial change in diaphragm thickness during mechanical ventilation Outcome >20% decrease (n=31) 10–20% decrease (n=46) <10% change (n=85) 10–20% increase (n=29) >20% increase (n=18) p-value Hospital mortality (%) 50% 29% 33% 36% 39% 0.41 Ventilator-free days at 60 days (median, IQR) 32 (0–50) 49 (12–55) 54 (0–57) 37 (0–53) 40 (0–50) 0.003 Duration of mechanical ventilation in survivors (median, IQR) 9 (7–14) 8.5 (5–18) 5 (3–8) 9 (5–22) 12 (9.5–20) 0.0002 Complications of acute respiratory failure (death/reintubation/tracheostomy/MV duration > 14 days) (%) 77% 57% 46% 64% 72% 0.023 Fig. 25 (abstract A473). Time to disconnection from the ventilator varies with the initial change in Tdi

ion in survivors (median, IQR) 9 (7–14) 8.5 (5–18) 5 (3–8) 9 (5–22) 12 (9.5–20) 0.0002 Complications of acute respiratory failure (death/reintubation/tracheostomy/MV duration > 14 days) (%) 77% 57% 46% 64% 72% 0.023 Fig. 25 (abstract A473). Time to disconnection from the ventilator varies with the initial change in Tdi A474 Impact of breathing variability on clinical outcomes in mechanically ventilated patients - a prospective multicentric study C. Rolland-Debord1, C. Bureau1, T. Poitou1, M. Clavel2, S. Perbet3, N. Terzi4, A. Kouatchet5, T. Similowski1, A. Demoule1 1Université Pierre et Marie Curie, UMR_S 1158 and Hôpital Pitié-Salpêtrière, Respiratory Division and Medical ICU, Paris, France; 2Hôpital Dupuytren, Limoges, France; 3CHU de Clermont-Ferrand and Université d'Auvergne, Clermont-Ferrand, France; 4Université Grenoble-Alpes and CHU Grenoble Alpes, Grenoble, France; 5CHU d'Angers, Angers, France Correspondence: C. Bureau - Université Pierre et Marie Curie, UMR_S 1158 and Hôpital Pitié-Salpêtrière, Respiratory Division and Medical ICU, Paris, France Introduction: Breathing is a cyclic activity that is not monotonous and is therefore variable. During mechanical ventilation (MV), breathing variability is reduced. Breathing variability has been mostly quantified with two indices: the coefficient of variation (CV) that is defined as the standard deviation to the mean ratio, and the amplitude ratio of the spectrum's first harmonic to its zero frequency (H1/DC). Previous reports have suggested that low variability was associated with weaning failure and increased mortality. However, these studies have quantified variability either on very specific time points or on the contrary on the whole intensive care unit stay.

the spectrum's first harmonic to its zero frequency (H1/DC). Previous reports have suggested that low variability was associated with weaning failure and increased mortality. However, these studies have quantified variability either on very specific time points or on the contrary on the whole intensive care unit stay. Objectives: The aim of our study was to quantify breathing variability at the early phase of weaning according to those two indices. We further evaluated the factors associated with and the prognosis impact of low breathing variability. Methods: Ancillary study of a multicentre, randomized controlled trial comparing neurally ventilator adjusted assist to pressure support ventilation during the early phase of weaning. Airway flow and pressure were recorded during 20 minutes 12, 24, 36 and 48 hours following inclusion. Respiratory rate, tidal volume and mean inspiratory flow were measured. Variability was assessed according to CV and H1/DC. Impact of variability on clinical outcome was determined with a Receiver Operating Characteristic. Results: 108 patients mechanically ventilated for 5 days (3–9) were included, 72 men (68 %), aged 66 (37–86) years, SAPS II 44 (35–59), 62 % were mechanically ventilated for de novo hypoxemic respiratory failure.

Methods: Ancillary study of a multicentre, randomized controlled trial comparing neurally ventilator adjusted assist to pressure support ventilation during the early phase of weaning. Airway flow and pressure were recorded during 20 minutes 12, 24, 36 and 48 hours following inclusion. Respiratory rate, tidal volume and mean inspiratory flow were measured. Variability was assessed according to CV and H1/DC. Impact of variability on clinical outcome was determined with a Receiver Operating Characteristic. Results: 108 patients mechanically ventilated for 5 days (3–9) were included, 72 men (68 %), aged 66 (37–86) years, SAPS II 44 (35–59), 62 % were mechanically ventilated for de novo hypoxemic respiratory failure. Sensitivity and specificity were plotted as functions of the ability of the CV of respiratory rate (RR), tidal volume (Vt) and mean inspiratory flow (Vt/Ti) to predict day-28 mortality. The curves intersected at a coefficient of variation of respectively 0.21 for respiratory rate, 0.16 for tidal volume and 0.14 for mean inspiratory flow (Fig. 26).

ere plotted as functions of the ability of the CV of respiratory rate (RR), tidal volume (Vt) and mean inspiratory flow (Vt/Ti) to predict day-28 mortality. The curves intersected at a coefficient of variation of respectively 0.21 for respiratory rate, 0.16 for tidal volume and 0.14 for mean inspiratory flow (Fig. 26). Day-28 mortality was significantly reduced in patients with a high CV-Vt (>16 %) (12 % vs 31 %, p = 0.02) and in patients with a low H1/DC (≤40 %) (16 % vs 60 % p = 0.04). Low variability was not associated with significant difference in term of hospital and intensive care unit length of stay, ventilator free days, and duration of MV. Pulmonary gas exchanges were greater in patients with high CV-Vt (PaO2/FiO2 = 193 (156–241) vs 230 (183–288), p = 0.03). No significant difference in term of gender, age, SAPS 2, Charlson score or length of MV prior to inclusion was observed with decreased variability. Conclusions: In MV patients at the early phase of weaning, low variability is associated with higher mortality. Low variability is associated with worse oxygenation. Whether low variability is a causative factor of mortality or a consequence of severity remains to be determined.Fig. 26 (abstract A474). Receiver Operating Characteristic (ROC)

ents at the early phase of weaning, low variability is associated with higher mortality. Low variability is associated with worse oxygenation. Whether low variability is a causative factor of mortality or a consequence of severity remains to be determined.Fig. 26 (abstract A474). Receiver Operating Characteristic (ROC) Therapies in neurointensive care: effects on physiology and outcomes A475 Relationship between trough haloperidol concentrations and clinical response in critically ill adults with delirium who are managed with a scheduled IV haloperidol protocol N. Hunfeld1,2, Z. Trogrlic1, S. Ladage1, R.J. Osse3, B. Koch2, W. Rietdijk1, J. Devlin4, M. van der Jagt1 1Erasmus Medical Center, Intensive Care, Rotterdam, Netherlands; 2Erasmus Medical Center, Pharmacy, Rotterdam, Netherlands; 3Erasmus Medical Center, Psychiatry, Rotterdam, Netherlands; 4Northeastern University, School of Pharmacy, Boston, MA, USA Correspondence: N. Hunfeld - Erasmus Medical Center, Intensive Care, Rotterdam, Netherlands Introduction: Intravenous haloperidol (IVH) is frequently administered on a scheduled basis to treat delirium in critically ill adults. However, dosing recommendations for this indication have been derived from older studies evaluating the use of IVH in psychiatric patients [trough level H: 8.4 μg/L (SD 6.7 μg/L), 3–6 mg per day] without delirium and who are not critically ill. To establish the optimal dose of IVH in critically ill adults with delirium, the relationship between IVH serum concentrations and clinical response is required.

es evaluating the use of IVH in psychiatric patients [trough level H: 8.4 μg/L (SD 6.7 μg/L), 3–6 mg per day] without delirium and who are not critically ill. To establish the optimal dose of IVH in critically ill adults with delirium, the relationship between IVH serum concentrations and clinical response is required. Objective: To establish the relationship between serum haloperidol concentrations and clinical response in critically ill adults with delirium who are managed with a scheduled IVH protocol. Methods: This single-center, prospective study included consecutive critically ill adults with delirium without prior exposure to haloperidol, neurological or psychiatric disease, end-stage liver disease, a QTc interval ≥ 450 msec, requiring deep sedation, or requiring a medication known to interact with haloperidol. Delirium was deemed to be present when the bedside nurse and the intensivist agreed that the Intensive Care Delirium Screening Checklist (ICDSC) was ≥ 4 and was managed with a scheduled IVH delirium treatment protocol. IVH was initiated at 1 mg q8h (0.5 mg q8h in patients ≥ 70; 2 mg q8h if agitation present) and titrated upwards q24h by 0.5 mg q8h to 2 mg q8h based on the ICDSC score and delirium symptoms present. Trough haloperidol (TH) serum concentrations were measured on days 2 and 3 of IVH and analyzed with a validated liquid chromatography mass spectometry Method:

q8h in patients ≥ 70; 2 mg q8h if agitation present) and titrated upwards q24h by 0.5 mg q8h to 2 mg q8h based on the ICDSC score and delirium symptoms present. Trough haloperidol (TH) serum concentrations were measured on days 2 and 3 of IVH and analyzed with a validated liquid chromatography mass spectometry Method: Results: Baseline characteristics of the 14 patients enrolled were: male (64 %), mean age of 65 (SD 8), Body Mass Index (BMI) of 28 (SD 5) and a SOFA score of 7 (IQR 4 to 8) and were initiated on IVH dose as follows: 0.5 mg q8h (n = 2), 1 mg q8h (n = 10) and 2 mg q8h (n = 2). IVH dosing across day 2 and 3 of haloperidol use was similar with the day it was initiated. The mean H dose (q8h) in the 48 h up to the TH determinations was 1.6 mg (SD 0.19). At the time of IVH initiation, the median (IQR) ICDSC score was 5 (4 to 5.3) and was relatively unchanged across day 2 [4 (3 to 5)] and day 3 [4 (3 to 5)], p = NS. The mean (SD) H serum concentration was not significantly different between days 2 and 3 [2.2 (1.5) vs. 1.5 (0.9) μg/L; p = 0.15]. Conclusions: This preliminary report is the first study to evaluate the pharmacokinetics and the pharmacodynamic response of low-dose IVH when administered to critically ill adults to treat delirium. We observed low trough levels of IVH and lasting delirium (for the first three days after initiation of IVH). The IVH dose that should be used to treat delirium in the ICU requires further investigation.

etics and the pharmacodynamic response of low-dose IVH when administered to critically ill adults to treat delirium. We observed low trough levels of IVH and lasting delirium (for the first three days after initiation of IVH). The IVH dose that should be used to treat delirium in the ICU requires further investigation. A476 Intracranial pressure after antipyretic therapy in acute brain injury E. Picetti1, P. Ceccarelli1, F. Mensi1, L. Malchiodi1, S. Risolo1, I. Rossi1, M.V. Antonini1, F. Servadei2, M.L. Caspani1 1I Servizio Anestesia Rianimazione, Parma, Italy; 2Neurochirurgia e Neurotraumatologia, Parma, Italy Correspondence: E. Picetti - I Servizio Anestesia Rianimazione, Parma, Italy Introduction: Fever is a dangerous secondary insult for the injured brain (1). In a recent study (2) examining Paracetamol administration for fever control in acute brain injury (ABI) patients (pts), a reduction in intracranial pressure (ICP) was observed only in subjects with a baseline ICP > 15 mmHg. Objectives: 1) to analyze ICP trend after antipyretics administration, 2) to evaluate if ICP variations are influenced by ICP value before antipyretics administration.

A476 Intracranial pressure after antipyretic therapy in acute brain injury E. Picetti1, P. Ceccarelli1, F. Mensi1, L. Malchiodi1, S. Risolo1, I. Rossi1, M.V. Antonini1, F. Servadei2, M.L. Caspani1 1I Servizio Anestesia Rianimazione, Parma, Italy; 2Neurochirurgia e Neurotraumatologia, Parma, Italy Correspondence: E. Picetti - I Servizio Anestesia Rianimazione, Parma, Italy Introduction: Fever is a dangerous secondary insult for the injured brain (1). In a recent study (2) examining Paracetamol administration for fever control in acute brain injury (ABI) patients (pts), a reduction in intracranial pressure (ICP) was observed only in subjects with a baseline ICP > 15 mmHg. Objectives: 1) to analyze ICP trend after antipyretics administration, 2) to evaluate if ICP variations are influenced by ICP value before antipyretics administration. Methods: Adults pts with ABI admitted to our Intensive Care Unit (ICU) were prospectively evaluated after antipyretics administration [4 time points: baseline (t-0), 30 minutes (t-30), 60 minutes (t-60) and 120 minutes (t-120)]. Inclusion criteria were: 1) monitoring of intra-arterial blood pressure, core temperature (Tc), and ICP and 2) a Tc ≥ 37.5 °C. Exclusion criteria were: 1) hypovolemia, 2) administration of drugs with hemodynamic effects during the study period, 3) administration of antipyretics in the 6 hours before the start of the study, 4) acute heart failure and 5) cerebral vasospasm.

Conclusions: ICP variations are influenced by ICP value before antipyretics administration. If these data are confirmed in future studies, the decision to start antipyretic therapy should take into account the baseline ICP value. References 1. Thompson HJ et Al. Hyperthermia following traumatic brain injury: a critical evaluation. Neurobiol Dis 2003; 12: 163–173. 2. Picetti E et Al. Intravenous paracetamol for fever control in acute brain injury patients: cerebral and hemodynamic effects. Acta Neurochir 2014; 156(10): 1953–9. A477 Continuous osmotherapy for the treatment of post-traumatic intracranial hypertension - a multicenter cohort study A. Roquilly1, S. Lasocki2, P. Seguin3, T. Geeraerts4, P.F. Perrigault5, C. Dahyot-Fizelier6, C. Paugam-Burtz7, F. Cook8, R. Cinotti9, D. Demeure dit Latte9, P.J. Mahe9, C. Fortuit9, F. Feuillet9, K. Asehnoune9 1Nantes University Hospital, Creteil, France; 2University Hospital of Angers, Angers, France; 3University Hospital of Rennes, Rennes, France; 4University Hospital of Toulouse, Toulouse, France; 5University hospital of Montpellier, Montpellier, France; 6University Hospital of Poitiers, Poitiers, France; 7APHP, Beaujon, Beaujon, France; 8APHP, Henri Mondor, Creteil, France; 9Nantes University Hospital, Nantes, France Correspondence: A. Roquilly - Nantes University Hospital, Creteil, France Introduction: Intracranial hypertension (ICH) is one of the main cause of death, and of poor neurological recovery after traumatic brain injury (TBI). The use of a continuous osmotherapy has been described in retrospective single-center studies, but its effectiveness is discussed.

s University Hospital, Creteil, France Introduction: Intracranial hypertension (ICH) is one of the main cause of death, and of poor neurological recovery after traumatic brain injury (TBI). The use of a continuous osmotherapy has been described in retrospective single-center studies, but its effectiveness is discussed. Objectives: The first objective was to describe the association between the use of a continuous osmotherapy and mortality in patients developing a post-traumatic ICH. Secondary objectives were to investigate the impact of continuous osmotherapy on long-term neurological recovery and on tolerance. Methods: We pooled the data from three prospective multicenter studies: the corti-TC trial (ref 1), the BI-VILI study (NCT01885507) and the Atlanrea cohort (NCT02426255). We included all patients with traumatic brain-injury, hospitalized in ICU and receiving mechanical ventilation for more than 24 hours. ICH was defined as one or more episodes of intracranial pressure higher than 20 mmHg and that has required a specific therapeutic intervention. In one participating center, continuous osmotherapy (NaCL20% (ref 2)) was initiated as the first line treatment of intracranial hypertension, and its administration was pursued up to the end of the period at risk of ICH. The primary endpoint was the survival rate at day 90. The secondary endpoint was the Glasgow Outcome Scale (GOS) at day 90 (GOS = 1 death, GOS = 5 minor sequelae).

In one participating center, continuous osmotherapy (NaCL20% (ref 2)) was initiated as the first line treatment of intracranial hypertension, and its administration was pursued up to the end of the period at risk of ICH. The primary endpoint was the survival rate at day 90. The secondary endpoint was the Glasgow Outcome Scale (GOS) at day 90 (GOS = 1 death, GOS = 5 minor sequelae). A crude comparison of patients with ICH treated or not with continuous osmotherapy was first performed. To consider the biases related to this observational study, an adjustment on potential confounders (age, glasgow coma scale, pupil reactivity, Marshall score, hypotension and hypoxemia) was performed by both a propensity analysis and a multivariate analysis. The protocol for this study was approved by an ethics committee. Results: Among the 1054 included patients, 545 (51.7 %) developed ICH. Out of 143 patients with ICH and treated with a continuous osmotherapy, 106 (74.1 %) were alive at day 90 as compared to 265 (65.9 %) patients with ICH and not treated with continuous osmotherapy (p = 0.07). With the continuous osmotherapy, the relative risk of survival was 1.43 (95 % CI, 0998–2062, p = 0.05). Adjusted hazard ratio for survival at day 90 was 1.43 (95 % CI, 1.02 - 1.99, p = 0.003) in propensity score-adjusted analysis and 1.67 (95%CI, 1.12-2.50, p = 0.01) in multivariate analysis. At day 90, GOS was higher in patients treated with continous osmotherapy (64 (45.2 %) patients had a GOS = 4–5 vs 115 (35.8 %), p = 0.01).

Out of 143 patients with ICH and treated with a continuous osmotherapy, 106 (74.1 %) were alive at day 90 as compared to 265 (65.9 %) patients with ICH and not treated with continuous osmotherapy (p = 0.07). With the continuous osmotherapy, the relative risk of survival was 1.43 (95 % CI, 0998–2062, p = 0.05). Adjusted hazard ratio for survival at day 90 was 1.43 (95 % CI, 1.02 - 1.99, p = 0.003) in propensity score-adjusted analysis and 1.67 (95%CI, 1.12-2.50, p = 0.01) in multivariate analysis. At day 90, GOS was higher in patients treated with continous osmotherapy (64 (45.2 %) patients had a GOS = 4–5 vs 115 (35.8 %), p = 0.01). Severe hypernatremia (≥160 mmol / l) was more frequent in treated patients (9.1 % vs 2.2 %, p < 0.001). No case of central pontine myelinolysis was recorded. Conclusions: Continuous osmotherapy for the treatment of post-traumatic ICH was associated with improved adjusted 90-day survival. A randomized clinical trial is warranted before definitive Conclusion: References 1. Lancet Respir Med 2014; 2(9):706–16 2. Crit Care 2011; 15(5):R260. Grant acknowledgement None

Severe hypernatremia (≥160 mmol / l) was more frequent in treated patients (9.1 % vs 2.2 %, p < 0.001). No case of central pontine myelinolysis was recorded. Conclusions: Continuous osmotherapy for the treatment of post-traumatic ICH was associated with improved adjusted 90-day survival. A randomized clinical trial is warranted before definitive Conclusion: References 1. Lancet Respir Med 2014; 2(9):706–16 2. Crit Care 2011; 15(5):R260. Grant acknowledgement None A478 Intracerebral hemorrhage in ICU: is it worth treating? C. Marzorati1, S. Spina1, V. Scaravilli1, A. Vargiolu2, M. Riva1, C. Giussani1,3, E. Sganzerla1,3, G. Citerio1,2 1University of Milan - Bicocca, School of Medicine and Surgery, Milan, Italy; 2San Gerardo Hospital, Neurointensive Care, Department of Emergency and Intensive Care, Monza, Italy; 3San Gerardo Hospital, Neurosurgical Clinic, Department of Neurosciences, Monza, Italy Correspondence: C. Marzorati - University of Milan - Bicocca, School of Medicine and Surgery, Milan, Italy Introduction: Prediction of prognosis in Intracerebral hemorrhage (ICH) is always challenging: the ICH score was developed to predict outcomes and limit futile overtreatment1. Recent guidelines on ICH management have significantly modified our traditional nihilistic approach2,3. Objectives: To compare observed and predicted 6-months mortality in a cohort of consecutive ICH patients admitted to the Neurosurgical ICU of San Gerardo Hospital (Monza, Italy) from 2013 to 2015.

A478 Intracerebral hemorrhage in ICU: is it worth treating? C. Marzorati1, S. Spina1, V. Scaravilli1, A. Vargiolu2, M. Riva1, C. Giussani1,3, E. Sganzerla1,3, G. Citerio1,2 1University of Milan - Bicocca, School of Medicine and Surgery, Milan, Italy; 2San Gerardo Hospital, Neurointensive Care, Department of Emergency and Intensive Care, Monza, Italy; 3San Gerardo Hospital, Neurosurgical Clinic, Department of Neurosciences, Monza, Italy Correspondence: C. Marzorati - University of Milan - Bicocca, School of Medicine and Surgery, Milan, Italy Introduction: Prediction of prognosis in Intracerebral hemorrhage (ICH) is always challenging: the ICH score was developed to predict outcomes and limit futile overtreatment1. Recent guidelines on ICH management have significantly modified our traditional nihilistic approach2,3. Objectives: To compare observed and predicted 6-months mortality in a cohort of consecutive ICH patients admitted to the Neurosurgical ICU of San Gerardo Hospital (Monza, Italy) from 2013 to 2015. Methods: Retrospective analysis of prospectively collected data of ICH patients with ICU length-of-stay >24 hours. We retrieved clinical data, CT scans, DNROs within 48 hours from admission, 30 days mortality and modified Rankin Scale (mRS) along with the ICH score and 6-months survival. Results: 100 consecutive patients (64 ± 14 years old, 57 % males, median GCS 7 at admission, median ICH volume 75 cm3) were included. DNRO: 28 %, all died, 71 ± 11 years old, median GCS 4 at admission, median ICH volume 110 cm3.

Methods: Retrospective analysis of prospectively collected data of ICH patients with ICU length-of-stay >24 hours. We retrieved clinical data, CT scans, DNROs within 48 hours from admission, 30 days mortality and modified Rankin Scale (mRS) along with the ICH score and 6-months survival. Results: 100 consecutive patients (64 ± 14 years old, 57 % males, median GCS 7 at admission, median ICH volume 75 cm3) were included. DNRO: 28 %, all died, 71 ± 11 years old, median GCS 4 at admission, median ICH volume 110 cm3. No DNRO: 72 %, 62 ± 15 years old, median GCS 9 at admission, median ICH volume 47 cm3 (p < 0.05 vs. DNRO). Surgery was performed in 71 %. Overall, ICU mortality was lower than expected accordingly to the ICH score (see Fig. 27) (i.e. 33 %). mRS was ≤3 and >3 in 21 (30 %) and 46 (64 %) of the discharged patients, respectively. 180-days survival globally approached 60 % (95%CI 50–70). Considering the severity on admission: - ICH score < 3 (33 %), median length of stay in ICU of 3 days. 17 (51 %) patients underwent to neurosurgery. Mortality was non-significantly different between patients receiving neurosurgery and patients receiving medical treatment. - ICH score ≥ 3 (67 %) had a median length of stay of 5.5 days. Neurosurgery was performed in 34 (50 %), and observed short- and long-term mortality was significantly lower for patients that underwent to neurosurgery than in patients receiving medical treatment (see Figs. 28 and 29).

- ICH score < 3 (33 %), median length of stay in ICU of 3 days. 17 (51 %) patients underwent to neurosurgery. Mortality was non-significantly different between patients receiving neurosurgery and patients receiving medical treatment. - ICH score ≥ 3 (67 %) had a median length of stay of 5.5 days. Neurosurgery was performed in 34 (50 %), and observed short- and long-term mortality was significantly lower for patients that underwent to neurosurgery than in patients receiving medical treatment (see Figs. 28 and 29). Conclusions: We observed mortality to be lower than predicted by ICH score and functional outcome to be acceptable in one-third of patients. Full treatment including neurosurgery significantly improved short and long term survival for patients with ICH score ≥3. Prognosis estimation after an ICH remains a challenging subject and early DNROs and limitations of treatment should be carefully evaluated. References 1. Hemphill, J. C. et al. The ICH score: a simple, reliable grading scale for intracerebral hemorrhage. Stroke J. Cereb. Circ. 32, 891–897 (2001). 2. Morgenstern, L. B. et al. Full medical support for intracerebral hemorrhage. Neurology 84, 1739–1744 (2015). 3. Jolink, W. M. T. et al. Time trends in incidence, case fatality, and mortality of intracerebral hemorrhage. Neurology 85, 1318–1324 (2015).Fig. 27 (abstract A478). Mortality (observed and expected) at 30 days according to ICH score. Fig. 28 (abstract A478). Global 180-days survival (±95%CI) of patients admitted to neuro-ICU with ICH (N= 100 patients).

3. Jolink, W. M. T. et al. Time trends in incidence, case fatality, and mortality of intracerebral hemorrhage. Neurology 85, 1318–1324 (2015).Fig. 27 (abstract A478). Mortality (observed and expected) at 30 days according to ICH score. Fig. 28 (abstract A478). Global 180-days survival (±95%CI) of patients admitted to neuro-ICU with ICH (N= 100 patients). Fig. 29 (abstract A478). 180-days survival of patients undergoing to neurosurgery and receiving only medical treatment Respiratory infections A479 Clinical differences between influenza A (H1N1)PDM09 and influenza A(H3N2) infection in critically ill patients S. Barbadillo1, F.J. González de Molina2,3, F. Álvarez-Lerma4, A. Rodríguez5, SEMICYUC/GETGAG Working Group 1Hospital General de Catalunya, Intensive Care Department, Sant Cugat del Valles, Spain; 2Hospital Universitari Mútua de Terrassa, Intensive Care Department, Terrassa, Spain; 3AGAUR, Grup Recerca Emergent, Terrassa, Spain; 4Hospital del Mar, Intensive Care Department, Barcelona, Spain; 5Hospital Universitari de Tarragona Joan XXIII, Intensive Care Department, Tarragona, Spain Correspondence: S. Barbadillo - Hospital General de Catalunya, Intensive Care Department, Sant Cugat del Valles, Spain Introduction: Influenza infection causes severe morbidity and mortality around the world. Pandemic influenza A(H1N1)pdm09 have been describe to increases disease severity comparing with seasonal flu viruses. There are limited data comparing clinical differences between influenza A (H1N1)pdm09 and influenza A(H3N2) infection in critically ill patients.

causes severe morbidity and mortality around the world. Pandemic influenza A(H1N1)pdm09 have been describe to increases disease severity comparing with seasonal flu viruses. There are limited data comparing clinical differences between influenza A (H1N1)pdm09 and influenza A(H3N2) infection in critically ill patients. Objectives: Our aim was to analyse the demographic and clinical differences between patients admitted to ICU due to pandemic influenza A(H1N1)pdm09 and seasonal influenza A(H3N2) infection. Methods: Prospective, observational, multicenter study conducted in 148 Spanish ICUs from 2009 to 2015. Individuals with Influenza A(H1N1)pdm09 were compared with those infected by influenza A (H3N2). All serotypes were confirmed using RT-PCR at ICU admission. Patients´ demographic, clinical, radiologic features, laboratory values, ICU and hospital length of stay (LOS) and outcomes were recorded. Discrete variables are expressed as percentage and continuous variables as medians with 25th to 75th interquartile range (IQR). Differences between groups were assessed using the x2 test and the Fisher exact test for categoric variables and Mann–Whitney U test for continuous variables.

(LOS) and outcomes were recorded. Discrete variables are expressed as percentage and continuous variables as medians with 25th to 75th interquartile range (IQR). Differences between groups were assessed using the x2 test and the Fisher exact test for categoric variables and Mann–Whitney U test for continuous variables. Results: Of 2898 patients with confirmed influenza infection at ICU admission, 110 were excluded due to influenza B (n = 56) or non-serotype A typification (n = 54). At all, 2421 patients with influenza A (H1N1) were compared to 367 patients with influenza A (H3N2). Patients with A(H1N1) were much younger (50 [38–61] vs 61 [50–73], P < 0.001), presented a higher multiorgan dysfunction (61,2 % vs 55,4 %, P = 0.038), and required more invasive mechanical ventilation (70,6 % vs 65.1 %, P = 0.046), and prone position(18.9 % vs 13.0 %, P = 0.007). Patients with A(H3N2) had more comorbidities: COPD (30.4 % vs 19.5 %, P < 0.001), heart failure(18,6 % vs 10.0 %, P < 0.001), chronic renal failure(14,5 % vs 7,6 %, P < 0.001), and diabetes(23,9 % vs 15,3 %, P < 0.001). Patients with A(H3N2) presented more bacterial coinfection pneumonia (26,5 % vs 14,4 %, P < 0.001) and a higher vaccination rate (23,5 % vs 5,7 %, P < 0.001). Influenza A(H1N1) group have an increased ICU length of stay (LOS) (11[5–21] vs 8[4–19], P < 0.001) but there is no difference in hospital LOS. No statistically significant difference in overall mortality was observed (21.9 % H1N1 vs 24.2 % H3N2, P = 0337).

1) and a higher vaccination rate (23,5 % vs 5,7 %, P < 0.001). Influenza A(H1N1) group have an increased ICU length of stay (LOS) (11[5–21] vs 8[4–19], P < 0.001) but there is no difference in hospital LOS. No statistically significant difference in overall mortality was observed (21.9 % H1N1 vs 24.2 % H3N2, P = 0337). Conclusions: Our data suggest than Infuenza A subtypes have different clinical presentation in critically ill patients. Influenza A(H1N1)pdm09 affect younger patients with less comorbidities suffering an acute and more severe respiratory disease. Influenza A (H3N2) is presented more frecuently with bacterial coinfection in older patients with more chronic illness and especially CPOD. The overall mortality was high without differences between groups. References: Schaffer A.et al. BMC Public Health. 2012; 12:869. Mitchell R. et al. Am J Infect Control. 2013; 41:1032–7.

Conclusions: Our data suggest than Infuenza A subtypes have different clinical presentation in critically ill patients. Influenza A(H1N1)pdm09 affect younger patients with less comorbidities suffering an acute and more severe respiratory disease. Influenza A (H3N2) is presented more frecuently with bacterial coinfection in older patients with more chronic illness and especially CPOD. The overall mortality was high without differences between groups. References: Schaffer A.et al. BMC Public Health. 2012; 12:869. Mitchell R. et al. Am J Infect Control. 2013; 41:1032–7. A480 The dynamics of the pulmonary microbiome during mechanical ventilation in the intensive care unit and the association with occurrence of pneumonia T. Zakharkina1, I. Martin-Loeches2, S. Matamoros1, P. Povoa3, A. Torres4, J. Kastelijn1, J.-J. Hofstra1, M. de Jong1, M. Schultz1, P. Sterk1, A. Artigas5, L.J. Bos1 1Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands; 2St James's University Hospital, Dublin, Ireland; 3Hospital São Francisco Xavier, Lisbon, Portugal; 4Hospital Clinic, Barcelona, Spain,;5Autonomous University of Barcelona, Barcelona, Spain Correspondence: L.J. Bos - Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands Introduction: During the past ten years the paradigm of a “healthy lung is a sterile lung” was challenged [1–5]. The healthy lung appeared to be populated by multiple resident bacterial species, that migrate to the distal airways from the oral cavity [2]. According to the adapted island model the respiratory microbiome represents a dynamic community, where the equilibrium point is achieved by the balance between immigration and elimination mechanisms [6].

to be populated by multiple resident bacterial species, that migrate to the distal airways from the oral cavity [2]. According to the adapted island model the respiratory microbiome represents a dynamic community, where the equilibrium point is achieved by the balance between immigration and elimination mechanisms [6]. Objectives: We hypothesized that mechanical ventilation and antibiotic administration decrease the diversity of the respiratory microbiome and that these changes are more profound in patients who develop ventilator-associated pneumonia (VAP). Methods: Intubated and mechanically ventilated ICU-patients were included. Tracheal aspirates were obtained three times a week. 16S RNA sequencing with the Roche 454 platform was used to measure the composition of the respiratory microbiome. Associations were tested with linear mixed model analysis. The relative changes were compared between patients that did and did not develop VAP.

ncluded. Tracheal aspirates were obtained three times a week. 16S RNA sequencing with the Roche 454 platform was used to measure the composition of the respiratory microbiome. Associations were tested with linear mixed model analysis. The relative changes were compared between patients that did and did not develop VAP. Results: 35 patients were included; 11 had VAP, 18 did not have VAP. Six additional patients developed pneumonia within the first 48 hours after intubation. Duration of mechanical ventilation was associated with a decrease in Shannon diversity (fixed-effect regression coefficient (ß): −0.03 [95%CI: −0.05 - -0.005]), but the administration of antibiotic therapy was not (fixed-effect ß: 0.06; 95%CI: −0.17 - 0.30). There was a statistically significant increase in abundance of Pseudomonadales and a decrease in abundance of Lactobacillales, Gemellales, Actinomycetales and Clostridiales between the moment of intubation and extubation. There were no statistically significant differences between patients that did and did not develop VAP. Conclusions: Mechanical ventilation, but not antibiotic administration, was associated with changes in the respiratory microbiome. Duration of mechanical ventilation was associated with increased abundances of Pseudomonadales. The dynamics in the respiratory microbiome were not different between patients that did and did not develop VAP. References 1. Dickson RP. Ann Am Thorac Soc 2015. 2. Bassis CM. mBio 2015. 3. Morris A. Am J Respir Crit Care Med 2013. 4. Zakharkina T. PLoS One 2013. 5. Segal LN. Ann Am Thorac Soc 2014.

Conclusions: Mechanical ventilation, but not antibiotic administration, was associated with changes in the respiratory microbiome. Duration of mechanical ventilation was associated with increased abundances of Pseudomonadales. The dynamics in the respiratory microbiome were not different between patients that did and did not develop VAP. References 1. Dickson RP. Ann Am Thorac Soc 2015. 2. Bassis CM. mBio 2015. 3. Morris A. Am J Respir Crit Care Med 2013. 4. Zakharkina T. PLoS One 2013. 5. Segal LN. Ann Am Thorac Soc 2014. 6. Dickson RP. Annu Rev Physiol 2015. Grant acknowledgement Grant by the Instituto de Salud Carlos III (ISCIII) (ISCIII/FIS-PI 12/01815) Spanish Government and Institut Merieux Research grant.

2. Bassis CM. mBio 2015. 3. Morris A. Am J Respir Crit Care Med 2013. 4. Zakharkina T. PLoS One 2013. 5. Segal LN. Ann Am Thorac Soc 2014. 6. Dickson RP. Annu Rev Physiol 2015. Grant acknowledgement Grant by the Instituto de Salud Carlos III (ISCIII) (ISCIII/FIS-PI 12/01815) Spanish Government and Institut Merieux Research grant. A481 Impact of immunosuppression on the incidence, etiology, and outcome of ventilator-associated lower respiratory tract infections: a post-hoc analysis of the TAVeM database A.-S. Moreau1, I. Martin-Loeches2, P. Povoa3, J. Salluh4, A. Rodriguez5, S. Nseir1, TAVeM study group 1Lille University Hospital, ICU, Lille, France; 2Trinity Centre for Health Sciences, St James's University Hospital, Critical Care Medicine, Dublin, Ireland; 3Centro Hospitalar de Lisboa Ocidental, São Francisco Xavier Hospital, ICU, Lisbon, Portugal; 4D'Or Institute for Research and Education, Rio de Janeiro, Brazil; 5Joan XXIII University Hospital, Institut d'Investigació Sanitària Pere Virgili, Tarragona, Spain Correspondence: A.-S. Moreau - Lille University Hospital, ICU, Lille, France Introduction: Immunocompromised (IC) patients have poor outcome in the ICU. To our knowledge, no study to date has specifically evaluated ventilator-associated (VA) lower respiratory tract infections (LRTI) in this population. Objectives: To determine the incidence, etiology and outcome of VA-LRTI in IC patients, and to compare it with patients with no apparent immunosuppression.

A481 Impact of immunosuppression on the incidence, etiology, and outcome of ventilator-associated lower respiratory tract infections: a post-hoc analysis of the TAVeM database A.-S. Moreau1, I. Martin-Loeches2, P. Povoa3, J. Salluh4, A. Rodriguez5, S. Nseir1, TAVeM study group 1Lille University Hospital, ICU, Lille, France; 2Trinity Centre for Health Sciences, St James's University Hospital, Critical Care Medicine, Dublin, Ireland; 3Centro Hospitalar de Lisboa Ocidental, São Francisco Xavier Hospital, ICU, Lisbon, Portugal; 4D'Or Institute for Research and Education, Rio de Janeiro, Brazil; 5Joan XXIII University Hospital, Institut d'Investigació Sanitària Pere Virgili, Tarragona, Spain Correspondence: A.-S. Moreau - Lille University Hospital, ICU, Lille, France Introduction: Immunocompromised (IC) patients have poor outcome in the ICU. To our knowledge, no study to date has specifically evaluated ventilator-associated (VA) lower respiratory tract infections (LRTI) in this population. Objectives: To determine the incidence, etiology and outcome of VA-LRTI in IC patients, and to compare it with patients with no apparent immunosuppression. Methods: Post-hoc analysis of the large prospective multinational TAVeM database, coming from 114 ICUs [1]. All consecutive patients receiving mechanical ventilation for >48 h were included. The incidence, etiology and outcome of VA-LRTI (ventilator-associated tracheobronchitis (VAT), and ventilator-associated pneumonia (VAP) [1]) were compared between IC (neoplasia, haematological malignancy, AIDS, allogeneic stem cell transplant, immunosuppressant drug, organ transplant) and non-IC patients.

re included. The incidence, etiology and outcome of VA-LRTI (ventilator-associated tracheobronchitis (VAT), and ventilator-associated pneumonia (VAP) [1]) were compared between IC (neoplasia, haematological malignancy, AIDS, allogeneic stem cell transplant, immunosuppressant drug, organ transplant) and non-IC patients. Results: Among the 2960 included patients, 663 (22 %) were IC. The incidence of VA-LRTI was significantly lower in IC compared with non-IC patients (116 (17.5 %) versus 573 (25 %), p < 0.0001, OR = 0.64 [95%CI 0.51-0.79]). Although VAT incidence was significantly lower in IC compared with non-IC patients (7.8 % vs 11.7 %, p = 0.04), no significant difference was found in VAP incidence between the two groups (9.7 % vs 13.3 %, p = 0.13, respectively). In patients with VA-LRTI, rate of prior antibiotic treatment was significantly higher in IC compared with non-IC patients (80 % vs 66 %, p = 0.001). The incidence of progression from VAT to VAP was similar in IC and non-IC patients (7/52 [13 %] vs 32/268 [12 %], p = 0.76). IC and non-IC patients received appropriate antibiotics for VA-LRTI in the same proportions (77 % vs 79 %, p = 0.5). Same results were obtained in VAT and VAP subgroups. Among IC patients with VAT, 39/52 (75 %) received appropriate antibiotics. Percentage of IC patients with progression from VAT to VAP was significantly lower in patients who received appropriate compared with those who received inappropriate antibiotic treatment (8 % vs 31 %, p = 0.035, OR 0.19 (95 % CI 0.03-0.99)).

s. Among IC patients with VAT, 39/52 (75 %) received appropriate antibiotics. Percentage of IC patients with progression from VAT to VAP was significantly lower in patients who received appropriate compared with those who received inappropriate antibiotic treatment (8 % vs 31 %, p = 0.035, OR 0.19 (95 % CI 0.03-0.99)). The incidence of multidrug resistant (MDR) bacteria was higher in IC compared to non-IC patients with VA-LRTI (72 % vs 59 %, p = 0.01), similar results were found in the subgroup of patients with VAP (78 % vs 58 %, p = 0.001), but not in those with VAT (65 % vs 61 %, p = 0.52). Among patients with VA-LRTI, MRSA (5.2 % vs 1.7 %, p = 0.025) and Enterobacter spp. (21 % vs 10 %, p = 0.001) were significantly more frequent in IC compared with non-IC patients. ICU mortality rate was higher in IC compared with non-IC patients with VA-LRTI (54 %, vs 30 %, p < 0.0001, OR 2.68 (95 % CI 1.78-4.02)). Similar results were obtained in VAT and VAP subgroups. Conclusions: Incidence of VA-LRTI is significantly lower in IC compared with non-IC patients. MDR bacteria and mortality rates are significantly higher in IC compared with non-IC patients with VA-LRTI. References 1. Martin-Loeches I et al. Lancet Respir Med 2015, 3:859–68Table 14 (abstract A481). Outcomes of patients with VA-LRTI

Conclusions: Incidence of VA-LRTI is significantly lower in IC compared with non-IC patients. MDR bacteria and mortality rates are significantly higher in IC compared with non-IC patients with VA-LRTI. References 1. Martin-Loeches I et al. Lancet Respir Med 2015, 3:859–68Table 14 (abstract A481). Outcomes of patients with VA-LRTI Immunocompromised n=663 Non-immunocompromised n = 2297 VAT n=52 VAP n=64 No VA-LRTI n=547 p VAT n=268 VAP n=305 No VA-LRTI n=1724 p Mechanical ventilation duration, d 16 (10–25.5) 15 (8–27) 7 (4–14) <0.001 13 (8–22) 14 (8–26) 7 (4–12) <0.001 Length of ICU stay, d 23 (16–38) 20 (13–30) 12 (7–20) <0.001 21 (14–33) 21 (13–34) 12 (8–19) <0.001 ICU mortality, % 42 64 39.5 0.001 26.5 34 26.5 0.016

o VA-LRTI n=547 p VAT n=268 VAP n=305 No VA-LRTI n=1724 p Mechanical ventilation duration, d 16 (10–25.5) 15 (8–27) 7 (4–14) <0.001 13 (8–22) 14 (8–26) 7 (4–12) <0.001 Length of ICU stay, d 23 (16–38) 20 (13–30) 12 (7–20) <0.001 21 (14–33) 21 (13–34) 12 (8–19) <0.001 ICU mortality, % 42 64 39.5 0.001 26.5 34 26.5 0.016 A482 Procalcitonin guided antibiotic therapy in severe community- acquired pneumonia. Randomized controlled trial E. de Jong1, J.A. van Oers2, A. Beishuizen3, A.R.J. Girbes1, M.W.N. Nijsten4, D.W. de Lange5 1VU University Medical Center Amsterdam, Intensive Care, Amsterdam, Netherlands; 2ST Elisabeth Twee Steden Hospital, Intensive Care, Tilburg, Netherlands; 3Medisch Spectrum Twente, Intensive Care, Enschede, Netherlands; 4University Medical Center Groningen, Intensive Care, Amsterdam, Netherlands; 5University Medical Center Utrecht, Intensive Care, Utrecht, Netherlands Correspondence: E. de Jong - VU University Medical Center Amsterdam, Intensive Care, Amsterdam, Netherlands Introduction: Severe community-acquired pneumonia (CAP) remains a significant cause of morbidity and mortality [1]. In the Netherlands the guidelines recommends a duration of antibiotic therapy in severe CAP for at least five till seven days. Objectives: The purpose of this trial was to evaluate whether procalcitonin measurements were able to reduce antibiotic usage in patients with severe CAP in Dutch intensive care units by reducing the duration of antibiotic treatment without increasing mortality or recurrent infections.

A482 Procalcitonin guided antibiotic therapy in severe community- acquired pneumonia. Randomized controlled trial E. de Jong1, J.A. van Oers2, A. Beishuizen3, A.R.J. Girbes1, M.W.N. Nijsten4, D.W. de Lange5 1VU University Medical Center Amsterdam, Intensive Care, Amsterdam, Netherlands; 2ST Elisabeth Twee Steden Hospital, Intensive Care, Tilburg, Netherlands; 3Medisch Spectrum Twente, Intensive Care, Enschede, Netherlands; 4University Medical Center Groningen, Intensive Care, Amsterdam, Netherlands; 5University Medical Center Utrecht, Intensive Care, Utrecht, Netherlands Correspondence: E. de Jong - VU University Medical Center Amsterdam, Intensive Care, Amsterdam, Netherlands Introduction: Severe community-acquired pneumonia (CAP) remains a significant cause of morbidity and mortality [1]. In the Netherlands the guidelines recommends a duration of antibiotic therapy in severe CAP for at least five till seven days. Objectives: The purpose of this trial was to evaluate whether procalcitonin measurements were able to reduce antibiotic usage in patients with severe CAP in Dutch intensive care units by reducing the duration of antibiotic treatment without increasing mortality or recurrent infections. Methods: From 2009 until 2013 a randomized intervention trial was performed in three university medical centres and 12 teaching hospitals in the Netherlands [2]. In this Dutch multicentre trial all patients admitted to the intensive care unit and who received antibiotics for presumed infection were assigned to a PCT-guided or standard-of-care antibiotic discontinuation. In this study a sub analysis for all patients with a severe community-acquired pneumonia was performed. Stopping advice for antibiotics was provided when PCT was ≤20 % of its peak value or ≤0.5 ug/L. Mortality and duration of antibiotic treatment (DOT) were the primary endpoints of this study.

tibiotic discontinuation. In this study a sub analysis for all patients with a severe community-acquired pneumonia was performed. Stopping advice for antibiotics was provided when PCT was ≤20 % of its peak value or ≤0.5 ug/L. Mortality and duration of antibiotic treatment (DOT) were the primary endpoints of this study. Results: In 15 ICUs 440 patients with severe community acquired pneumonia were randomized for PCT-guidance (n = 208) or standard-of-care (n = 232). The median DOT was 5.5 (3–8) and 7 days (4–10) respectively (P < 0 · 001). No difference in reinstitution of antibiotics or CRP-levels within 28 days after randomization was observed. Mortality at 28 days was 43/208 (20.7 %) and 58/232 (25 · 0 %) in the PCT and standard-of-care group respectively (P = 0.31). One-year mortality was 63/208 (30.3 %) and 90/232 (38.8 %) in the PCT and standard-of-care group respectively. Conclusions: In this prospective randomized trial, the addition of PCT-measurements to assist intensivists in duration of antibiotic therapy in severe community-acquired pneumonia, resulted in a clear reduction of antibiotic treatment from 7 days (IQR 4–10 days) in the control group to 5.5 days (IQR 3–8 days) in the PCT-guided patient group, without increase of mortality or subsequent antibiotic prescriptions. References 1. Ramírez P1, Ferrer M, Martí V, Reyes S, Martínez R, Menéndez R, Ewig S, Torres. Inflammatory biomarkers and prediction for intensive care unit admission in severe community-acquired pneumonia. Crit Care Med. 2011 Oct;39(10):2211–7. doi: 10.1097/CCM.0b013e3182257445.

Conclusions: In this prospective randomized trial, the addition of PCT-measurements to assist intensivists in duration of antibiotic therapy in severe community-acquired pneumonia, resulted in a clear reduction of antibiotic treatment from 7 days (IQR 4–10 days) in the control group to 5.5 days (IQR 3–8 days) in the PCT-guided patient group, without increase of mortality or subsequent antibiotic prescriptions. References 1. Ramírez P1, Ferrer M, Martí V, Reyes S, Martínez R, Menéndez R, Ewig S, Torres. Inflammatory biomarkers and prediction for intensive care unit admission in severe community-acquired pneumonia. Crit Care Med. 2011 Oct;39(10):2211–7. doi: 10.1097/CCM.0b013e3182257445. 2. Wiersinga WJ, Bonten MJ, Boersma WG, Jonkers RE, Aleva RM, Kullberg BJ, Schouten JA, Degener JE, Janknegt R, Verheij TJ, Sachs AP, Prins JM. SWAB/NVALT guidelines on the management of community-acquired pneumonia in adults. Neth J Med. 2012 Mar;70(2):90–101. 3. de Jong E, van Oers JA, Beishuizen A, Vos P, Vermeijden WJ, Haas LE, et al. Efficacy and safety of procalcitonin guidance in reducing the duration of antibiotic treatment in critically ill patients: a randomised, controlled, open-label trial. Lancet Infect Dis. 2016 Feb 29Fig. 30 (abstract A482). Kaplan-Meier plot for survival

E, van Oers JA, Beishuizen A, Vos P, Vermeijden WJ, Haas LE, et al. Efficacy and safety of procalcitonin guidance in reducing the duration of antibiotic treatment in critically ill patients: a randomised, controlled, open-label trial. Lancet Infect Dis. 2016 Feb 29Fig. 30 (abstract A482). Kaplan-Meier plot for survival A483 Type of stress-ulcer prophylaxis and incidence of ventilator-associated pneumonia D. Bonvicini1, D. Labate1, L. Benacchio2, A. Olivieri2, E. Pizzirani1 1Ulss 15 Alta Padovana, Anesthesia and Intensive Care Unit, Camposampiero, Italy; 2Ulss 15 Alta Padovana, Epidemiology Unit, Camposampiero, Italy Correspondence: D. Labate - Ulss 15 Alta Padovana, Anesthesia and Intensive Care Unit, Camposampiero, Italy Introduction: Ventilator-Associated Pneumonia (VAP) is a leading nosocomial infection in intensive care unit (ICU) patients undergoing mechanical ventilation (MV). Giving that such patients show an increasing risk of important gastrointestinal (GI) bleeding, stress-ulcer prophylaxis (SUP) has been recommended for the prevention of upper GI hemorrhage. SUP strategy rely on drugs that block the secretion of gastric acid and increase the gastric pH (histamine-2-receptor antagonists - H2RA, and proton pump inhibitor - PPI) and those that does not alter gastric pH (sucralfate). The increase of gastric pH leads to bacterial overgrowth and potential colonization of trachea determining a higher risk of VAP. Objectives: In the setting of a study designated to assess VAP outcomes, we collected data on known VAP risk factors (SUP strategy).

A483 Type of stress-ulcer prophylaxis and incidence of ventilator-associated pneumonia D. Bonvicini1, D. Labate1, L. Benacchio2, A. Olivieri2, E. Pizzirani1 1Ulss 15 Alta Padovana, Anesthesia and Intensive Care Unit, Camposampiero, Italy; 2Ulss 15 Alta Padovana, Epidemiology Unit, Camposampiero, Italy Correspondence: D. Labate - Ulss 15 Alta Padovana, Anesthesia and Intensive Care Unit, Camposampiero, Italy Introduction: Ventilator-Associated Pneumonia (VAP) is a leading nosocomial infection in intensive care unit (ICU) patients undergoing mechanical ventilation (MV). Giving that such patients show an increasing risk of important gastrointestinal (GI) bleeding, stress-ulcer prophylaxis (SUP) has been recommended for the prevention of upper GI hemorrhage. SUP strategy rely on drugs that block the secretion of gastric acid and increase the gastric pH (histamine-2-receptor antagonists - H2RA, and proton pump inhibitor - PPI) and those that does not alter gastric pH (sucralfate). The increase of gastric pH leads to bacterial overgrowth and potential colonization of trachea determining a higher risk of VAP. Objectives: In the setting of a study designated to assess VAP outcomes, we collected data on known VAP risk factors (SUP strategy). Our objective was to evaluate the risk of VAP accounting for exposure to H2RA, PPI and sucralfate.

A483 Type of stress-ulcer prophylaxis and incidence of ventilator-associated pneumonia D. Bonvicini1, D. Labate1, L. Benacchio2, A. Olivieri2, E. Pizzirani1 1Ulss 15 Alta Padovana, Anesthesia and Intensive Care Unit, Camposampiero, Italy; 2Ulss 15 Alta Padovana, Epidemiology Unit, Camposampiero, Italy Correspondence: D. Labate - Ulss 15 Alta Padovana, Anesthesia and Intensive Care Unit, Camposampiero, Italy Introduction: Ventilator-Associated Pneumonia (VAP) is a leading nosocomial infection in intensive care unit (ICU) patients undergoing mechanical ventilation (MV). Giving that such patients show an increasing risk of important gastrointestinal (GI) bleeding, stress-ulcer prophylaxis (SUP) has been recommended for the prevention of upper GI hemorrhage. SUP strategy rely on drugs that block the secretion of gastric acid and increase the gastric pH (histamine-2-receptor antagonists - H2RA, and proton pump inhibitor - PPI) and those that does not alter gastric pH (sucralfate). The increase of gastric pH leads to bacterial overgrowth and potential colonization of trachea determining a higher risk of VAP. Objectives: In the setting of a study designated to assess VAP outcomes, we collected data on known VAP risk factors (SUP strategy). Our objective was to evaluate the risk of VAP accounting for exposure to H2RA, PPI and sucralfate. Methods: We recruited 772 patients in a multicenter prospective observational study, conducted in 21 Italian ICUs. Patients admitted to the ICU for at least 48 hours, received MV from the admission, and were followed until discharge, the day of withholding MV, the day of transfer in other hospital ward, or death.We recorded demographic data, admission category, and severity scores on admission. Each day of ICU staying SUP approach (H2RA, PPI or sucralfate) was recorded. VAP was defined as pneumonia occurring after a MV period of at least 48 hours during the ICU stay.Independent relationships between VAP incidence and the specific SUP approach was assessed by using a logistic regression analysis.

on admission. Each day of ICU staying SUP approach (H2RA, PPI or sucralfate) was recorded. VAP was defined as pneumonia occurring after a MV period of at least 48 hours during the ICU stay.Independent relationships between VAP incidence and the specific SUP approach was assessed by using a logistic regression analysis. Results: Among 772 patients experiencing on average 11.3 days of MV, 116 (15 %) developed VAP. Sucralfate was less frequently used (8 %), while 28 % has been receiving H2RA and 77 % PPI. The risk of VAP was higher for patients treated mainly with PPI (OR 4.27, 95%CI 2.57-7.09) than those who received H2RA (OR 2.30, 95%CI 1.08-5.07). Conclusions: Our study show that the use of ranitidine is associated with a lower risk of VAP, compared with PPI. Sucralfate was less frequently used in the participating ICUs. Although some meta-analysis indicate a lower risk of VAP associated with sucralfate, an important study have not established a superiority of this medication compared with ranitidine [1]. Miano et al. compared the use of pantoprazole vs ranitidine in Cardiothoracic Surgery patients showing that pantoprazole was associated with a higher rate of VAP [2]. Such VAP rate associated with PPI would seem to be linked to a greater antacid power than ranitidine. This can lead to a more extensive gastric bacterial colonization in patients treated with these drugs. Randomized-Controlled Trial of Stress-ulcer prophylaxis are needed to assess which drug show the highest risk of VAP and the relationship with the incidence of major GI bleeding. References

Conclusions: Our study show that the use of ranitidine is associated with a lower risk of VAP, compared with PPI. Sucralfate was less frequently used in the participating ICUs. Although some meta-analysis indicate a lower risk of VAP associated with sucralfate, an important study have not established a superiority of this medication compared with ranitidine [1]. Miano et al. compared the use of pantoprazole vs ranitidine in Cardiothoracic Surgery patients showing that pantoprazole was associated with a higher rate of VAP [2]. Such VAP rate associated with PPI would seem to be linked to a greater antacid power than ranitidine. This can lead to a more extensive gastric bacterial colonization in patients treated with these drugs. Randomized-Controlled Trial of Stress-ulcer prophylaxis are needed to assess which drug show the highest risk of VAP and the relationship with the incidence of major GI bleeding. References 1 Cook D et al. N Eng J Med 1998; 338:791–7 2 Miano TA et al. Chest 2009; 136(2):440–7

Conclusions: Our study show that the use of ranitidine is associated with a lower risk of VAP, compared with PPI. Sucralfate was less frequently used in the participating ICUs. Although some meta-analysis indicate a lower risk of VAP associated with sucralfate, an important study have not established a superiority of this medication compared with ranitidine [1]. Miano et al. compared the use of pantoprazole vs ranitidine in Cardiothoracic Surgery patients showing that pantoprazole was associated with a higher rate of VAP [2]. Such VAP rate associated with PPI would seem to be linked to a greater antacid power than ranitidine. This can lead to a more extensive gastric bacterial colonization in patients treated with these drugs. Randomized-Controlled Trial of Stress-ulcer prophylaxis are needed to assess which drug show the highest risk of VAP and the relationship with the incidence of major GI bleeding. References 1 Cook D et al. N Eng J Med 1998; 338:791–7 2 Miano TA et al. Chest 2009; 136(2):440–7 Clinical studies in perioperative ICU A484 Influence of postoperative albumin levels in the outcome of cardiac surgery J.C. Lopez-Delgado1, M. Gonzalez-Romero1, V. Fuentes-Mila1, D. Berbel-Franco1, I. Romera-Peregrina1, A. Martinez-Pascual1, J. Perez-Sanchez1, R. Abellan-Lencina1, R.E. Ávila-Espinoza1, G. Moreno-Gonzalez1, F. Sbraga2 1Hospital Universitari de Bellvitge, Intensive Care, L' Hospitalet de Llobregat, Spain; 2Hospital Universitari de Bellvitge, Cardiac Surgery, L' Hospitalet de Llobregat, Spain Correspondence: J.C. Lopez-Delgado - Hospital Universitari de Bellvitge, Intensive Care, L' Hospitalet de Llobregat, Spain Introduction: Low preoperative albumin reflects a poor nutritional status that influences outcome in cardiac surgery (CS). However, postoperative albumin levels are influenced by nutritional status and inflammatory response. Thus, its value and utility as a prognostic marker is unclear.

' Hospitalet de Llobregat, Spain Introduction: Low preoperative albumin reflects a poor nutritional status that influences outcome in cardiac surgery (CS). However, postoperative albumin levels are influenced by nutritional status and inflammatory response. Thus, its value and utility as a prognostic marker is unclear. Objectives: To evaluate the influence of postoperative plasma levels of albumin in the outcome of patients who underwent cardiac surgery (CS). Methods: Prospective, observational study in Surgical ICU in a referral hospital. Albumin was measured during the first 24 h after CS, together with clinical data and outcomes including in-hospital and long-term mortality (follow-up of 4.6 ± 2.4 years). Patients were classified into different categories based on plasma albumin levels: normal(≥35 g•L−1), low deficit(30–34.9 g•L−1), moderate deficit(25–29.9 g•L−1) and severe deficit (<25 g•L−1). Results: 2818 patients were included. Mean age was 64.5 ± 11.6 years; 63.8 % (n = 1799) were male; Body Mass Index: 28 ± 4.3Kg · m-2. 5.8 %(n = 162) had normal levels, 32.8 %(n = 924) low deficit, 44.3 %(n = 1249) moderate deficit and 17.1 %(n = 483) severe deficit. We showed that not having low albumin levels 24 h after CS was protector for in-hospital (HR:0.844;95 % IC:0.805-0.844;P = 0.007) and long-term mortality (HR:0.846; 95 % IC:0.821-0.871; P < 0.001).

162) had normal levels, 32.8 %(n = 924) low deficit, 44.3 %(n = 1249) moderate deficit and 17.1 %(n = 483) severe deficit. We showed that not having low albumin levels 24 h after CS was protector for in-hospital (HR:0.844;95 % IC:0.805-0.844;P = 0.007) and long-term mortality (HR:0.846; 95 % IC:0.821-0.871; P < 0.001). Subgroups of patients with low albumin levels showed worst survival during hospital stay (98.1 % for normal subgroup, 97.3 % low deficit, 95 % moderate deficit and 85.9 % severe deficit; P < 0.001) and from the long-term mortality scenario (5-year mortality was 94.3 % for normal subgroup, 87.4 % low deficit, 83.1 % moderate deficit and 72.4 % severe deficit; P < 0.001). Multivariable analysis showed higher in-hospital mortality, sepsis and hemorrhage related complications and higher ICU stay in subgroups of patients with low albumin levels Predictors for the presence of any levels of hypoalbuminemia were suffering from any degree of chronic kidney disease (OR:1.316;95 % IC:1.085-1.595;P = 0.005), preoperative low hemoglobin levels (OR:1.060;95 % IC:1.033-1.088;P < 0.001), previous CS (OR:1.229;1.067-1.415;P = 0.004) and longer Cardiopulmonary bypass time (OR:1.904;95 % IC:1.902-2.128;P < 0.001). Conclusions: Physicians may be aware about the degree of hypoalbuminemia in the postoperative of CS due to its influence in outcomes, even in long-term survival. Relationship between factors associated with nutritional and inflammatory status may be associated with the development of postoperative hypoalbuminemia. References

Conclusions: Physicians may be aware about the degree of hypoalbuminemia in the postoperative of CS due to its influence in outcomes, even in long-term survival. Relationship between factors associated with nutritional and inflammatory status may be associated with the development of postoperative hypoalbuminemia. References 1. Karas PL, Goh SL, Dhital K. Is low serum albumin associated with postoperative complications in patients undergoing cardiac surgery? Interact Cardiovasc Thorac Surg. 2015 Dec;21(6):777–86. doi: 10.1093/icvts/ivv247.

Conclusions: Physicians may be aware about the degree of hypoalbuminemia in the postoperative of CS due to its influence in outcomes, even in long-term survival. Relationship between factors associated with nutritional and inflammatory status may be associated with the development of postoperative hypoalbuminemia. References 1. Karas PL, Goh SL, Dhital K. Is low serum albumin associated with postoperative complications in patients undergoing cardiac surgery? Interact Cardiovasc Thorac Surg. 2015 Dec;21(6):777–86. doi: 10.1093/icvts/ivv247. A485 Reducing the incidence of post-operative pulmonary complications (POPC) in high-risk surgical patients: development of a care bundle (CB) to be applied in addition to standard enhanced recovery pathways - a survey of the ESICM POIC section and a Delphi expert consensus by the POPC-CB investigators S. Griffiths1, M.P.W. Grocott1,2, B. Creagh-Brown3,4, POPC-CB investigators 1University of Southampton, Southampton, UK; 2University Hospital Southampton, Southampton, UK; 3Royal Surrey County Hospital, ICU and SPACeR Research Group, Guildford, UK; 4University of Surrey, Guildford, UK Correspondence: B. Creagh-Brown - University of Surrey, Guildford, UK Introduction: Post-operative pulmonary complications (POPC) are common with a reported incidence ranging from 2 %-40 %. Adverse outcomes include death, longer hospital stays and lower long term survival, which is independent of pre-operative risk. The most widely recognised prediction tool for POPCs is ARISCAT. Interventions to reduce POPCs have been studied individually but the use of a care bundle (CB) has not been widely investigated. Enhanced recovery (ER) pathways are 'standard of care' for major surgery - however some patients may benefit from additional focused interventions. An example of a CB for POPCs is the ´I COUGH´ study, which demonstrated a trend towards reduction in POPCs. Results may be superior if the intervention is targeted to a high-risk group.

very (ER) pathways are 'standard of care' for major surgery - however some patients may benefit from additional focused interventions. An example of a CB for POPCs is the ´I COUGH´ study, which demonstrated a trend towards reduction in POPCs. Results may be superior if the intervention is targeted to a high-risk group. Objectives: 1) To perform a survey of members of the ESICM POIC section: to choose international experts for the Delphi and to share their opinions on a POPC-CB. 2) To recruit a team of experts to participate in and complete an email-based Delphi consensus, leading to the formulation of a CB. Methods: A SurveyMonkey of members of the ESICM POIC section was conducted. International experts were identified from this survey. A Delphi consensus was conducted anonymously via email over 3 rounds, to 36 independently chosen experts. Each round comprised of a series of statements which experts were asked to respond to using a five-point Likert scale. There were free text spaces for comments and the possibility for papers to be circulated. After each round, anonymous feedback was circulated to the Delphi group and the previous results were used to create the statements for the subsequent round. At the end of the process, final feedback was sent to the group with the consensus and suggested care bundle. Results: Survey: 362 respondents. ·Routine screening for high-risk of POPC occurs in 50 % of centres. ·Information about current practice in relation to various pre-, intra- and post-operative interventions was ascertained.

Methods: A SurveyMonkey of members of the ESICM POIC section was conducted. International experts were identified from this survey. A Delphi consensus was conducted anonymously via email over 3 rounds, to 36 independently chosen experts. Each round comprised of a series of statements which experts were asked to respond to using a five-point Likert scale. There were free text spaces for comments and the possibility for papers to be circulated. After each round, anonymous feedback was circulated to the Delphi group and the previous results were used to create the statements for the subsequent round. At the end of the process, final feedback was sent to the group with the consensus and suggested care bundle. Results: Survey: 362 respondents. ·Routine screening for high-risk of POPC occurs in 50 % of centres. ·Information about current practice in relation to various pre-, intra- and post-operative interventions was ascertained. ·Almost 80 % of respondents would be happy to assess the feasibility of implementing a POPC-CB. Delphi: The preferred number of components in the POPC care bundle was 7. Care bundle elements supported by the consensus process were: ·PRE-operatively: a supervised exercise programme and inspiratory muscle training. ·INTRA-operatively: low tidal volume ventilation with individualised PEEP, use of routine monitoring to avoid hyperoxia and efforts made to limit neuromuscular blockade. ·POST-operatively: deep breathing exercises and mandatory elevation of the head of the bed.

·PRE-operatively: a supervised exercise programme and inspiratory muscle training. ·INTRA-operatively: low tidal volume ventilation with individualised PEEP, use of routine monitoring to avoid hyperoxia and efforts made to limit neuromuscular blockade. ·POST-operatively: deep breathing exercises and mandatory elevation of the head of the bed. Conclusions: A POPC care bundle has been suggested for use, in addition to enhanced recovery pathways, in major surgical patients at high-risk of POPC. Prospective evaluation of feasibility, before an evaluation of effectiveness, is now indicated.

is a robust predictor of mortality following discharge, hospital readmission, and discharge to a care facility. Eosinopenia is a marker for ICU survivors at an especially high risk for adverse outcomes. Thus, patients with eosinopenia may benefit from closer post discharge follow-up and higher intensity rehabilitation. Nutritional aspects in the ICU A494 Protein intake, nutritional status and outcomes in ICU survivors P.J.M. Weijs1,2, K.M. Mogensen3, J.D. Rawn4, M.K. Robinson4, K.B. Christopher5,6 1VU University Medical Center Amsterdam, Department of Nutrition and Dietetics, Internal Medicine, Amsterdam, Netherlands; 2Amsterdam University of Applied Sciences, Amsterdam, Netherlands; 3Brigham and Women's Hospital, Department of Nutrition, Boston, USA; 4Brigham and Women's Hospital, Department of Surgery, Boston, USA; 5Brigham And Women's Hospital, Renal Division, Boston, USA; 6Brigham and Women's Hospital, Channing Division of Network Medicine, Boston, USA Correspondence: P.J.M. Weijs - Amsterdam University of Applied Sciences, Amsterdam, Netherlands Introduction: Critical illness is marked by hypermetabolism and increased protein catabolism. While studies suggest that protein delivery may be beneficial for critical illness outcomes, to date, limited information exists regarding the association between protein delivery during hospitalization and outcomes in ICU survivors following hospital discharge. Objectives: We hypothesized that higher protein intake might have a protective effect in patients with malnutrition.

·POST-operatively: deep breathing exercises and mandatory elevation of the head of the bed. Conclusions: A POPC care bundle has been suggested for use, in addition to enhanced recovery pathways, in major surgical patients at high-risk of POPC. Prospective evaluation of feasibility, before an evaluation of effectiveness, is now indicated. A486 The development and impact of implementation of a quality improvement care pathway for patients undergoing an emergency laparotomy J. Doyle1, P. Wilkerson2, Y. Soon2, S. Huddart3, M. Dickinson3, A. Riga4, A. Zuleika5 1Department of Intensive Care Medicine and Surrey Peri-Operative Anaesthesia and Critical Care Collaborative Research Group (SPACER), Guildford, UK; 2Department of Surgery, Royal Surrey County Hospital NHS Foundation Trust, Guildford, UK; 3Department of Anaesthesia, Royal Surrey County Hospital NHS Foundation Trust, Guildford, UK; 4Department of Surgery, Royal Surrey County Hospital NHS Foundation Trust, Guilford, UK; 5Department of Intensive Care Medicine and Surrey Peri-Operative Anaesthesia and Critical Care Collaborative Research Group (SPACER) and Department of Anaesthesia, Guildford, UK Correspondence: J. Doyle - Department of Intensive Care Medicine and Surrey Peri-Operative Anaesthesia and Critical Care Collaborative Research Group (SPACER), Guildford, UK Introduction: 80 % of UK in-hospital surgical mortality occurs in those undergoing high-risk surgical intervention such as emergency laparotomy (EL) (1) these patients are among the largest consumers of hospital resources (2, 3). Furthermore it was suggested that UK outcomes are worse than comparable countries (4).

UK Introduction: 80 % of UK in-hospital surgical mortality occurs in those undergoing high-risk surgical intervention such as emergency laparotomy (EL) (1) these patients are among the largest consumers of hospital resources (2, 3). Furthermore it was suggested that UK outcomes are worse than comparable countries (4). In the 2011 review of peri-operative care of high-risk patients NCEPOD reported care was good in less than half of patients leading to considerable health and financial costs (5). As such there was a clear need for a change in the approach to patients having an EL. Since this study the first national emergency laparotomy audit (NELA) has been published demonstrating a 30-day mortality of 11 % (6). Objectives: The aims were to create a focused, evidence-based Emergency Laparotomy Pathway (ELP) delivered by a multidisciplinary (MDT) team, utilising care bundles to improve delivery of care and assess the impact of the pathway on hospital mortality and length of stay (LOS). Methods Pathway design: A Emergency Laparotomy Group (ELG) was convened and designed the ELP based on principles of best basic practice and the recommendations of the RCS (2) and NCEPOD (5). The ELP covered the entire surgical stay of the patient in 4 stages. Study protocol: Data was collected for consecutive patients, 50 pre and 96 post-ELP during 2011/2012. The primary outcome measure was in hospital mortality. Secondary outcomes were ICU mortality, ICU and hospital LOS.

Pathway design: A Emergency Laparotomy Group (ELG) was convened and designed the ELP based on principles of best basic practice and the recommendations of the RCS (2) and NCEPOD (5). The ELP covered the entire surgical stay of the patient in 4 stages. Study protocol: Data was collected for consecutive patients, 50 pre and 96 post-ELP during 2011/2012. The primary outcome measure was in hospital mortality. Secondary outcomes were ICU mortality, ICU and hospital LOS. Results: The in-hospital mortality fell from 21.7 % Pre-ELP to 9.6 % Post-ELP (p = 0.048). There is a strong trend towards significance in a survival analysis (p = 0.053). There was no significant difference in hospital LOS between Pre and Post cohorts. (Median 15.5 days vs 15 days, p = 0.875). There was no difference in the ICU LOS (3 vs 2 days, p = 0.3339), nor in the utilisation of high level beds. Conclusion: The introduction of an ELP reduced the in hospital mortality significantly, this has since been further evidenced with a 4-centre (including our own) study demonstrating a case-mix adjusted decrease risk of death from 15.6 % to 9.6 %, p = 0.002 (7). ELP is a useful tool reducing the risk of failure to deliver optimal care. References 1. Pearse RM. Identification and characterisation of the high-risk surgical population in the UK. Crit care. 2006;10(3):R81 2. Anderson ID. Report on Peri-op Care of the High Risk General Surgical Patient. 2011 3. Clarke A. Mortality and postop care after EL. Eur J of Anaes. 2011;28(1):16

Conclusion: The introduction of an ELP reduced the in hospital mortality significantly, this has since been further evidenced with a 4-centre (including our own) study demonstrating a case-mix adjusted decrease risk of death from 15.6 % to 9.6 %, p = 0.002 (7). ELP is a useful tool reducing the risk of failure to deliver optimal care. References 1. Pearse RM. Identification and characterisation of the high-risk surgical population in the UK. Crit care. 2006;10(3):R81 2. Anderson ID. Report on Peri-op Care of the High Risk General Surgical Patient. 2011 3. Clarke A. Mortality and postop care after EL. Eur J of Anaes. 2011;28(1):16 4. Bennett-Guerrero E. Comparison of P-POSSUM risk-adjusted mortality rates after surgery between patients in the USA and UK. BJS. 2003;90(12):1593 5. Findlay GP. NCEPOD Knowing the Risk. review of the peri-op care of surgical patients. 2011 6. NELA project team. First patient report of NELA. RCoA 2015. 7. Huddart S. Use of a pathway quality improvement care bundle to reduce mortality after EL. BJS. 2015;102(1):57 Grant Nil received

4. Bennett-Guerrero E. Comparison of P-POSSUM risk-adjusted mortality rates after surgery between patients in the USA and UK. BJS. 2003;90(12):1593 5. Findlay GP. NCEPOD Knowing the Risk. review of the peri-op care of surgical patients. 2011 6. NELA project team. First patient report of NELA. RCoA 2015. 7. Huddart S. Use of a pathway quality improvement care bundle to reduce mortality after EL. BJS. 2015;102(1):57 Grant Nil received A487 Dexmedetomidine for ventilated septic patients in ICU: a multicenter randomized controlled trial K. Miyamoto1, Y. Kawazoe2, T. Morimoto3, T. Yamamoto4, A. Fuke5, A. Hashimoto6, H. Koami7, S. Beppu8, Y. Katayama9, M. Ito10, Y. Ohta11, H. Yamamura12, DESIRE (DExmedetomidine for Sepsis in ICU Randomized Evaluation) Trial Investigators 1Wakayama Medical University, Department of Emergency and Critical Care Medicine, Wakayama, Japan; 2Tohoku University Hospital Emergency Center, Division of Emergency and Critical Care Medicine, Sendai, Japan; 3Hyogo College of Medicine, Department of Clinical Epidemiology, Nishinomiya, Japan; 4Osaka City University Graduate School of Medicine, Department of Trauma and Critical Care Medicine, Osaka, Japan; 5Osaka City General Hospital, Emergency and Urgent Medical Care Center, Osaka, Japan; 6Hyogo College of Medicine, Emergency and Critical Care Center, Nishinomiya, Japan; 7Saga University Hospital, Advanced Emergency and Critical Care Center, Saga, Japan; 8National Hospital Organization Kyoto Medical Center, Department of Emergency Medicine, Critical Care, Kyoto, Japan; 9Sapporo Medical University, Department of Emergency Medicine, Sapporo, Japan; 10Yamaguchi Grand Medical Center, Department of Anesthesiology, Yamaguchi, Japan; 11Hyogo College of Medicine, Division of General Medicine, Department of Internal Medicine, Nishinomiya, Japan; 12Hirosaki University Graduate School of Medicine, Department of Disaster and Emergency Medicine, Hirosaki, Japan Correspondence: K. Miyamoto - Wakayama Medical University, Department of Emergency and Critical Care Medicine, Wakayama, Japan Introduction: Dexmedetomidine was shown to improve 28-day mortality compared with lorazepam in a subgroup analysis of sepsis patients in the MENDS randomized controlled trial1).

Hirosaki, Japan Correspondence: K. Miyamoto - Wakayama Medical University, Department of Emergency and Critical Care Medicine, Wakayama, Japan Introduction: Dexmedetomidine was shown to improve 28-day mortality compared with lorazepam in a subgroup analysis of sepsis patients in the MENDS randomized controlled trial1). Objectives: This study aimed to demonstrate whether a sedation strategy using dexmedetomidine was superior to that without dexmedetomidine in septic patients, in terms of 28-day mortality and ventilator-free days within a 28-day period. Methods: A multicenter randomized controlled trial was conducted to evaluate the dexmedetomidine-based strategy in septic patients in eight ICUs. We included adult septic patients who were expected to require mechanical ventilation for at least 24 hours. Patients were randomized to receive sedation strategy either with dexmedetomidine (DEX group) or without dexmedetomidine (non-DEX group). Additionally, fentanyl, propofol, and midazolam were used as required to achieve the sedation goal in both groups which was set to Richmond Agitation Sedation Scale 0 (calm) for daytime and −2 (light sedation) for nighttime.

on strategy either with dexmedetomidine (DEX group) or without dexmedetomidine (non-DEX group). Additionally, fentanyl, propofol, and midazolam were used as required to achieve the sedation goal in both groups which was set to Richmond Agitation Sedation Scale 0 (calm) for daytime and −2 (light sedation) for nighttime. Results: We included 201 patients (mean age 68.8 years), of whom 127 (63 %) were men. Mean Acute Physiology And Chronic Health Evaluation (APACHE) II score was 23. The baseline characteristics were similar between groups. In the DEX group, well-controlled sedation rate was higher (p = 0.002) and delirium rate was lower (p = 0.03) during ventilation compared with those in the non-DEX group. At 28 days, 19 of 100 patients (19 %) in the DEX group and 28 of 101 patients (28 %) in the non-DEX group had died (p = 0.14). The ventilator-free days in a 28-day period did not differ between the groups, at 22 days (interquartile range [IQR], 17–25) and 21 days (IQR, 15.5-24.5) (p = 0.49). In a predefined subgroup analysis that included 104 patients with APACHE II score of 23 or more, 28-day mortality in the DEX group was significantly lower than that in the non-DEX group (odds ratio 0.40, 95 % confidential interval 0.17-0.93).

s (interquartile range [IQR], 17–25) and 21 days (IQR, 15.5-24.5) (p = 0.49). In a predefined subgroup analysis that included 104 patients with APACHE II score of 23 or more, 28-day mortality in the DEX group was significantly lower than that in the non-DEX group (odds ratio 0.40, 95 % confidential interval 0.17-0.93). Conclusions: A sedation strategy using dexmedetomidine compared with a strategy without dexmedetomidine induced well-controlled sedation in ventilated septic patients but did not significantly improve 28-day mortality and the number of ventilator-free days. Although more critically ill patients may benefit from a dexmedetomidine strategy, further studies are needed to confirm this result. References Pandharipande PP et al. Crit Care 2010; 14: R38. Grant acknowledgement The design of this study was supported in part by Hospira Japan who also provided a non-contractual research grant to Wakayama Medical University for start-up purposes. However, data collection, data management, statistical analyses, interpretation and manuscript preparation were conducted solely by the academic investigators without any support from pharmaceutical companies. The Institute for Clinical Effectiveness, a non-profit academic research organization, supported the implementation of this study.

The design of this study was supported in part by Hospira Japan who also provided a non-contractual research grant to Wakayama Medical University for start-up purposes. However, data collection, data management, statistical analyses, interpretation and manuscript preparation were conducted solely by the academic investigators without any support from pharmaceutical companies. The Institute for Clinical Effectiveness, a non-profit academic research organization, supported the implementation of this study. A488 Lower versus higher haemoglobin threshold for blood transfusion in septic shock: exploratory subgroup analyses of a randomised trial S.L. Rygård1, L.B. Holst1, J. Wetterslev2, P.I. Johansson3, A. Perner1 1University of Copenhagen, Rigshospitalet, Department of Intensive Care, København, Denmark; 2Copenhagen Trial Unit, Center for Clinical Intervention Research, Copenhagen, Denmark; 3University of Copenhagen, Rigshospitalet, Section for Transfusion Medicine, Copenhagen, Denmark Correspondence: S.L. Rygård - University of Copenhagen, Rigshospitalet, Department of Intensive Care, København, Denmark Introduction: Using a restrictive transfusion strategy appears to be safe in critically ill patients [1,2] but there may be subgroups of patients benefiting from transfusion at a higher haemoglobin level [3]. Objectives: We explored if subgroups of patients with septic shock had worse outcome when transfused at a lower versus a higher haemoglobin threshold.

A488 Lower versus higher haemoglobin threshold for blood transfusion in septic shock: exploratory subgroup analyses of a randomised trial S.L. Rygård1, L.B. Holst1, J. Wetterslev2, P.I. Johansson3, A. Perner1 1University of Copenhagen, Rigshospitalet, Department of Intensive Care, København, Denmark; 2Copenhagen Trial Unit, Center for Clinical Intervention Research, Copenhagen, Denmark; 3University of Copenhagen, Rigshospitalet, Section for Transfusion Medicine, Copenhagen, Denmark Correspondence: S.L. Rygård - University of Copenhagen, Rigshospitalet, Department of Intensive Care, København, Denmark Introduction: Using a restrictive transfusion strategy appears to be safe in critically ill patients [1,2] but there may be subgroups of patients benefiting from transfusion at a higher haemoglobin level [3]. Objectives: We explored if subgroups of patients with septic shock had worse outcome when transfused at a lower versus a higher haemoglobin threshold. Methods: By performing post-hoc analyses of the whole trial population of 998 patients from the Transfusion Requirements in Septic Shock (TRISS) trial [2], we investigated the intervention effect in patients with severe co-morbidity (chronic lung disease, haematological malignancy or metastatic cancer), patients who had undergone surgery (either elective or acute) and patients with septic shock as defined by the new definition of septic shock: lactate above 2 mmol/l and treatment with vasopressors [4].

n effect in patients with severe co-morbidity (chronic lung disease, haematological malignancy or metastatic cancer), patients who had undergone surgery (either elective or acute) and patients with septic shock as defined by the new definition of septic shock: lactate above 2 mmol/l and treatment with vasopressors [4]. Results: The baseline characteristics were mostly similar between the 2 intervention groups in the different subgroups. There were no differences in the intervention effect on 90-day mortality in patients with chronic lung disease (test of interaction P = 0.31), haematological malignancy (P = 0.47), metastatic cancer (P = 0.51), septic shock by the new definition (P = 0.20) or in those who had undergone surgery (P = 0.99), see Table. Conclusions: In exploratory analyses of a randomised trial in patients with septic shock, we observed no mortality benefit in any subgroups of transfusion at a haemoglobin threshold of 90 g/dl vs. 70 g/dl. References 1. Forminsky E, Putzu A, Monaco F et al. Liberal transfusion strategy improves survival in perioperative but not in critically ill patients. A meta-analysis of randomized trials BJA 2015, 115: 511–19 2. Holst LB, Haase N, Wetterslev J, et al. Lower versus higher hemoglobin threshold for transfusion in septic shock N Engl J Med 2014, 371: 1381–91 3. Vincent JL, Which carries the biggest risk: Aneamia or blood transfusion? Transfusion Clinique et Biologique 2015, 22, 148–50 4. Singer M, Deutschman CS, Seymour CW et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3) JAMA 2016, 315:801–10

2. Holst LB, Haase N, Wetterslev J, et al. Lower versus higher hemoglobin threshold for transfusion in septic shock N Engl J Med 2014, 371: 1381–91 3. Vincent JL, Which carries the biggest risk: Aneamia or blood transfusion? Transfusion Clinique et Biologique 2015, 22, 148–50 4. Singer M, Deutschman CS, Seymour CW et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3) JAMA 2016, 315:801–10 Grant acknowledgement None.Fig. 31 (abstract A488). Results

3. Vincent JL, Which carries the biggest risk: Aneamia or blood transfusion? Transfusion Clinique et Biologique 2015, 22, 148–50 4. Singer M, Deutschman CS, Seymour CW et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3) JAMA 2016, 315:801–10 Grant acknowledgement None.Fig. 31 (abstract A488). Results Factors affecting long-term ICU outcome A489 Long-term ICU prognosis: how well do subjective ICU physician prognoses comply with the actual observed one-year outcomes? I.W. Soliman1, D.W. de Lange1, D. van Dijk1, J.J.M. van Delden2, O.L. Cremer1, A.J.C. Slooter1, L.M. Peelen1,3 1University Medical Center Utrecht, Department of Intensive Care, Utrecht, Netherlands; 2Julius Center for Health Sciences and Primary Care University Medical Center Utrecht, Department of Medical Humanities, Utrecht, Netherlands; 3Julius Center for Health Sciences and Primary Care University Medical Center Utrecht, Department of Epidemiology, Utrecht, Netherlands Correspondence: I.W. Soliman - University Medical Center Utrecht, Department of Intensive Care, Utrecht, Netherlands Introduction: At discharge from the Intensive Care Unit (ICU) physicians may integrate information of a patient's condition before ICU admission (medical history, functional status, quality of life) with the sequence of events during the ICU stay to estimate the prognosis of future well-being [1,2]. In 2006, the Sabadell group developed a simple but subjective prognostic score which made this prognosis explicit [1,2]. However, the Sabadell score has not been validated beyond hospital survival or for long-term health related quality of life (HRQoL).

the ICU stay to estimate the prognosis of future well-being [1,2]. In 2006, the Sabadell group developed a simple but subjective prognostic score which made this prognosis explicit [1,2]. However, the Sabadell score has not been validated beyond hospital survival or for long-term health related quality of life (HRQoL). Objectives: To compare ICU physicians' subjective long-term prognosis with the observed long-term survival and HRQoL. Methods: The first admissions of all patients discharged alive from the ICU of the University Medical Center Utrecht between March 2012 and December 2014 were included in the study. The Sabadell score was determined and recorded by the treating physician at the time of ICU discharge ('0 - Good prognosis', '1 - Poor long-term prognosis (>6 months) with unlimited ICU readmission', '2 - Poor short-term prognosis (<6 months); ICU readmission debatable', or '3 - Death expected during hospitalisation, ICU readmission not recommended'). Survival status was verified using the municipal registry one year after ICU discharge. In one-year ICU survivors HRQoL was measured using the EuroQoL 5D-3LTM index score, dichotomized at 0.4 into high or low HRQoL. The analyses included agreement between Sabadell score and one-year outcome categories, and a description of correctly and incorrectly estimated prognosis groups.

one year after ICU discharge. In one-year ICU survivors HRQoL was measured using the EuroQoL 5D-3LTM index score, dichotomized at 0.4 into high or low HRQoL. The analyses included agreement between Sabadell score and one-year outcome categories, and a description of correctly and incorrectly estimated prognosis groups. Results: Of the 4,874 eligible patients, 3,796 (77.9 %) had completed follow-up and were included into the study (see Fig. 32). In-hospital mortality was 3.1 %, 9.3 %, 39.1 % and 81.6 % for the Sabadell scores 0–3 respectively. In 2,849/3796 (75.1 %) the Sabadell score accurately predicted one-year outcome (Fig. 33). Table 15 shows patient characteristics for the study population, and the groups where prognosis was over- and underestimated respectively. Compared to patients with a correct estimate of prognosis, both patient groups with over- or underestimated prognoses were more often admitted for non-surgical reasons, and more severely ill at admission and during ICU stay. Conclusions: In three out of four patients, the Sabadell score was in accordance with one-year prognosis. Only in less than 10 % of patients surviving ICU there was major disagreement between a physician's prediction and the observed long-term outcome. References 1. Fernandez R et al. Crit Care. 2006;10(6):R179 2. Fernandez R et al. Intensive Care Med. 2010;36(7):1196–201 Grant acknowledgement This study was supported by the NutsOhra Foundation, project nr 1404–013Fig. 32 (abstract A489). Flowchart Fig. 33 (abstract A489). Agreement Table 15 (abstract A489) Patient characteristics

References 1. Fernandez R et al. Crit Care. 2006;10(6):R179 2. Fernandez R et al. Intensive Care Med. 2010;36(7):1196–201 Grant acknowledgement This study was supported by the NutsOhra Foundation, project nr 1404–013Fig. 32 (abstract A489). Flowchart Fig. 33 (abstract A489). Agreement Table 15 (abstract A489) Patient characteristics Total population Prognosis correctly estimated Prognosis overestimated Prognosis underestimated p-value N 3,796 2,849 (75.1%)a 765 (20.2%)a 182 (4.8%)a -- Sex (male) 2,405 (63.4%) 1,833 (64.3%) 459 (60.0%) 113 (62.1%) .081 Age at ICU admission 65 (55–73) 65 (55–72) 67 (58–75) 65 (53–73) <.001 Admission type - <.001 Elective surgical 2,317 (61.1%) 1,951 (68.5%) 302 (39.6%) 64 (35.4%) Urgent surgical 593 (15.6%) 401 (14.1%) 160 (21.0%) 32 (17.7%) Medical 881 (23.2%) 495 (17.4%) 301 (39.4%) 85 (47.0%) APACHE IV score 44 (33–60) 41 (31–54) 55 (41–76) 58 (42–81) <.001 Total maximum SOFA scoreb 5 (3–8) 5 (3–8) 6 (4–10) 8 (4–11) <.001 SOFA score at discharge 4 (3–6) 4 (3–6) 4 (3–6) 4 (3–6) <.001 ICU length of stay 1 (1–2) 1 (1–2) 1 (1–4) 3 (1–8) <.001 Continuous variables are presented as median (interquartile range) and tested for differences using a Kruskal Wallis test, categorical variables are presented as n (percentage) and tested for differences using a Chi-square test ICU intensive care unit, LoS length of stay, APACHE IV acute physiology and chronic health evaluation fourth edition, SOFA sequential organ failure assessment apercentage reflects proportion of total sample size bSum of highest SOFA component scores during admission

Total population Prognosis correctly estimated Prognosis overestimated Prognosis underestimated p-value N 3,796 2,849 (75.1%)a 765 (20.2%)a 182 (4.8%)a -- Sex (male) 2,405 (63.4%) 1,833 (64.3%) 459 (60.0%) 113 (62.1%) .081 Age at ICU admission 65 (55–73) 65 (55–72) 67 (58–75) 65 (53–73) <.001 Admission type - <.001 Elective surgical 2,317 (61.1%) 1,951 (68.5%) 302 (39.6%) 64 (35.4%) Urgent surgical 593 (15.6%) 401 (14.1%) 160 (21.0%) 32 (17.7%) Medical 881 (23.2%) 495 (17.4%) 301 (39.4%) 85 (47.0%) APACHE IV score 44 (33–60) 41 (31–54) 55 (41–76) 58 (42–81) <.001 Total maximum SOFA scoreb 5 (3–8) 5 (3–8) 6 (4–10) 8 (4–11) <.001 SOFA score at discharge 4 (3–6) 4 (3–6) 4 (3–6) 4 (3–6) <.001 ICU length of stay 1 (1–2) 1 (1–2) 1 (1–4) 3 (1–8) <.001 Continuous variables are presented as median (interquartile range) and tested for differences using a Kruskal Wallis test, categorical variables are presented as n (percentage) and tested for differences using a Chi-square test ICU intensive care unit, LoS length of stay, APACHE IV acute physiology and chronic health evaluation fourth edition, SOFA sequential organ failure assessment apercentage reflects proportion of total sample size bSum of highest SOFA component scores during admission A490 Impact of an enhanced rehabilitation quality improvement project on long term survival of mechanically ventilated patients D. McWilliams1, C. Snelson2 1University Hospitals Birmingham NHS Trust, Critical Care, Birmingham, UK; 2University Hospital Birmingham, Critical Care, Birmingham, UK Correspondence: C. Snelson - University Hospital Birmingham, Critical Care, Birmingham, UK Introduction: Early and enhanced rehabilitation of patients in critical care is associated with improved patient outcomes (1). We have previously published the results of a quality improvement project promoting early and enhanced rehabilitation for patients mechanically ventilated for greater than 5 days, and shown an improvement in mobility levels at ICU discharge that was associated with a significant reduction in ICU and hospital length of stay, ventilator days and in-hospital mortality (2). The impact of such mobility programs on longer term outcomes is unknown.

ients mechanically ventilated for greater than 5 days, and shown an improvement in mobility levels at ICU discharge that was associated with a significant reduction in ICU and hospital length of stay, ventilator days and in-hospital mortality (2). The impact of such mobility programs on longer term outcomes is unknown. Objectives: To assess the 3 year mortality of patients mechanically ventilated for more than 5 days and who survived to ICU discharge pre and post a quality improvement project enhancing rehabilitation on critical care. Methods: The study took place within a large, tertiary referral ICU. A supportive rehabilitation team was created within the ICU in April 2012 with a focus on promoting early and enhanced rehabilitation for patients at high risk for prolonged ICU and hospital length of stays. The intervention has been described in detail in a previous publication, and was successful in improving mobility levels at ICU discharge. For all patients who survived to ICU discharge, mortality at 3 years was assessed. Results: The mean age of the total court was 53 years. A total of 201 patients in the pre-quality improvement (Pre QI) group and 222 patients in the post quality improvement (Post QI) were discharged from ICU alive. The 3 year mortality rate was 81/201 (40.3 %) in the Pre QI group vs 60/222 (27 %) in the Post QI group; p = 0.004.

e mean age of the total court was 53 years. A total of 201 patients in the pre-quality improvement (Pre QI) group and 222 patients in the post quality improvement (Post QI) were discharged from ICU alive. The 3 year mortality rate was 81/201 (40.3 %) in the Pre QI group vs 60/222 (27 %) in the Post QI group; p = 0.004. Conclusions: The introduction of early and enhanced rehabilitation within the ICU was associated with a significant reduction in 3 year mortality of ICU survivors. This hypothesis generating data warrants further evaluation in future randomized controlled trials of the intervention. References 1. Hashem, M et al. Early mobilisation and rehabilitation of the critically ill patient. Chest available online March 2016 2. McWilliams et al. (2015) Enhancing rehabilitation of mechanically ventilated patients in the intensive care unit: a quality improvement project. J of Crit Care 30; 8–13 Grant acknowledgement We are grateful to Queen Elizabeth Hospital Birmingham Charity for their support with this project

1. Hashem, M et al. Early mobilisation and rehabilitation of the critically ill patient. Chest available online March 2016 2. McWilliams et al. (2015) Enhancing rehabilitation of mechanically ventilated patients in the intensive care unit: a quality improvement project. J of Crit Care 30; 8–13 Grant acknowledgement We are grateful to Queen Elizabeth Hospital Birmingham Charity for their support with this project A491 Epidemiological, clinical characteristics and determinants of 1-year mortality after ICU discharge. The Caviuci study (evaluation of quality of life after ICU in Argentine) A. Das Neves1, C.I. Loudet1, M. Busico2, D. Vazquez3, D. Villalba4, M. Veronesi5, A. Lischinsky6, F.J.L. López7, L. Benito Mori8, G. Plotnikow3, A. Díaz9, S. Giannasi10, R. Hernandez11, L. Krzisnik12, C. Cecotti13, L. Viola14, R. Lopez15, J.P. Sottile16, G. Benavent17, E. Estenssoro1 1HIGA Gral San Martin, La Plata, Argentina; 2Hospital de Trauma F Abete, Municipio Islas Malvinas, Buenos Aires, Argentina; 3Sanatorio Anchorena, Buenos Aires, Argentina; 4Hospital Municipal Chivilcoy, Chivilcoy, Argentina; 5Instituto Fleni, Escobar, Argentina; 6Instituto De Neurología Cognitiva-INECO, Buenos Aires, Argentina; 7Hospital Escuela de Agudos Dr. Ramón Madariaga, Posadas, Argentina; 8HIGA Prof. Dr. Luis Güemes, Haedo, Argentina; 9Hospital Misericordia, Cordoba, Argentina; 10Hospital Italiano Buenos Aires, Buenos Aires, Argentina; 11Hospital Francisco López Lima, Gral Roca, Argentina; 12Alta Complejidad en Red, Hospital El Cruce, Dr. Néstor Carlos Kirchner, Florencio Varela, Argentina; 13HZGA Mi Pueblo Florencio Varela, Florencio Varela, Argentina; 14Hospital Español, Mendoza, Argentina; 15Clinica Pueyrredon, Mar del Plata, Argentina; 16Hospital Zonal Bariloche, Bariloche, Argentina; 17Clinica CEMEP, Ushuaia, Argentina Correspondence: A. Das Neves - HIGA Gral San Martin, La Plata, Argentina Introduction: Admission to the ICU deeply impacts on post-ICU mortality and quality of life, but the information on the issue is still scarce(1).

ata, Argentina; 16Hospital Zonal Bariloche, Bariloche, Argentina; 17Clinica CEMEP, Ushuaia, Argentina Correspondence: A. Das Neves - HIGA Gral San Martin, La Plata, Argentina Introduction: Admission to the ICU deeply impacts on post-ICU mortality and quality of life, but the information on the issue is still scarce(1). Objectives: To analyze epidemiological data and evolution of patients on mechanical ventilation (MV) in the ICU and to define independent variables associated to long-term (1 year) mortality. Methods: National, prospective, multicenter cohort study sponsored by the Argentinian Society of Intensive Care Medicine, including patients receiving MV >72 hr (3/5-8/31/2014). Epidemiological and evolution variables were recorded, and after hospital discharge patients/their relatives were interviewed face-to-face or by telephone at 2, 6 and 12 months. Comparisons were made between survivors (S) and non-survivors (NS). Data are analyzed according to their nature; a p < .05 was considered significant. Independent predictors of post-ICU mortality were identified by logistic regression. Results: 320 patients from 26 hospitals across Argentina were included. The data are shown on Tables 16, 17 and 18. Global post-ICU mortality was 22 %; respectively 15 %, 5 % and 2 % at 2,6, and 12 months post hospital discharge. Independent predictors of mortality were: age (OR 1.07[.04-1.08;]; p 0.000) and SOFA24hs (OR 1.12[1.02-.1.23];p 0.013). Calibration and discrimination of the model were adequate (GOF 0.407; AUROC 0.79, respectively). Conclusions:

Global post-ICU mortality was 22 %; respectively 15 %, 5 % and 2 % at 2,6, and 12 months post hospital discharge. Independent predictors of mortality were: age (OR 1.07[.04-1.08;]; p 0.000) and SOFA24hs (OR 1.12[1.02-.1.23];p 0.013). Calibration and discrimination of the model were adequate (GOF 0.407; AUROC 0.79, respectively). Conclusions: 1. Most patients were middle-aged men, admitted to public hospitals for medical causes, 60 % with ≥1 comorbidities and very acutely ill, according to APACHE II and SOFA scores. 2. Complications as ARDS, shock and infections were frequent, occurring in more than a half of patients. 3. 1-year mortality was 22 %. Most patients died within two months after hospital discharge. Yet no single ICU event was associated with post-ICU mortality; only age and organ failures were independent predictors of this outcomes. Grant acknowledgement This study was supported by de Argentinian Society of Intensive Care (SATI).Table 16 (abstract A491). Characteristics of Patients

3. 1-year mortality was 22 %. Most patients died within two months after hospital discharge. Yet no single ICU event was associated with post-ICU mortality; only age and organ failures were independent predictors of this outcomes. Grant acknowledgement This study was supported by de Argentinian Society of Intensive Care (SATI).Table 16 (abstract A491). Characteristics of Patients All Survivors Non Survivors p Age 50[29–66] 41[25–60] 66[53–74] 0.000 Male gender 184(58%) 118(57%) 59(41%) 0.79 Marital status Married/cohabiting Single Widower 153(48%)/ 133(43%)/ 34(11%) 46%/ 47%/ 8% 53%/ 29%/ 18% 0.005 Widower vs single P<0.001 Diagnostic category Medical Scheduled surgery Emergency surgery Trauma Trauma + Cranial trauma 51%/ 17%/ 8%/ 8%/ 15% 47%/ 18%/ 8%/ 9%/ 19% 66%/ 13%/ 9%/ 7%/ 6% 0.025 Occupational status Occupied (any type of employment/study) 51% 62% 40% 0.000 Years of study 9±4 9±4 9±4 0.45 Health Insurance 60% 57% 70% 0.05 Admission to public hospitals 65% 68% 56% 0.06 Transferred from Emergency Ward Home Operating room Another Hospital 49%/ 15%/ 1%/ 18%/ 17% 49%/ 13%/ 1%/ 19%/ 18% 49%/ 21%/ 3%/ 16%/ 11% 0.20 Table 17 (abstract A491). Severity Scores All Survivivors Non survivors p Charlson score 0 1 ≥ 2 118(37%) 72(22%)/ 130(41%)/ 43% 21%/ 36%/ 21% 31%/ 46%/ 0.008 APACHE II 18±7 17±7 20±8 0.004 SOFA24hs 7±4 6±3 8±3 0.007 Admission Glasgow 14[8–15] 14[9–15] 14[8–15] 0.64 Table 18 (abstract A491). Events during ICU Stay

All Survivors Non Survivors p Age 50[29–66] 41[25–60] 66[53–74] 0.000 Male gender 184(58%) 118(57%) 59(41%) 0.79 Marital status Married/cohabiting Single Widower 153(48%)/ 133(43%)/ 34(11%) 46%/ 47%/ 8% 53%/ 29%/ 18% 0.005 Widower vs single P<0.001 Diagnostic category Medical Scheduled surgery Emergency surgery Trauma Trauma + Cranial trauma 51%/ 17%/ 8%/ 8%/ 15% 47%/ 18%/ 8%/ 9%/ 19% 66%/ 13%/ 9%/ 7%/ 6% 0.025 Occupational status Occupied (any type of employment/study) 51% 62% 40% 0.000 Years of study 9±4 9±4 9±4 0.45 Health Insurance 60% 57% 70% 0.05 Admission to public hospitals 65% 68% 56% 0.06 Transferred from Emergency Ward Home Operating room Another Hospital 49%/ 15%/ 1%/ 18%/ 17% 49%/ 13%/ 1%/ 19%/ 18% 49%/ 21%/ 3%/ 16%/ 11% 0.20 Table 17 (abstract A491). Severity Scores All Survivivors Non survivors p Charlson score 0 1 ≥ 2 118(37%) 72(22%)/ 130(41%)/ 43% 21%/ 36%/ 21% 31%/ 46%/ 0.008 APACHE II 18±7 17±7 20±8 0.004 SOFA24hs 7±4 6±3 8±3 0.007 Admission Glasgow 14[8–15] 14[9–15] 14[8–15] 0.64 Table 18 (abstract A491). Events during ICU Stay All Survivors Non Survivors p Shock at admission 51% 49% 59% 0.15 Shock on evolution 57% 55% 63% 0.27 Duration of shock (days) 3[2–6] 3[2–5] 3[2–4] 0.3 ARDS 45% 48% 36% 0.07 Ventilator-associated pneumonia 26% 29% 19% 0.09 Catheter-related infection 41% 40% 43% 0.71 Length of MV(days) 10[6–19] 10[6–19] 10[6–19] 0.85 ICU stay (days) 16[10–26] 15[9–24] 18[11–27] 0.07 Hospital stay (days) 29[17–52] 28[16–52] 42[21–57] 0.06

uration of shock (days) 3[2–6] 3[2–5] 3[2–4] 0.3 ARDS 45% 48% 36% 0.07 Ventilator-associated pneumonia 26% 29% 19% 0.09 Catheter-related infection 41% 40% 43% 0.71 Length of MV(days) 10[6–19] 10[6–19] 10[6–19] 0.85 ICU stay (days) 16[10–26] 15[9–24] 18[11–27] 0.07 Hospital stay (days) 29[17–52] 28[16–52] 42[21–57] 0.06 A492 The outcomes of patients with severe dengue admitted to intensive care units C.-M. Chen1,2, C.-C. Lai3, K.-C. Cheng4, W. Chou1, K.-S. Chan2 1Chia Nan University of Pharmacy and Science, Recreation and Health-Care Management, Tainan, Taiwan, Province of China; 2Chi-Mei Medical Center, Intensive Care Medicine, Tainan, Taiwan, Province of China; 3Chi-Mei Medical Center, Liouying District, Intensive Care Medicine, Tainan, Taiwan, Province of China; 4Chi Mei Medical Center, Internal Medicine, Tainan, Taiwan, Province of China Correspondence: C.-M. Chen - Chi-Mei Medical Center, Intensive Care Medicine, Tainan, Taiwan, Province of China Introduction: To survey the outcome of patients with dengue infection admitted to intensive care unit (ICU). Objectives: Because the outcome of adult patients with dengue infections requiring ICU admissions remains unclear, this study was conducted to assess the clinical manifestations and the prognostic factors of critical ill patients with severe dengue.

A492 The outcomes of patients with severe dengue admitted to intensive care units C.-M. Chen1,2, C.-C. Lai3, K.-C. Cheng4, W. Chou1, K.-S. Chan2 1Chia Nan University of Pharmacy and Science, Recreation and Health-Care Management, Tainan, Taiwan, Province of China; 2Chi-Mei Medical Center, Intensive Care Medicine, Tainan, Taiwan, Province of China; 3Chi-Mei Medical Center, Liouying District, Intensive Care Medicine, Tainan, Taiwan, Province of China; 4Chi Mei Medical Center, Internal Medicine, Tainan, Taiwan, Province of China Correspondence: C.-M. Chen - Chi-Mei Medical Center, Intensive Care Medicine, Tainan, Taiwan, Province of China Introduction: To survey the outcome of patients with dengue infection admitted to intensive care unit (ICU). Objectives: Because the outcome of adult patients with dengue infections requiring ICU admissions remains unclear, this study was conducted to assess the clinical manifestations and the prognostic factors of critical ill patients with severe dengue. Methods: This study was conducted in a tertiary referral hospital of 96 adult ICU beds. In this retrospective study, all of the patients with laboratory-confirmed severe dengue infections and admitted to ICU were enrolled between July 31 and November 31 in 2015 during the large outbreak period. The medical records of all the recruited patients were retrospectively reviewed and the following information was collected: age, gender, clinical manifestations, disease severity scores, underlying conditions, laboratory examinations, and outcome. The primary endpoint is to find the predictors of in-hospital mortality.

he medical records of all the recruited patients were retrospectively reviewed and the following information was collected: age, gender, clinical manifestations, disease severity scores, underlying conditions, laboratory examinations, and outcome. The primary endpoint is to find the predictors of in-hospital mortality. Results: During study period, there were a total of 4787 patients with dengue infections, and 143 (2.99 %) patients with severe dengue infection required ICU admission. Among 143 critically ill patients, their mean age was 69.7 years. Hypertension (n = 90, 62.9 %) and diabetes mellitus (n = 70, 49.0 %) were the two most common underlying diseases. Eighty patients (55.9 %) had co-bacterial infections, and 33 patients had co-bacteremia. Hematologic system was the most common failure organ, followed by chest and cardiovascular systems. Fever was the most common presentation (n = 112, 78.3 %), followed by anorexia (n = 47, 32.9 %), and abdominal pain (n = 46, 32.2 %). Overall, a total of 33 patients had mortality, and the rate of mortality was 23.1 %. In multivariate analysis, we found that in-hospital mortality was significantly associated with lower Glasgow coma scale, lower platelet count before ICU discharge, and higher number of organ failures. Conclusions: The outcome of severe dengue patients requiring ICU admission remains high and the mortality was associated with lower Glasgow coma scale, lower platelet counts and more organ failures. Additionally, more than half of patients would have bacterial infections during severe dengue infections. References

Results: During study period, there were a total of 4787 patients with dengue infections, and 143 (2.99 %) patients with severe dengue infection required ICU admission. Among 143 critically ill patients, their mean age was 69.7 years. Hypertension (n = 90, 62.9 %) and diabetes mellitus (n = 70, 49.0 %) were the two most common underlying diseases. Eighty patients (55.9 %) had co-bacterial infections, and 33 patients had co-bacteremia. Hematologic system was the most common failure organ, followed by chest and cardiovascular systems. Fever was the most common presentation (n = 112, 78.3 %), followed by anorexia (n = 47, 32.9 %), and abdominal pain (n = 46, 32.2 %). Overall, a total of 33 patients had mortality, and the rate of mortality was 23.1 %. In multivariate analysis, we found that in-hospital mortality was significantly associated with lower Glasgow coma scale, lower platelet count before ICU discharge, and higher number of organ failures. Conclusions: The outcome of severe dengue patients requiring ICU admission remains high and the mortality was associated with lower Glasgow coma scale, lower platelet counts and more organ failures. Additionally, more than half of patients would have bacterial infections during severe dengue infections. References 1.Murray NE, Quam MB, Wilder-Smith A: Epidemiology of dengue: past, present and future prospects. Clinical epidemiology 2013, 5:299–309. 2.Guzman MG, Harris E: Dengue. Lancet (London, England) 2015, 385(9966):453–465. Grant acknowledgement no

Conclusions: The outcome of severe dengue patients requiring ICU admission remains high and the mortality was associated with lower Glasgow coma scale, lower platelet counts and more organ failures. Additionally, more than half of patients would have bacterial infections during severe dengue infections. References 1.Murray NE, Quam MB, Wilder-Smith A: Epidemiology of dengue: past, present and future prospects. Clinical epidemiology 2013, 5:299–309. 2.Guzman MG, Harris E: Dengue. Lancet (London, England) 2015, 385(9966):453–465. Grant acknowledgement no A493 Eosinopenia in ICU survivors and post-hospital outcomes L.E. Roeker1, C.M. Horkan1, F.K. Gibbons2, K.B. Christopher3,4 1Brigham and Women's Hospital, Department of Medicine, Boston, MA, USA; 2Massachusetts General Hospital, Pulmonary and Critical Care Medicine, Boston, MA, USA; 3Brigham and Women's Hospital, Renal Division, Boston, MA, USA; 4Brigham and Women's Hospital, Channing Division of Network Medicine, Boston, MA, USA Correspondence: L.E. Roeker - Brigham and Women's Hospital, Department of Medicine, Boston, MA, USA Introduction: Eosinopenia is associated with short term adverse outcomes in the critically ill. In survivors of critical care, it is not know if eosinopenia is predictive of adverse outcomes following hospital discharge. Objectives: We hypothesized that eosinopenia at ICU admission would be associated with increased hospital readmission rates and higher mortality following hospital discharge.

A493 Eosinopenia in ICU survivors and post-hospital outcomes L.E. Roeker1, C.M. Horkan1, F.K. Gibbons2, K.B. Christopher3,4 1Brigham and Women's Hospital, Department of Medicine, Boston, MA, USA; 2Massachusetts General Hospital, Pulmonary and Critical Care Medicine, Boston, MA, USA; 3Brigham and Women's Hospital, Renal Division, Boston, MA, USA; 4Brigham and Women's Hospital, Channing Division of Network Medicine, Boston, MA, USA Correspondence: L.E. Roeker - Brigham and Women's Hospital, Department of Medicine, Boston, MA, USA Introduction: Eosinopenia is associated with short term adverse outcomes in the critically ill. In survivors of critical care, it is not know if eosinopenia is predictive of adverse outcomes following hospital discharge. Objectives: We hypothesized that eosinopenia at ICU admission would be associated with increased hospital readmission rates and higher mortality following hospital discharge. Methods: We performed a two center observational study of patients treated in medical and surgical intensive care units. We studied 68,648 patients, age ≥ 18 years, who received critical care between 1998 and 2012 and survived hospitalization. The exposure of interest was absolute eosinophil count within 48 hours of ICU admission and categorized a priori as ≤10/ul, 10-50/ul, 50-350/ul (normal range) and >350/ul. The primary outcome was all cause mortality in the 90 days following hospital discharge determined using the US Social Security Administration Death Master File. 365-day follow-up was present in all cohort patients. Adjusted odds ratios were estimated by multivariable logistic regression models with inclusion of covariate terms for age, race, gender, Deyo-Charlson Index, patient type (medical versus surgical), sepsis and number of organs with acute failure.

n Death Master File. 365-day follow-up was present in all cohort patients. Adjusted odds ratios were estimated by multivariable logistic regression models with inclusion of covariate terms for age, race, gender, Deyo-Charlson Index, patient type (medical versus surgical), sepsis and number of organs with acute failure. Results: The cohort patients were 57.8 % male, 20.3 % nonwhite and 48.8 % surgical. 10.1 % of the cohort had sepsis, and the mean age was 61.7 years. Median [IQR] absolute eosinophil count was 90 [30–190]. 90-day post discharge mortality was 7.7 %. 30-day readmission rate was 14.3 %. A decreased absolute eosinophil count was a robustly associated with all-cause post discharge mortality as well as hospital readmission and discharge to a care facility instead of home. Patients with an absolute eosinophil count of ≤10/ul or 10-50/ul have an adjusted OR of 90-day post-discharge mortality of 1.58 (95%CI, 1.46-1.71; P < 0.001) or 1.49 (95%CI, 1.38-1.60; P < 0.001) relative to patients with an absolute eosinophil count of 50-350/ul. Patients with an absolute eosinophil count of ≤10/ul or 10-50/ul have an adjusted OR of 30-day readmission of 1.08 (95%CI, 1.01-1.15; P = 0.018) or 1.10 (95%CI, 1.04-1.17; P = 0.001) relative to patients with an absolute eosinophil count of 50-350/ul. Finally, patients with an absolute eosinophil count of ≤10/ul or 10-50/ul have an adjusted OR of discharge to a care facility of 1.34 (95%CI, 1.28-1.41; P < 0.001) or 1.30 (95%CI, 1.24-1.36; P < 0.001) relative to patients with an absolute eosinophil count of 50-350/ul. An absolute eosinophil count >350/ul was not associated with study outcomes.

Nutritional aspects in the ICU A494 Protein intake, nutritional status and outcomes in ICU survivors P.J.M. Weijs1,2, K.M. Mogensen3, J.D. Rawn4, M.K. Robinson4, K.B. Christopher5,6 1VU University Medical Center Amsterdam, Department of Nutrition and Dietetics, Internal Medicine, Amsterdam, Netherlands; 2Amsterdam University of Applied Sciences, Amsterdam, Netherlands; 3Brigham and Women's Hospital, Department of Nutrition, Boston, USA; 4Brigham and Women's Hospital, Department of Surgery, Boston, USA; 5Brigham And Women's Hospital, Renal Division, Boston, USA; 6Brigham and Women's Hospital, Channing Division of Network Medicine, Boston, USA Correspondence: P.J.M. Weijs - Amsterdam University of Applied Sciences, Amsterdam, Netherlands Introduction: Critical illness is marked by hypermetabolism and increased protein catabolism. While studies suggest that protein delivery may be beneficial for critical illness outcomes, to date, limited information exists regarding the association between protein delivery during hospitalization and outcomes in ICU survivors following hospital discharge. Objectives: We hypothesized that higher protein intake might have a protective effect in patients with malnutrition. Methods: We performed a single center observational study of patients treated in medical intensive care units in Boston, Massachusetts. We studied 801 patients age ≥ 18 years, who received critical care following between 2004 and 2011 and survived to hospital discharge. All patients underwent a Registered Dietitian formal assessment within 48 hours of ICU admission. The exposure of interest, grams of protein per kilogram body weight delivered per day, was determined from all oral, enteral and parenteral sources for up to 28 days. Nutrition status was categorized as non-specific malnutrition, protein-energy malnutrition, or at risk for malnutrition via anthropometric measurements, clinical signs of malnutrition, malnutrition risk factors, and metabolic stress. The primary outcome was all cause 90-day post-discharge mortality. Adjusted odds ratios were estimated by mixed- effects logistic regression models to describe how 90-day post-discharge mortality differed with changes in protein delivery.

clinical signs of malnutrition, malnutrition risk factors, and metabolic stress. The primary outcome was all cause 90-day post-discharge mortality. Adjusted odds ratios were estimated by mixed- effects logistic regression models to describe how 90-day post-discharge mortality differed with changes in protein delivery. Results: The cohort was 55 % male, 79 % white with a mean age of 62.3 years. 22 % of the cohort had sepsis, 10 % had acute kidney injury and 35 % had non-cardiac acute respiratory failure. 59 % had non-specific malnutrition or protein-energy malnutrition. The 30, 90 and 365-day post-discharge mortality was 7.1 and 13.9 and 24.4 %. The average number of nutrition delivery days recorded was 15. In a mixed-effect logistic regression model adjusted for age, gender, race, Deyo-Charlson Index, acute organ failures, sepsis and percent energy needs met, 90-day post-discharge mortality rate was 17 % (95%CI: 6–26) lower for each 1 g/kg increase in daily protein delivery (P = 0.002) compared with the 90-day post-discharge mortality rate in the entire cohort [OR = 0.83 (95%CI 0.74-0.94; P = 0.002)]. In a subset of patients with malnutrition (n = 473): 90-day post-discharge mortality rate was 30 % (95%CI: 6–26) lower for each 1 g/kg increase in daily protein delivery (P < 0.001) compared with the 90-day post-discharge mortality rate in the entire cohort [OR = 0.70 (95%CI 0.61-0.81; P < 0.001)].

4-0.94; P = 0.002)]. In a subset of patients with malnutrition (n = 473): 90-day post-discharge mortality rate was 30 % (95%CI: 6–26) lower for each 1 g/kg increase in daily protein delivery (P < 0.001) compared with the 90-day post-discharge mortality rate in the entire cohort [OR = 0.70 (95%CI 0.61-0.81; P < 0.001)]. Conclusions: In adult MICU patients who survive to hospital discharge, protein intake appears to be predictive of out of hospital outcomes. Patients with improvements in protein intake during hospitalization independent of energy intake have decreased mortality in the 3 months following hospital discharge. The achievement of ideal protein delivery may be an important factor in ICU survivorship, especially in malnutrition. A495 Safety and efficacy of a new parenteral lipid emulsion (SMOFlipid) in surgical critically ill patients Z. Tang, C. Qiu, B. Ouyang, C. Cai, X. Guan The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou, China Correspondence: Z. Tang - The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou, China Introduction: SMOFlipid 20 % is intravenous lipid emulsion (ILE) containing long-chain triglycerides (LCT), medium-chain triglycerides(MCT), olive oil, and fish oil as a mixed emulsion containing α-tocopherol. The aim was to assess the efficacy of this new ILE in surgery compared with MCT/LCT,and it was tested for safety, tolerance, metabolic and clinical efficacy in surgical patients. Objectives: To assess the efficacy of this new ILE in gastrointestinal surgery compared with MCT/LCT.

A495 Safety and efficacy of a new parenteral lipid emulsion (SMOFlipid) in surgical critically ill patients Z. Tang, C. Qiu, B. Ouyang, C. Cai, X. Guan The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou, China Correspondence: Z. Tang - The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou, China Introduction: SMOFlipid 20 % is intravenous lipid emulsion (ILE) containing long-chain triglycerides (LCT), medium-chain triglycerides(MCT), olive oil, and fish oil as a mixed emulsion containing α-tocopherol. The aim was to assess the efficacy of this new ILE in surgery compared with MCT/LCT,and it was tested for safety, tolerance, metabolic and clinical efficacy in surgical patients. Objectives: To assess the efficacy of this new ILE in gastrointestinal surgery compared with MCT/LCT. Methods: In this prospective study, 42 patients were randomized to SMOFlipid 20 % or MCT/LCT (Lipovenoes 20 %) group. Clinical and biochemistry data were collected. Inflammatory markers (CRP, IL-6) and liver function indicators (ALT,AST,TBIL)were measured.

Objectives: To assess the efficacy of this new ILE in gastrointestinal surgery compared with MCT/LCT. Methods: In this prospective study, 42 patients were randomized to SMOFlipid 20 % or MCT/LCT (Lipovenoes 20 %) group. Clinical and biochemistry data were collected. Inflammatory markers (CRP, IL-6) and liver function indicators (ALT,AST,TBIL)were measured. Results: 32 patients (17 males and 15 females) with a mean age of 51 years completed the study. The patients' baseline datas(age, gender, APACHE II) were similar in two groups(p > 0.05). The patients' baseline datas(age, gender, APACHE II)were similar in two groups.The increment of triglyceride on day 5 from baseline was significantly lower in SMOFlipid group than in Lipovenoes MCT/LCT group(P < 0.05). The concentrations of alanine transaminase (ALT), aspartate transaminase (AST) and totalbilirubin on day 5 were significantly lower in SMOFlipid group than in control group(P < 0.05). Inflammatory markers (CRP, IL-6)were not different between groups(P > 0.05). Conclusions: SMOFlipid had a better triglyceride-lowering effectas compared with MCT/LCT and may be associated with a better liver tolerance in surgical critically ill. References 1. Burrin DG, Ng K, Stoll B, et al. Impact of new-generation lipid emulsions on cellular mechanisms of parenteral nutrition-associated liverdisease. Adv Nutr, 2014,5(1):82–91.

Conclusions: SMOFlipid had a better triglyceride-lowering effectas compared with MCT/LCT and may be associated with a better liver tolerance in surgical critically ill. References 1. Burrin DG, Ng K, Stoll B, et al. Impact of new-generation lipid emulsions on cellular mechanisms of parenteral nutrition-associated liverdisease. Adv Nutr, 2014,5(1):82–91. 2. Wu MH, Wang MY, Yang CY,et al.Randomized clinical trial of new intravenous lipid (SMOFlipid 20 %) versus medium-chain triglycerides/long-chaintriglycerides in adult patients undergoing gastrointestinal surgery,JPEN J Parenter Enteral Nutr, 2014,38(7):800–8.

1. Burrin DG, Ng K, Stoll B, et al. Impact of new-generation lipid emulsions on cellular mechanisms of parenteral nutrition-associated liverdisease. Adv Nutr, 2014,5(1):82–91. 2. Wu MH, Wang MY, Yang CY,et al.Randomized clinical trial of new intravenous lipid (SMOFlipid 20 %) versus medium-chain triglycerides/long-chaintriglycerides in adult patients undergoing gastrointestinal surgery,JPEN J Parenter Enteral Nutr, 2014,38(7):800–8. A496 Dexmethasone-induced muscular atrophy is mediated by the expression of functional connexin based hemichannels T. Regueira1, L. Cea2, S. Juan Carlos3, B. Elisa3, C. Puebla3, A. Vargas3 1Clínica Las Condes, Medicina Intensiva, Santiago, Chile; 2Universidad de Chile, Program of Anatomy and Developmental Biology, Institute of Biomedical Sciences, Santiago, Chile; 3Universidad Católica de Chile, Departamento de Fisiología, Santiago, Chile Correspondence: T. Regueira - Clínica Las Condes, Medicina Intensiva, Santiago, Chile Introduction: Muscle atrophy is one of the most important and frequent problems observed in patients in Intensive Care Units (ICU). Several mechanisms and risk factors have been described to explain this syndrome. Glucocorticoids, which are frequently used in ICU, induce muscle atrophy, but the mechanism still remains to be elucidated. Connexin based hemichannels (Cx HCs), are channels formed by six connexin proteins and communicates the intra with extra cellular space since they are permeable to ions and small molecules (e.g., ATP). Recently, CxHCs have been associated to denervation-induced skeletal muscle atrophy (1), and may also participate in the pathogenesis of corticoids induced muscle atrophy.

rmed by six connexin proteins and communicates the intra with extra cellular space since they are permeable to ions and small molecules (e.g., ATP). Recently, CxHCs have been associated to denervation-induced skeletal muscle atrophy (1), and may also participate in the pathogenesis of corticoids induced muscle atrophy. Objectives: To evaluate the possible involvement of Cx HCs in dexamethasone-induced muscular atrophy. Methods: C57Bl/6 wild type (WT) mice and Cx HCs knock out (KO) mice (Cx43fl/flCx45fl/fl and Cx43fl/flCx45fl/fl:Myo-Cre) mice were used. KO mice were skeletal muscle deficient for connexin 43 and 45. Also, Lanthanum, a connexin blocker, was used in experiments. Presence and functionality of Cx HCs was evaluated acutely after 5 hours of dexamethasone exposure, by cellular Ethidium uptake, Evans blue permeability, intracellular free Ca2+, resting membrane potential, immunohistochemistry, and western blot. NFkb Phosphorylation and mRNA levels of TNF-alpha were also assessed. Finally, muscle atrophy was evaluated by cross sectional area (CSA) after 7 days of dexamethasone exposure. Also, atrogin-1 and MurF-1, which are two E3 ligases from ubiquitin-proteasome system (intracellular degradation system), levels were evaluated.

kb Phosphorylation and mRNA levels of TNF-alpha were also assessed. Finally, muscle atrophy was evaluated by cross sectional area (CSA) after 7 days of dexamethasone exposure. Also, atrogin-1 and MurF-1, which are two E3 ligases from ubiquitin-proteasome system (intracellular degradation system), levels were evaluated. Results: Freshly isolated myofibers from mice treated with dexamethasone by 5 h showed de novo expression of functional connexins 39, 43, 45 and pannexins (P2X7R) which were mainly distributed on sarcolemma. Cx HCs expression was present only in WT myofibers, was blocked by La, and was absent in KO mice. Accordingly, a significant increase in basal intracellular free Ca2+ levels, and a decrease of resting membrane potential of about 10 mV, was observed in myofibers from WT mice treated with dexamethasone compared with mice injected with saline buffer or in KO mice. Moreover, dexamethasone induced the phosphorylation of NFkb and increased the mRNA levels of TNF-alpha. Finally, seven days treatment with dexamethasone drastically reduced the CSA of myofibers from WT mice but not KO animals. Both, atrogin-1 and MurF-1 were elevated in WT animals in comparison with KO mice. Conclusions: Dexamethasone induces the expression of Ca2+ permeable channels such as connexins, which leads to muscle atrophy. Connexins are upstream to the activation of the ubiquitin-proteasome system. Connexins are a novel family of proteins which should be studied for prevention of corticoids-induced muscle atrophy. References 1. Cea et al. Proc Natl Acad Sci USA 2013,110(40). Grant acknowledgement

Conclusions: Dexamethasone induces the expression of Ca2+ permeable channels such as connexins, which leads to muscle atrophy. Connexins are upstream to the activation of the ubiquitin-proteasome system. Connexins are a novel family of proteins which should be studied for prevention of corticoids-induced muscle atrophy. References 1. Cea et al. Proc Natl Acad Sci USA 2013,110(40). Grant acknowledgement Fondecyt 1141092 A497 Comparison of two indirect calorimetry devices at varying levels of inspired oxygen M.K. Poulsen1, L.P. Thomsen1, S. Kjærgaard2, S.E. Rees1, D.S. Karbing1 1Aalborg University, Respiratory and Critical Care Group (Rcare), Department of Health Science and Technology, Aalborg, Denmark; 2Aalborg University Hospital, Department of Anesthesiology, Aalborg, Denmark Correspondence: S.E. Rees - Aalborg University, Respiratory and Critical Care Group (Rcare), Department of Health Science and Technology, Aalborg, Denmark Introduction: Malnutrition increases morbidity, length of hospitalization and mortality in ICU patients (1), stressing the importance of energy expenditure (EE) assessment. Indirect calorimetry is the most accurate method, but available calorimeters do not support EE measurement at inspired oxygen (FiO2) >80 %, preventing measurement in the most severely ill. The Beacon Caresystem (Mermaid Care, Nørresundby, Denmark) for advising on mechanical ventilation includes indirect calorimetry designed for use in the full FiO2 range, but has not been compared to other calorimeters.

pport EE measurement at inspired oxygen (FiO2) >80 %, preventing measurement in the most severely ill. The Beacon Caresystem (Mermaid Care, Nørresundby, Denmark) for advising on mechanical ventilation includes indirect calorimetry designed for use in the full FiO2 range, but has not been compared to other calorimeters. Objectives: To compare the Beacon Caresystem to the E-COVX (Datex-Ohmeda, GE Health Care, Helsinki, Finland) at varying FiO2. Methods: Twenty healthy male subjects were included. Informed consent and ethical approval were obtained in all cases. Measurements from 16 subjects (age 32 ± 12 yrs) were analysed, the remainder excluded due to gas leakage. Subjects were mechanically ventilated by mask (Servo-I, Maquet, Solna, Sweden) in pressure support (PS: 3 cm H2O, PEEP: 3 cm H2O) at FiO2 of 21, 50 and 85 %, in that order, for 15 min at each FiO2. Analysis periods were selected where COV for average oxygen consumption (VO2) and CO2 production (VCO2) were < 5 % in five consecutive 1 min intervals (2). VO2, VCO2 and EE were compared between devices (Beacon - E-COVX) at each FiO2 by Bland-Altman analysis of bias and limits of agreement (LOA) and within devices between 21 and 85 % FiO2 by paired t-tests.

V for average oxygen consumption (VO2) and CO2 production (VCO2) were < 5 % in five consecutive 1 min intervals (2). VO2, VCO2 and EE were compared between devices (Beacon - E-COVX) at each FiO2 by Bland-Altman analysis of bias and limits of agreement (LOA) and within devices between 21 and 85 % FiO2 by paired t-tests. Results: VO2, VCO2 and EE bias [LOA] between devices were −27 [−73 - 19] ml/min, −5 [−28 - 18] ml/min, and −157 [−434 - 121] kcal/day at 21 % FiO2; −5 [−50 - 50] ml/min, 6 [−22 - 34] ml/min, and −12 [−345 - 320] kcal/day at 50 % FiO2; and 26 [−87 - 139] ml/min, 50 [16–84] ml/min, and 327 [−179 - 834] kcal/day at 85 % FiO2. Analysis was performed in 12 subjects at 85 % FiO2 as E-COVX was unable to measure EE in 4 subjects. Mean ± SD changes in VO2, VCO2 and EE within device from 21 to 85 % as percentage of mean were 11 ± 6 % (p < 0.001), 3 ± 6 % (p = 0.11) and 9 ± 6 % (p < 0.001) for Beacon Caresystem, and −4 ± 14 % (p = 0.415), −16 ± 10 % (p < 0.001) and −9 ± 12 % (p = 0.017) for E-COVX. Conclusions: Measurements of VO2, VCO2 and EE were comparable between Beacon Caresystem and E-COVX at 21 and 50 % FiO2. Significant differences were observed at 85 % FiO2 where E-COVX slightly underestimated VCO2 and EE and Beacon Caresystem slightly overestimated VO2 and EE. Only Beacon Caresystem allowed EE measurement at 85 % FiO2 in all 16 subjects. References 1. Goiburu ME et al. The impact of malnutrition on morbidity, mortality and length of hospital stay in trauma patients. Nutr Hosp. 2006;21(5):604–10.

Conclusions: Measurements of VO2, VCO2 and EE were comparable between Beacon Caresystem and E-COVX at 21 and 50 % FiO2. Significant differences were observed at 85 % FiO2 where E-COVX slightly underestimated VCO2 and EE and Beacon Caresystem slightly overestimated VO2 and EE. Only Beacon Caresystem allowed EE measurement at 85 % FiO2 in all 16 subjects. References 1. Goiburu ME et al. The impact of malnutrition on morbidity, mortality and length of hospital stay in trauma patients. Nutr Hosp. 2006;21(5):604–10. 2. Frankenfield D et al. Validation of a 5-minute steady state indirect calorimetry protocol for resting energy expenditure in critically ill patients. J Am Coll Nutr. 1996;15(4):397–402. Grant acknowledgement DSK and SER are minor shareholders and perform consultancy for Mermaid Care.

1. Goiburu ME et al. The impact of malnutrition on morbidity, mortality and length of hospital stay in trauma patients. Nutr Hosp. 2006;21(5):604–10. 2. Frankenfield D et al. Validation of a 5-minute steady state indirect calorimetry protocol for resting energy expenditure in critically ill patients. J Am Coll Nutr. 1996;15(4):397–402. Grant acknowledgement DSK and SER are minor shareholders and perform consultancy for Mermaid Care. A498 Measuring resting energy expenditure in patients with veno-venous extracorporeal membrane oxygenation as a necessary tool to guarantee goal directed feeding (MEEP) T. Wollersheim1,2, S. Frank1, M.C. Müller1, N.M. Carbon1, V. Skrypnikov1, P.A. Pickerodt1, R. Falk1, A. Mahlau1, S. Weber-Carstens1,2 1Charité - Universitaetsmedizin Berlin, Berlin, Germany; 2Berlin Institute of Health (BIH), Berlin, Germany Correspondence: T. Wollersheim - Berlin Institute of Health (BIH), Berlin, Germany Introduction: Acute respiratory distress syndrome (ARDS) is a common lung disease. Extracorporeal lung support (ECLS) can be an additional lifesaving tool. Patients with severe ARDS often suffer from a catabolic state with massive loss of weight including skeletal muscle and fat tissue. For ICU patients in general there is evidence, that goal directed feeding could improve outcome and indirect calorimetry (IC) is the gold standard to measure resting energy expenditure (REE). Due to the O2 uptake and CO2 removal by lung and ECLS, an IC in the common way is not feasible. So far it is unclear if and which impact ECLS has on caloric needs of patients with ARDS.

directed feeding could improve outcome and indirect calorimetry (IC) is the gold standard to measure resting energy expenditure (REE). Due to the O2 uptake and CO2 removal by lung and ECLS, an IC in the common way is not feasible. So far it is unclear if and which impact ECLS has on caloric needs of patients with ARDS. Objectives: We introduce our MEEP (Measuring Energy Expenditure in ECMO Patients) protocol, which enables us to determine the REE in patients with ECLS. We investigated the impact of ECLS in ARDS patients on REE. Methods: We perform a common IC and extend it by a calculation of the O2 uptake and the CO2 elimination by the ECMO filter due to blood gas analyses (BGAs) and the ECMO blood flow [1]. Sum O2 uptake and CO2 elimination were used in the equation of Weir [2] to calculate REE. With this protocol we performed a monocentric, controlled, prospective, observational pilot study. We included 20 patients with ARDS and ECMO treatment and 20 matched ARDS patients without ECLS as control. Informed consent was given by legal proxy. The COSMED Quark RMR® was used to perform IC, BGAs were done by ABL Flex 800 and filter blood flow was measured by the ECMO device. For all patients we measured resting energy expenditure and calculated the most prevalent predicting equations for REE for comparison. Nonparametric tests were performed. Data were shown as median [IQR]. P-values ≤ 0.05 are assumed as being significant. Ethic vote(EA1/293/13).

lood flow was measured by the ECMO device. For all patients we measured resting energy expenditure and calculated the most prevalent predicting equations for REE for comparison. Nonparametric tests were performed. Data were shown as median [IQR]. P-values ≤ 0.05 are assumed as being significant. Ethic vote(EA1/293/13). Results: Patient´s baseline characteristics were without significant differences except SOFA Score at time-point of measurement for patients with ECMO 12.5 [8.0/14.5] and without ECMO 7.5 [6.0/9.5] (p = 0.002). We report that our MEEP-protocol is safe, easy and feasible to perform in ECMO patients. We show for the first time, that measured REE values according to our MEEP protocol do not significantly differ between ARDS patients with ECMO (2013 kcal/d [1786/2333]) compared to ARDS patients without ECMO (1857 kcal/d [1602/2085]) (p = 0.165). All investigated equations predicting REE do not fit the measured REE according to the MEEP protocol neither in non-ECMO patients (data not shown) nor in ECMO patients. Conclusions: Due to our MEEP-protocol REE is easy measurable in ARDS patients with ECLS. ECMO per se does not to influence REE. Equations to estimate caloric needs do not fit the REE of critically ill patients with ECMO. We recommend to implement sequential measurements of REE in critically ill during ECMO treatment to improve goal directed feeding.Fig. 34 (abstract A498). Resting energy expenditure in ECMO patients

e does not to influence REE. Equations to estimate caloric needs do not fit the REE of critically ill patients with ECMO. We recommend to implement sequential measurements of REE in critically ill during ECMO treatment to improve goal directed feeding.Fig. 34 (abstract A498). Resting energy expenditure in ECMO patients Trauma critical care A499 CRASH-2 from the coal face - experience from a major trauma centre in the United Kingdom A. Lee, R. Inglis, R. Morgan, G. Barker John Radcliffe Hospital, Adult Intensive Care Unit, Oxford, UK Correspondence: A. Lee - John Radcliffe Hospital, Adult Intensive Care Unit, Oxford, UK Introduction: The administration of tranexamic acid to trauma patients with risk of haemorrhage is one of only seven treatments shown to improve mortality in critically ill patients in multi-centre randomized controlled trials [1,2]. Nonetheless, four years on from CRASH-2, the implementation of this simple, low risk intervention, which has the potential to save 600 lives a year in the UK, remains suboptimal [3]. Objectives: To measure the proportion of trauma patients deemed to be at risk of haemorrhage (as evidenced by the administration of a first dose of tranexamic acid) that correctly received the second dose of tranexamic acid. We designed a quality improvement project to improve compliance with this measure, and here we report preliminary Results:

proportion of trauma patients deemed to be at risk of haemorrhage (as evidenced by the administration of a first dose of tranexamic acid) that correctly received the second dose of tranexamic acid. We designed a quality improvement project to improve compliance with this measure, and here we report preliminary Results: Methods: All patients admitted under a 'trauma call' between January 2015 and January 2016 at the John Radcliffe Hospital in Oxford, a major trauma centre in the United Kingdom, were included in our study. Tranexamic acid administration information was collected for all patients. Interventions included disseminating tranexamic acid information to trauma doctors verbally and during departmental induction. We also added both doses of tranexamic acid into the electronic prescribing bundle for trauma patients in our hospital. Further interventions are ongoing. Results: The John Radcliffe Hospital received 1373 patients as 'trauma calls' between January 2015 and January 2016. Of the 184 patients who received the first dose of tranexamic acid, 82 patients correctly received the second dose (45 %). We present a run-chart of percentage of patients receiving both doses of tranexamic acid over time, out of all the patients who received the first dose of tranexamic acid.

ry 2015 and January 2016. Of the 184 patients who received the first dose of tranexamic acid, 82 patients correctly received the second dose (45 %). We present a run-chart of percentage of patients receiving both doses of tranexamic acid over time, out of all the patients who received the first dose of tranexamic acid. Conclusions: We have demonstrated that a majority of patients who would have received two doses of tranexamic acid under the terms of CRASH 2 are not receiving this life-saving treatment. This finding highlighted the difficulty in translating evidence-based recommendations into everyday practice and the need for quality improvement projects in this area. Responsibility for the administration of tranexamic acid lies across multiple specialties, and so measures to improve its administration need to target a diverse group of healthcare staff. Engaging stakeholders from all relevant parties has proved challenging and an early focus on inter-specialty collaboration is our key recommendation to centres who intend to implement similar initiatives. References 1. Landoni G et al. Mortality in Multicenter Critical Care Trials: An Analysis of Interventions With a Significant Effect. Crit Care Med. 2015 Aug;43(8):1559–68. 2. Shakur H et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant hemorrhage (CRASH-2): a randomized, placebo-controlled trial. The Lancet. 2010;376(9734):23–32. 3. Trauma Audit and Research Network (TARN) database.Fig. 35 (abstract A499). Run Chart TxA Audit Jan 2015 – Jan 2016

2. Shakur H et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant hemorrhage (CRASH-2): a randomized, placebo-controlled trial. The Lancet. 2010;376(9734):23–32. 3. Trauma Audit and Research Network (TARN) database.Fig. 35 (abstract A499). Run Chart TxA Audit Jan 2015 – Jan 2016 A500 Dynamic vital signs of patients with trauma and in-hospital mortality K. Kamata1, T. Abe2,3, D. Saitoh4, Y. Tokuda5 1Ministry of Health, Labour and Welfare, Government of Japan, Tuberculosis and Infectious Diseases Control, Tokyo, Japan; 2Tsukuba Medical Center Hospital, Emergency Medicine and Critical Care Medicine, Tsukuba, Japan; 3University of Tsukuba, Health Services Research, Tsukuba, Japan; 4National Defense Medical College, Traumatology and Emergency Medicine, Tokorozawa, Japan; 5Japan Community Healthcare Organization, Tokyo, Japan Correspondence: K. Kamata - Ministry of Health, Labour and Welfare, Government of Japan, Tuberculosis and Infectious Diseases Control, Tokyo, Japan Introduction: Emergency physicians are required to assess the severity of patient's illness and predict their prognosis with restricted situation at emergency department (ED). Trauma is one of the most severe injuries at ED and its outcome highly depends on initial appropriate management. Therefore, several scoring systems have been developed on the last decade. However, the environment on ED has changed and physicians have been still occasionally misled.

uation at emergency department (ED). Trauma is one of the most severe injuries at ED and its outcome highly depends on initial appropriate management. Therefore, several scoring systems have been developed on the last decade. However, the environment on ED has changed and physicians have been still occasionally misled. Objectives: Vital signs are widely used in both emergency medical service and ED. It is a role of not only the static parameter but also the dynamic one to help emergency physicians classify trauma patients' severity. Thus, we aimed to analyze the association between changes in vital sings (ΔVS) and in-hospital mortality among trauma patients. Methods: This study was multicenter observational retrospective study from Japan Trauma Data Bank. It consisted of major emergency hospitals in Japan from January 2004 to December 2013. In this period, a total of 159,157 trauma patients were registered. Of these, 75,833 patients (48 %) were analyzed since all vital signs (systolic blood pressure, heart rate, and respiratory rate) were available both at site and at ED. We developed a new scoring system to predict in-hospital mortality for trauma patients using three ΔVS and age. The patients were judged on these 4 items with one point added for each positive one; when systolic blood pressure decreased more than 30 mmHg, heart rate increased more than 20/min, respiratory rate increased 10/min, or old (age ≧65).

system to predict in-hospital mortality for trauma patients using three ΔVS and age. The patients were judged on these 4 items with one point added for each positive one; when systolic blood pressure decreased more than 30 mmHg, heart rate increased more than 20/min, respiratory rate increased 10/min, or old (age ≧65). Results: Mean age (±SD) was 54 (±24) years old. 29,309 (39 %) were aged over 65 years old. 25,854 (34 %) were women. Overall, in-hospital mortality rate was 5,834 (7.7 %). The in-hospital mortality rate increased from 6.4 % to 18.9 % when systolic blood pressure decreased more than 30 mmHg. In a similar way, it increased from 7.0 % to 16.5 % when heart rates increased 20/min. Also, it increased from 7.5 % to 10.7 % when respiratory rates increase 10/min. Old patients (> = 65 years old) had severe outcome than young adults (it increased from 5.1 % to 11.8 %). As using our scoring system, in-hospital mortality rate increased: 0 point was 3.5 %; 1 point was 9.5 %; 2 points was 18.5 %; 3 points was 33.0 %; and 4 points was 48.6 %. The area under the ROC curve of this predictive model was 0.675 (95 % CI 0.668-0.683). Conclusions: Our new trauma scoring system, ΔVS scoring system, which focus on changing vital signs, predicts in-hospital mortality. It is useful to improve that emergency physician's decision and identification of patients who need critical care. References 1. Kondo et al. Critical Care 2011,15:R191 2. Mehmood et al. BMC Emergency Medicine 2015, 15(Suppl 2):S10 3. Gerdin et al. BMC Emergency Medicine(2016)16:15 Grant acknowledgment

Conclusions: Our new trauma scoring system, ΔVS scoring system, which focus on changing vital signs, predicts in-hospital mortality. It is useful to improve that emergency physician's decision and identification of patients who need critical care. References 1. Kondo et al. Critical Care 2011,15:R191 2. Mehmood et al. BMC Emergency Medicine 2015, 15(Suppl 2):S10 3. Gerdin et al. BMC Emergency Medicine(2016)16:15 Grant acknowledgment This work was supported by JSPS KAKENHI Grant Number 16 K15388. A501 Increased mortality in trauma patients who develop post-intubation hypotension R.S. Green1, M.B. Butler2, M. Erdogan1 1Dalhousie University, Nova Scotia Trauma Program, Halifax, Canada; 2Dalhousie University, Undergraduate Medical Education, Halifax, Canada Correspondence: R.S. Green - Dalhousie University, Nova Scotia Trauma Program, Halifax, Canada Introduction: Systemic hypotension is a predictor of increased mortality in patients with traumatic brain injuries. Post-intubation hypotension (PIH) has been demonstrated to be common and associated with poor patient outcomes in other critically ill patient populations requiring emergent endotracheal intubation (ETI). The importance of PIH remains unclear in the trauma population. Objectives: To determine the incidence of PIH in adult trauma patients and assess the association of PIH with mortality.

A501 Increased mortality in trauma patients who develop post-intubation hypotension R.S. Green1, M.B. Butler2, M. Erdogan1 1Dalhousie University, Nova Scotia Trauma Program, Halifax, Canada; 2Dalhousie University, Undergraduate Medical Education, Halifax, Canada Correspondence: R.S. Green - Dalhousie University, Nova Scotia Trauma Program, Halifax, Canada Introduction: Systemic hypotension is a predictor of increased mortality in patients with traumatic brain injuries. Post-intubation hypotension (PIH) has been demonstrated to be common and associated with poor patient outcomes in other critically ill patient populations requiring emergent endotracheal intubation (ETI). The importance of PIH remains unclear in the trauma population. Objectives: To determine the incidence of PIH in adult trauma patients and assess the association of PIH with mortality. Methods: This study was a retrospective case series of adult (≥16 years) trauma patients requiring intubation after referral to a provincial trauma team located in a level 1 center in Halifax, Nova Scotia, Canada between 2000 and 2015. PIH was defined as (a) decrease in systolic blood pressure (SBP) to < 90 mmHg; (b) reduction in SBP of 20 % from baseline; (c) decrease in mean arterial pressure to < 60 mmHg; or (d) initiation of, or increased infusion dosage of, any vasopressor medication (bolus or infusion) during the 15-minute period after ETI. Data was collected from a provincial trauma registry and the patient chart, and included demographics, co-morbidities, trauma characteristics, intubation time, as well as all fluids, medications, adverse events, interventions, and vital signs during the 15 minutes before/after ETI. We evaluated the incidence of PIH and created a logistic regression model to determine the likelihood of mortality in the PIH and non-PIH groups after controlling for gender, age, intubator, mechanism of trauma, injury severity, traumatic brain injury and volume of fluid administered.

g the 15 minutes before/after ETI. We evaluated the incidence of PIH and created a logistic regression model to determine the likelihood of mortality in the PIH and non-PIH groups after controlling for gender, age, intubator, mechanism of trauma, injury severity, traumatic brain injury and volume of fluid administered. Results: Overall, 477 patients arrived unintubated and required ETI by the trauma team over the study period, of which 444 patients met eligibility criteria and were included in the analysis. The incidence of PIH was 35.6 % (158/444) in our study population. In-hospital mortality occurred in 47 (29.7 %) in the PIH group, compared to 45 (15.7 %) in the non-PIH group. After controlling for possible confounding factors, the development of PIH was associated with increased mortality (OR = 2.17; CI: {1.25, 3.77}; P = 0.006). Conclusions: In our study, development of PIH was common (35.6 %) and associated with increased mortality (OR 2.17). Clinicians should attempt to minimize hemodynamic instability during ETI in patients with traumatic injuries. Further investigation of PIH in the trauma population is warranted. References 1. Green RS, Turgeon AF, et al. Postintubation hypotension in intensive care unit patients: A multicenter cohort study. J Crit Care. 2015;30(5):1055–60.

Conclusions: In our study, development of PIH was common (35.6 %) and associated with increased mortality (OR 2.17). Clinicians should attempt to minimize hemodynamic instability during ETI in patients with traumatic injuries. Further investigation of PIH in the trauma population is warranted. References 1. Green RS, Turgeon AF, et al. Postintubation hypotension in intensive care unit patients: A multicenter cohort study. J Crit Care. 2015;30(5):1055–60. A502 Clinical characteristics and injury pattern of geriatric trauma patients H. Tae Hwa, L. Jae Gil Yonsei University College of Medicine, Department of Surgery, Seoul, Republic of Korea Correspondence: H. Tae Hwa - Yonsei University College of Medicine, Department of Surgery, Seoul, Republic of Korea Introduction: Recently, the population of elderly people has been increasing rapidly all over the world. Especially, Republic of Korea is now progressing to an aging society. The social activities of the aging population have increased, which has also increased the number of elderly patients injured. Objectives: The aim of this study is to analyze the clinical characteristics and injury pattern of geritaric patients involved in trauma. Methods: This study was conducted retrospectively from January 2015 to March 2016 among trauma patients who visited single trauma center in Seoul, Korea. The patients divided in two groups, a geriatrics group and an adult group on the basis of an age of 65. Patients under 18 years of age were excluded. The variables related with trauma were extracted and compared.

ely from January 2015 to March 2016 among trauma patients who visited single trauma center in Seoul, Korea. The patients divided in two groups, a geriatrics group and an adult group on the basis of an age of 65. Patients under 18 years of age were excluded. The variables related with trauma were extracted and compared. Results: Total number of the included patients was 384, and the number of elderly patients was 74. There were no significant differences in Revised Trauma Score (RTS), intensive care unit (ICU) admission, hospital stay between the two groups. There were differences in the mechanism and the locations of the injury, gender, injury severity score (ISS), trauma and injury severity score (TRISS), mortality, drinking at admission between the two groups. Most common injury mechanism in the adult group was driver traffic accident (24.5 %) and in the geriatric group was pedestrian traffic accident (45.9 %). Injury of extremity (44.8 % vs 59.5 %) (p = 0.028) was more common in the geriatric group than in the adult group. And female patient was more common in the geriatric group (26.8 % vs 40.5 %) (p = 0.020). The average ISS of the geriatric group was higher than that of the adult group (17.8+/−11.52 vs. 14.0+/−12.06, p = 0.015). The mortality was higher in the geriatric group (14.9 %) than in the adult group (4.5 %) (p = 0.003). But drinking at admission was higher in the adult group (33.9 % vs 10.8 %) (p < 0.00).

0.020). The average ISS of the geriatric group was higher than that of the adult group (17.8+/−11.52 vs. 14.0+/−12.06, p = 0.015). The mortality was higher in the geriatric group (14.9 %) than in the adult group (4.5 %) (p = 0.003). But drinking at admission was higher in the adult group (33.9 % vs 10.8 %) (p < 0.00). Conclusions: The mortality rate and the average ISS were greater within the geriatric group than the adult group despite there was no difference in the RTS, ICU admission. Also elderly people were prone to pedestrian traffic accident with the lower drinking rate compare than the adult group.

0.020). The average ISS of the geriatric group was higher than that of the adult group (17.8+/−11.52 vs. 14.0+/−12.06, p = 0.015). The mortality was higher in the geriatric group (14.9 %) than in the adult group (4.5 %) (p = 0.003). But drinking at admission was higher in the adult group (33.9 % vs 10.8 %) (p < 0.00). Conclusions: The mortality rate and the average ISS were greater within the geriatric group than the adult group despite there was no difference in the RTS, ICU admission. Also elderly people were prone to pedestrian traffic accident with the lower drinking rate compare than the adult group. Prediction models and outcome A503 Sequential sofa score as predictor of outcome in septic patients R. Hernández Vaquero1, E. Rodriguez-Ruiz1, A. Lopez Lago1, J.L. Garcia Allut1, A. Estany Gestal2, M.A. Garcia Gonzalez3 1Hospital Clínico Universitario, Intensive Care Unit, Santiago de Compostela, Spain; 2Fundación Ramón Dominguez, Unidad de Epidemiología e Investigación Clínica, Santiago de Compostela, Spain; 3Instituto de Investigaciones Sanitarias, Grupo de Genética y Biología del Desarrollo de las Enfermedades Renales, Santiago de Compostela, Spain Correspondence: R. Hernández Vaquero - Hospital Clínico Universitario, Intensive Care Unit, Santiago de Compostela, Spain Introduction: Sepsis is a dynamic medical condition where patients´ clinical status can change rapidly in either direction. Different biomarkers and scores have been studied to assess sepsis severity and prognosis. Procalcitonin, C-reactive protein and interleukin-6 are biomarkers referred as helpful to follow evolution and outcomes. APACHE II model was developed to predict hospital outcomes, based on data from the first 24 hours. SOFA score describes the degree of organ dysfunction over time and evaluates morbidity, but has been used also for predicting mortality.

d interleukin-6 are biomarkers referred as helpful to follow evolution and outcomes. APACHE II model was developed to predict hospital outcomes, based on data from the first 24 hours. SOFA score describes the degree of organ dysfunction over time and evaluates morbidity, but has been used also for predicting mortality. Objective: Our aim was to study the ability of different biomarkers (procalcitonin, C-reactive protein and interleukin-6) and scores (APACHE II and SOFA), determined on day 1, to predict outcome in septic patients. Besides, we studied sequential SOFA score to find out its dynamics at early sepsis and its correlation with mortality. Methods: We assessed 48 consecutive septic patients admitted to our medico-surgical ICU. Procalcitonin, C-reactive protein and interleukin-6 were determined on day 1. APACHE II was calculated on day 1, and daily SOFA score was calculated for 5 consecutive days, determining also the maximum SOFA. Mann–Whitney U test and ROC curves were carried out to study association of these variables with 28-day mortality.

ocalcitonin, C-reactive protein and interleukin-6 were determined on day 1. APACHE II was calculated on day 1, and daily SOFA score was calculated for 5 consecutive days, determining also the maximum SOFA. Mann–Whitney U test and ROC curves were carried out to study association of these variables with 28-day mortality. Results: In our cohort of septic patients (mean age 69 ± 13) there were 43 survivors and 5 non-survivors. Procalcitonin, C-reactive protein and interleukin-6 did not show correlation with 28-day mortality. APACHE II, SOFA scores on days 1 and 2, and maximum SOFA, also showed no correlation with 28-day mortality. However, when analyzing SOFA scores on days 3, 4 and 5, we found significant lower scores in survivors compared to non-survivors: SOFA-3 (p = 0.048), SOFA-4 (p = 0.016), and SOFA-5 (p = 0.018). The ROC curve analysis also showed association with 28-day mortality: SOFA-3 (AUC 0.772, p = 0.048), and stronger for SOFA-4 (AUC 0.823, p = 0.019), and SOFA-5 (AUC 0.820, p = 0.021).

e found significant lower scores in survivors compared to non-survivors: SOFA-3 (p = 0.048), SOFA-4 (p = 0.016), and SOFA-5 (p = 0.018). The ROC curve analysis also showed association with 28-day mortality: SOFA-3 (AUC 0.772, p = 0.048), and stronger for SOFA-4 (AUC 0.823, p = 0.019), and SOFA-5 (AUC 0.820, p = 0.021). Conclusion: Day 1 values for procalcitonin, C-reactive protein, interleukin-6, APACHE II and SOFA did not correlate with outcome. But, low SOFA scores on day 3, and specially in days 4 and 5, are good predictors of survival, meaning that after 48 hours of admission, regardless of the initial score, those patients with a low SOFA score have higher probability to survive. While SOFA scores at admission, and maximum SOFA, had bad performance in predicting mortality, sequential SOFA scores add information on the course of the disease, and improve the ability to predict the likelihood of survival. Larger studies are required to find out SOFA based prediction models with improved accuracy. References (1) Biron et al. (2015) Biomarkers for sepsis: what is and what might be? Biomark Insights, 10 (Suppl 4):7–17. (2) Minne et al. (2008) Evaluation of SOFA-based models for predicting mortality in the ICU: A systematic review. Critical Care, 12:R161.

Conclusion: Day 1 values for procalcitonin, C-reactive protein, interleukin-6, APACHE II and SOFA did not correlate with outcome. But, low SOFA scores on day 3, and specially in days 4 and 5, are good predictors of survival, meaning that after 48 hours of admission, regardless of the initial score, those patients with a low SOFA score have higher probability to survive. While SOFA scores at admission, and maximum SOFA, had bad performance in predicting mortality, sequential SOFA scores add information on the course of the disease, and improve the ability to predict the likelihood of survival. Larger studies are required to find out SOFA based prediction models with improved accuracy. References (1) Biron et al. (2015) Biomarkers for sepsis: what is and what might be? Biomark Insights, 10 (Suppl 4):7–17. (2) Minne et al. (2008) Evaluation of SOFA-based models for predicting mortality in the ICU: A systematic review. Critical Care, 12:R161. A504 Development of a sepsis mortality prediction model for use with German hospital claims data D.O. Thomas-Rüddel1,2, D. Schwarzkopf2, C. Fleischmann1,2, K. Reinhart1,2 1Universitaetsklinikum Jena, Department of Anesthesiology and Intensive Care, Jena, Germany; 2Universitaetsklinikum Jena, Center for Sepsis Control and Care (CSCC), Jena, Germany Correspondence: D.O. Thomas-Rüddel - Universitaetsklinikum Jena, Center for Sepsis Control and Care (CSCC), Jena, Germany Introduction: Epidemiological sepsis research in administrative data is widely used in the USA. Risk adjustment models [1] enable adjusted outcome comparisons and benchmarks between regions, hospitals, or over time. For Germany we recently reported first results of an administrative data Approach [2] but a risk model is lacking.

on: Epidemiological sepsis research in administrative data is widely used in the USA. Risk adjustment models [1] enable adjusted outcome comparisons and benchmarks between regions, hospitals, or over time. For Germany we recently reported first results of an administrative data Approach [2] but a risk model is lacking. Objectives: To develop and asses a risk model of hospital mortality for sepsis based on German hospital claims data. Methods: The hospital claims data of almost all inpatients in Germany is combined into one dataset at the Federal Statistical Office. Anonymized statistical analysis can be performed via remote data access. In data from 2010–2013 all adult cases with pathogen-based or clinical sepsis codes (A & R codes, ICD-10) were classified by severity. Age, sex, type of hospital admission, coded foci of infection and comorbidities based on categories of the Charlson and Elixhauser comorbidity indices, coding of sepsis as main diagnosis and year of treatment were assessed as possible risk factors. Predictors of hospital mortality were identified by backwards binary logistic regression in a derivation cohorts separately for each severity group. Performance of models was tested in validation cohorts. Models were then fitted with the selected predictors for the different severities in all cases and predicted mortality risk was calculated. Combined model performance across all severities was assessed by discrimination using the “area under the curve” (AUC), by calibration plotting observed against predicted mortality over risk quantiles, and by explained variance.

tors for the different severities in all cases and predicted mortality risk was calculated. Combined model performance across all severities was assessed by discrimination using the “area under the curve” (AUC), by calibration plotting observed against predicted mortality over risk quantiles, and by explained variance. Results: 441 207 cases with sepsis without organ dysfunction (hospital mortality 13,5 %), 284 883 with severe sepsis (mortality 40,3 %) and 112 609 with septic shock (mortality 59,6 %) were identified. 41–43 of 53 assessed predictors remained in the final models. Overall discrimination (Fig. 36), explained variance (R2 = 0.25), and calibration (Fig. 37) were good. Strongest risk factors were age and advanced liver and hemato-oncologic disease. A urogenital focus and sepsis as main diagnosis were associated with lower hospital mortality. Mortality decreased from 2010 to 2013 for sepsis (p < 0.0001) and severe sepsis (p < 0.0001) but not for septic shock (p = 0.11). Conclusions: We developed a well performing risk adjustment model for sepsis mortality in German administrative data. This enables comparisons and benchmarks from administrative data in Germany. Hospital characteristics will be included into the model in the next development step. References 1. Ford DW et al. Critical care medicine 2016 2. Fleischmann C et al. Deutsches Arzteblatt international 2016 Grant acknowledgement The CSCC is funded by the German Federal Ministry of Education and Research (BMBF), Germany, FKZ: 01EO1502.Fig. 36 (abstract A504). ROC curve of combined model

Conclusions: We developed a well performing risk adjustment model for sepsis mortality in German administrative data. This enables comparisons and benchmarks from administrative data in Germany. Hospital characteristics will be included into the model in the next development step. References 1. Ford DW et al. Critical care medicine 2016 2. Fleischmann C et al. Deutsches Arzteblatt international 2016 Grant acknowledgement The CSCC is funded by the German Federal Ministry of Education and Research (BMBF), Germany, FKZ: 01EO1502.Fig. 36 (abstract A504). ROC curve of combined model Fig. 37 (abstract A504). Predicted and observed hospital mortality A505 Modified early warning score to predict mortality in hospital: a systematic review and meta-analysis S. Suwanpasu1, Y. Sattayasomboon2 1KCMH, Nursing Department, Bangkok, Thailand; 2Mahidol University, Faculty of Public Health, Bangkok, Thailand Correspondence: S. Suwanpasu - KCMH, Nursing Department, Bangkok, Thailand Introduction: Patients at risk of rapid deterioration and critical illness often have preceding changes in their physiological parameters. Objectives: The systematic review quantifies the prognostic accuracy of screening instruments of the Modified Early Warning Score (MEWS) to detect risk of in-hospital mortality.

A505 Modified early warning score to predict mortality in hospital: a systematic review and meta-analysis S. Suwanpasu1, Y. Sattayasomboon2 1KCMH, Nursing Department, Bangkok, Thailand; 2Mahidol University, Faculty of Public Health, Bangkok, Thailand Correspondence: S. Suwanpasu - KCMH, Nursing Department, Bangkok, Thailand Introduction: Patients at risk of rapid deterioration and critical illness often have preceding changes in their physiological parameters. Objectives: The systematic review quantifies the prognostic accuracy of screening instruments of the Modified Early Warning Score (MEWS) to detect risk of in-hospital mortality. Methods: The review process followed guidelines consisting of 5 steps suggested for systematic reviews. Relevant studies from January 2000 to December 2015 were obtained from electronic databases. Standards for Reporting of Diagnostic accuracy (STARD) and Quality Assessment of Diagnostic Accuracy Studies instrument (QUADAS-2) were used to assess individual study quality and bias. MedCalc statistic software was used to assess and combine the data and diagnostic accuracy across studies. Pooled diagnostic odds ratio and area under The ROC Curve (AUC) analysis was used to evaluate the prognostic accuracy of MEWS tool.

dies instrument (QUADAS-2) were used to assess individual study quality and bias. MedCalc statistic software was used to assess and combine the data and diagnostic accuracy across studies. Pooled diagnostic odds ratio and area under The ROC Curve (AUC) analysis was used to evaluate the prognostic accuracy of MEWS tool. Results: A total of 402 citations were identified yielding 16 studies for inclusion in this systematic review. Studies were significantly heterogeneous in terms of age and sample size. For predicting in-hospital death, MEWS with high risk group as cut-off ≥ 4 and ≥ 5 had the Diagnostic odds ratio (DOR) of 14.278 (95 % Confidence Interval [CI] 12.185 to 16.730, I2 = 56.59 %]) and 3.28 (95%CI:2.489 to 4.323, I2 = 48.64 %). On pooled AUC analysis, there was a trend for MEWS to estimate fair at the discriminative power of test. AUC of MEWS ≥ 4 was 0.778 (95%CI:0.715 to 0.841, I2 = 89.54 %) and of MEWS ≥ 5 was 0.646 (95%CI:0.611 to 0.682, I2 = 49.69 %). Conclusions: On Meta-analysis of sixteen of MEWS studies, this showed a robust positive trend to predict in-hospital death. MEWS equal 4 or greater may be the favored tool to alert the deterioration in hospitalized patients. Grant acknowledgement Nursing Department, King Chulalongkorn Memorial Hospital, Thai Red Cross Soceity Note: This abstract has been previously published and is available at [1]. It is included here as a complete record of the abstracts from the conference. References

Conclusions: On Meta-analysis of sixteen of MEWS studies, this showed a robust positive trend to predict in-hospital death. MEWS equal 4 or greater may be the favored tool to alert the deterioration in hospitalized patients. Grant acknowledgement Nursing Department, King Chulalongkorn Memorial Hospital, Thai Red Cross Soceity Note: This abstract has been previously published and is available at [1]. It is included here as a complete record of the abstracts from the conference. References 1. Suwanpasu S, Sattayasomboon Y (2016) Accuracy of modified early warning scores for predicting mortality in hospital: a systematic review and meta-analysis. J Intensive & Crit Care 2:2.

Note: This abstract has been previously published and is available at [1]. It is included here as a complete record of the abstracts from the conference. References 1. Suwanpasu S, Sattayasomboon Y (2016) Accuracy of modified early warning scores for predicting mortality in hospital: a systematic review and meta-analysis. J Intensive & Crit Care 2:2. A506 Validation of SAPS 3 E APACHE II in elderly patients admitted to brazilian intensive care unit N.M. Filgueiras Filho1,2, J.C.A. Oliveira1, C.S. Ballalai1,2, C.V. De Lucia1,2, G.P. Araponga1, L.N. Veiga1, C.S. Silva1,2, M.E. Garrido1,2, B.B. Ramos1,2, E.F. Ricaldi1,2, S.S. Gomes1,2, GEMINI 1Núcleo de Ensino e Pesquisa Hospital da Cidade, Intensive Care, Salvador, Brazil; 2Universidade Salvador - UNIFACS, Núcleo de Pesquisa Clínica, Salvador, Brazil Correspondence: N.M. Filgueiras Filho - Universidade Salvador - UNIFACS, Núcleo de Pesquisa Clínica, Salvador, Brazil Introduction: The world population is ageing. Together with this fact there is increase rate of elderly patient's admissions to Intensive Care Units. Therefore, studies aimed in elderly population are important. The prognostic scores have been widely used in intensive care medicine, among them SAPS 3 and APACHE II have a significant importance. The sensitivity and specificity of all the systems are vary depending on the score systems and the different groups of patients. Objectives: Validate and compare SAPS 3 and APACHE II in elderly patients admitted in ICU in Salvador-BA, Brazil.

A506 Validation of SAPS 3 E APACHE II in elderly patients admitted to brazilian intensive care unit N.M. Filgueiras Filho1,2, J.C.A. Oliveira1, C.S. Ballalai1,2, C.V. De Lucia1,2, G.P. Araponga1, L.N. Veiga1, C.S. Silva1,2, M.E. Garrido1,2, B.B. Ramos1,2, E.F. Ricaldi1,2, S.S. Gomes1,2, GEMINI 1Núcleo de Ensino e Pesquisa Hospital da Cidade, Intensive Care, Salvador, Brazil; 2Universidade Salvador - UNIFACS, Núcleo de Pesquisa Clínica, Salvador, Brazil Correspondence: N.M. Filgueiras Filho - Universidade Salvador - UNIFACS, Núcleo de Pesquisa Clínica, Salvador, Brazil Introduction: The world population is ageing. Together with this fact there is increase rate of elderly patient's admissions to Intensive Care Units. Therefore, studies aimed in elderly population are important. The prognostic scores have been widely used in intensive care medicine, among them SAPS 3 and APACHE II have a significant importance. The sensitivity and specificity of all the systems are vary depending on the score systems and the different groups of patients. Objectives: Validate and compare SAPS 3 and APACHE II in elderly patients admitted in ICU in Salvador-BA, Brazil. Methods: Prospective cohort study placed in a general ICU. In the study were included patients above age of 65 admitted to general ICU, from March 2010 to May 2013. All the patients were assessed using the SAPS 3 (Global Equation and Central & South American Equation) and APACHE II scoring systems. Logistic regressions analysis was used to calculate the sensitivity, specificity and accuracy as well as the OR probability of mortality. The ability to predict mortality was performed with ROC curves. The differences between observed-to-predicted mortality were analyzed with the Hosmer-Lemeshow test.

scoring systems. Logistic regressions analysis was used to calculate the sensitivity, specificity and accuracy as well as the OR probability of mortality. The ability to predict mortality was performed with ROC curves. The differences between observed-to-predicted mortality were analyzed with the Hosmer-Lemeshow test. Results: There were analyzed 1106 patients, 497 (44.9 %) males. Median age: 76 years (IQR: 71–83 years). Hospital mortality 207 (18.7 %). Median SAPS 3 and APACHE II: 48 (IQR: 42–55), 14 (IQR: 11–18). The sensitivity, specificity and accuracy were respectively 70 %, 68.7 % and 71.4 % for SAPS 3 Global Equation; 70 %, 68.7 % and 71.4 % for SAPS 3 Central & South American Equation (CSA); 67.6 %, 69.1 % and 68.81 % for APACHE II. The area under the curve (AUC) of these models was: 0.748 for SAPS 3 Global Equation, 0.745 for SAPS 3 Central & South American Equation (CSA) and 0.740 for APACHE II. But the Hosmer-Lemeshow goodness-of-fit test H statistic also revealed poor performance for SAPS 3 Central & South American Equation (CSA) scoring system. Conclusions: In our series, the SAPS 3 Global Equation is the model that best predicts mortality in patients with at least 65 years admitted to a Brazilian ICU. Also, SAPS 3 Central & South American Equation (CSA) revealed poor calibration for critical care elderly patients. References 1- Sacanella E, Pérez-Castejón JM, Nicolás JM et al. Intensive Care Med. 2009 Mar;35(3):550–5. 2- Soares M, Silva UV, Teles JM et al. Intensive Care Med. 2010 Jul;36(7):1188–95.

Conclusions: In our series, the SAPS 3 Global Equation is the model that best predicts mortality in patients with at least 65 years admitted to a Brazilian ICU. Also, SAPS 3 Central & South American Equation (CSA) revealed poor calibration for critical care elderly patients. References 1- Sacanella E, Pérez-Castejón JM, Nicolás JM et al. Intensive Care Med. 2009 Mar;35(3):550–5. 2- Soares M, Silva UV, Teles JM et al. Intensive Care Med. 2010 Jul;36(7):1188–95. A507 Evaluation of QSOFA as a predictive tool using a restrospective ICU cohort L. Gemmell, A. MacKay, C. Wright, R.I. Docking, P. Doherty, E. Black, P. Stenhouse Queen Elizabeth University Hospital, Anaesthetics and Intensive Care, Glasgow, UK Correspondence: L. Gemmell - Queen Elizabeth University Hospital, Anaesthetics and Intensive Care, Glasgow, UK Introduction: Recent guidelines on the definition of sepsis from the Sepsis 3 working group have suggested the use of a simplified version of the SOFA score as an initial way to screen for sepsis (1). This score allocates a point each for new altered mentation (GCS < 15), respiratory rate greater than 22 and a systolic blood pressure < 100 mmHg. It has been given the term qSOFA based on the original SOFA (sepsis related organ failure assessment). The purpose of the score is to identify patients early with sepsis as dealy in recognition is associated with an increased mortality and increased length of stay. We decided to evaluate whether the qSOFA score when applied to ICU patients could be validated for use in our patient population.

organ failure assessment). The purpose of the score is to identify patients early with sepsis as dealy in recognition is associated with an increased mortality and increased length of stay. We decided to evaluate whether the qSOFA score when applied to ICU patients could be validated for use in our patient population. Methods: A historical patient cohort was identified by searching the WardWatcher CIS dataset for all patients admitted from 01/01/204 - 31/12/2014 to a five bedded teaching hospital Intensive Care Unit in Glasgow. These patients demographic data was collected along with the parameters from qSOFA (as above) along with outcome data. After removal of patients with incomplete datasets, statistical analysis was performed. Results: There were 3357 patients included in the study, with analysis of 1353 due to incomplete qSOFA data. Baeline data for these patients was as follows: 58.9 % male, mean age 57.4 +/− 0.9 years, mean APACHE-II score 19.7 +/− 0.5, mean predicted mortality 37 +/− 1.5 %, mean length of stay 5.0 +/− 0. days and mortality of 33 %. Using the available data from Wardwatcher we generated qSOFA data as below. Conclusions: As can be seen from above there is a good correlation between the qSOFA and APACHE-II in terms of recognition of severity. There is a significant correlation between qSOFA and increased length of stay and mortality. This would suggest that qSOFA is a quick and useful tool that can be used on ICU admission to aid prognostication which can assist in discussions with patients, relatives and colleagues. References

Conclusions: As can be seen from above there is a good correlation between the qSOFA and APACHE-II in terms of recognition of severity. There is a significant correlation between qSOFA and increased length of stay and mortality. This would suggest that qSOFA is a quick and useful tool that can be used on ICU admission to aid prognostication which can assist in discussions with patients, relatives and colleagues. References 1. Seymour et al. Assessment of clinical criteria for sepsis. JAMA 2016;315(8):762–774Table 19 (abstract A507). qSOFA qSOFA 0 1 2 3 p-value n 97 384 574 296 APACHE II 13.5 +/− 1.2 16.2 +/− 0.8 20.4 +/− 0.7 24.8 +/− 1.3 <0.001 ICU stay 3.6 +/− 0.8 4.4 +/− 0.5 5.1 +/− 0.5 5.7 +/− 0.8 0.006 Mortality 12.4% 21.9% 35.5% 50.3% <0.001

1. Seymour et al. Assessment of clinical criteria for sepsis. JAMA 2016;315(8):762–774Table 19 (abstract A507). qSOFA qSOFA 0 1 2 3 p-value n 97 384 574 296 APACHE II 13.5 +/− 1.2 16.2 +/− 0.8 20.4 +/− 0.7 24.8 +/− 1.3 <0.001 ICU stay 3.6 +/− 0.8 4.4 +/− 0.5 5.1 +/− 0.5 5.7 +/− 0.8 0.006 Mortality 12.4% 21.9% 35.5% 50.3% <0.001 Poster Corner Sessions ACID-BASE BALANCE AND INTERMEDIARY METABOLISM A508 The risk of type 2 diabetes in survivors of critical illness with stress induced hyperglycaemia M.P. Plummer1,2, M.E. Finnis1,2, L.K. Phillips3,4, P. Kar1,2, S. Bihari5,6, V. Biradar7, S. Moodie2, M. Horowitz3,4, J.E. Shaw8, A.M. Deane1,2 1University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia; 2Royal Adelaide Hospital, Intensive Care Unit, Adelaide, Australia; 3University of Adelaide, Discipline of Medicine, Adelaide, Australia; 4Royal Adelaide Hospital, Department of Endocrinology, Adelaide, Australia; 5Flinders University, Department of Critical Care Medicine, Adelaide, Australia; 6Flinders Medical Centre, Department of Intensive Care Medicine, Adelaide, Australia; 7Lyell McEwin Hospital, Department of Intensive Care Medicine, Adelaide, Australia; 8Baker IDI Heart and Diabetes Institute, Melbourne, Australia Correspondence: M.P. Plummer - Royal Adelaide Hospital, Intensive Care Unit, Adelaide, Australia Introduction: Hyperglycaemia occurs frequently in the critically ill and may be secondary to either diabetes (recognised or not), or stress induced hyperglycaemia. Stress induced hyperglycaemia occurs in patients who have normal glucose tolerance following resolution of their acute illness. Whether stress induced hyperglycaemia unmasks latent insulin resistance and/or impaired β-cell function has not been adequately explored.

(recognised or not), or stress induced hyperglycaemia. Stress induced hyperglycaemia occurs in patients who have normal glucose tolerance following resolution of their acute illness. Whether stress induced hyperglycaemia unmasks latent insulin resistance and/or impaired β-cell function has not been adequately explored. Objectives: To evaluate the association between stress induced hyperglycaemia, the development of diabetes and long-term mortality in survivors of critical illness. Methods: This is a retrospective, multi-centre observational study. All adult patients surviving admission to a tertiary intensive care unit (ICU) in South Australia between 2004 and 2011 were included. Stress induced hyperglycaemia was defined as a blood glucose ≥ 11.1 mmol/l within the first 24 hours of ICU admission. Prevalent diabetes was identified through ICD-10 coding or prior registration with the Australian National Diabetes Service Scheme (NDSS). Incident diabetes was identified as NDSS registration > 30 days after hospital discharge until July 2015. The predicted risk of developing diabetes was described as sub-hazard ratios using competing risk regression. Survival was assessed using Cox proportional hazards regression.

ational Diabetes Service Scheme (NDSS). Incident diabetes was identified as NDSS registration > 30 days after hospital discharge until July 2015. The predicted risk of developing diabetes was described as sub-hazard ratios using competing risk regression. Survival was assessed using Cox proportional hazards regression. Results: Stress induced hyperglycaemia was identified in 2,883 (17 %) of 17,074 patients without diabetes. The overall incidence of subsequent type 2 diabetes following critical illness was 4.8 % (821 of 17,074). The risk of diabetes in patients with stress induced hyperglycaemia was approximately double that of those without (HR 1.91 (95 % CI 1.62, 2.26), p < 0.001) and was sustained regardless of age or severity of illness. Subdividing age into approximate deciles the greatest risk was seen in the 50–59 year age group with over a seven-fold risk (HR 7.90 (5.38, 11.60), p < 0.001). Stress induced hyperglycaemia was not associated with increased long-term mortality in patients who survived their hospital admission. Conclusions: Stress induced hyperglycaemia within 24 hours of admission to the ICU identifies patients at greater risk of subsequent diabetes. Grant acknowledgement This project was supported by grants from the Diabetes Australia Research Trust and a University of Adelaide, Discipline of Acute Care Medicine Maurice Sando Project Grant.

Conclusions: Stress induced hyperglycaemia within 24 hours of admission to the ICU identifies patients at greater risk of subsequent diabetes. Grant acknowledgement This project was supported by grants from the Diabetes Australia Research Trust and a University of Adelaide, Discipline of Acute Care Medicine Maurice Sando Project Grant. A509 Identifying an optimal target for blood glucose management among critically ill patients: a network meta-analysis T. Yatabe1, S. Inoue2, M. Sakaguchi3, M. Egi4 1Kochi Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan; 2Tokai University Hachioji Hospital, Department of Emergency and Critical Care Medicine, Hachioji, Japan; 3Kochi Medical School, Integrated Center for Advanced Medical Technologies, Nankoku, Japan; 4Kobe University Hospital, Department of Anesthesiology, Kobe, Japan Correspondence: T. Yatabe - Kochi Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan Introduction: Although several clinical guidelines recommend maintaining blood glucose between 144 and 180 mg/dL, these are insufficient evidences to assess the benefit and harm of this glycemic target. Objectives: We conducted a network meta-analysis to assess the impact of glycemic target 144-180 mg/dL on the incidence of hypoglycemia, the risk of infection and mortality in compared with other glycemic target as < 110 mg/dL, 110-144 mg/dL and >180 mg/dL.

A509 Identifying an optimal target for blood glucose management among critically ill patients: a network meta-analysis T. Yatabe1, S. Inoue2, M. Sakaguchi3, M. Egi4 1Kochi Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan; 2Tokai University Hachioji Hospital, Department of Emergency and Critical Care Medicine, Hachioji, Japan; 3Kochi Medical School, Integrated Center for Advanced Medical Technologies, Nankoku, Japan; 4Kobe University Hospital, Department of Anesthesiology, Kobe, Japan Correspondence: T. Yatabe - Kochi Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan Introduction: Although several clinical guidelines recommend maintaining blood glucose between 144 and 180 mg/dL, these are insufficient evidences to assess the benefit and harm of this glycemic target. Objectives: We conducted a network meta-analysis to assess the impact of glycemic target 144-180 mg/dL on the incidence of hypoglycemia, the risk of infection and mortality in compared with other glycemic target as < 110 mg/dL, 110-144 mg/dL and >180 mg/dL. Methods: We considered all studies from three recent systematic reviews [1–3] and searched the PubMed database for studies that compared target blood glucose levels among critically ill patients. The primary outcomes were in-hospital mortality, and the secondary outcomes were the incidence of hypoglycemia, and risk of sepsis or bloodstream infection.

Methods: We considered all studies from three recent systematic reviews [1–3] and searched the PubMed database for studies that compared target blood glucose levels among critically ill patients. The primary outcomes were in-hospital mortality, and the secondary outcomes were the incidence of hypoglycemia, and risk of sepsis or bloodstream infection. Results: We identified 71 studies through the re-analysis of the three systematic reviews, and 431 studies through the PubMed search. Thirty-nine potentially eligible studies were considered for inclusion, although 12 studies were excluded after screening the full texts. Thus, the network meta-analysis included 14,495 patients from 27 studies. The mean patient age was 61.6 years, and 21.7 % of the patients had diabetes. There was no significant difference of in-hospital mortality and risk of sepsis or bloodstream infection among 3 glycemic bands (<110 mg/dL, 110-144 mg/dL and 144-180 mg/dL). Target blood glucose levels of 144–180 mg/dL had significantly lower in-hospital mortality and risk of sepsis or bloodstream infection in compared with >180 mg/dL (odds ratio [OR]:0.82, 95 % credible intervals [CI]: 0.69-0.96; OR: 0.69, 95 % CI: 0.52-0.92, respectively). Our network meta-analysis revealed that target blood glucose levels of < 110 mg/dL and 110–144 mg/dL were associated with a 6–7 fold higher risk of hypoglycemia, compared to targets of 144–180 mg/dL and >180 mg/dL. There were no significant differences in the risk of hypoglycemia between < 110 mg/dL and 110–144 mg/dL (OR = 0.76 (95 % CI: 0.49-1.11). There were also no significant differences in the risk of hypoglycemia between and between 144–180 mg/dL and >180 mg/dL (OR = 1.0 (95 % CI: 0.30-2.70). The rank probability analysis indicated that a target blood glucose level of 144–180 mg/dL had the highest probability of being the best target blood glucose level for reducing hypoglycemia and hospital mortality rates and was the second best target level, following a target level of 110–144 mg/dL, for reducing sepsis or bloodstream infection rates.

ndicated that a target blood glucose level of 144–180 mg/dL had the highest probability of being the best target blood glucose level for reducing hypoglycemia and hospital mortality rates and was the second best target level, following a target level of 110–144 mg/dL, for reducing sepsis or bloodstream infection rates. Conclusions: The results of our network meta-analysis indicate that target blood glucose levels of 144–180 mg/dL is an optimal target for balancing the risks and benefits of insulin therapy among critically ill patients. References [1] Wiener RS,et al. JAMA. 2008;300:933–44. [2] Friedrich JO, et al. Crit Care. 2010;14:324. [3] Song F, et al. Biomed Res Int. 2014;2014:698265. Grant acknowledgement None.

Conclusions: The results of our network meta-analysis indicate that target blood glucose levels of 144–180 mg/dL is an optimal target for balancing the risks and benefits of insulin therapy among critically ill patients. References [1] Wiener RS,et al. JAMA. 2008;300:933–44. [2] Friedrich JO, et al. Crit Care. 2010;14:324. [3] Song F, et al. Biomed Res Int. 2014;2014:698265. Grant acknowledgement None. A510 Long-term mortality of critically ill patients with diabetes who survive admission to intensive care Y. Ali Abdelhamid1,2, M.P. Plummer1,2, M.E. Finnis1,2, L.K. Phillips3,4, P. Kar1,2, S. Bihari5,6, V. Biradar7, S. Moodie1, M. Horowitz3,4, J.E. Shaw8, A.M. Deane1,2 1Royal Adelaide Hospital, Department of Critical Care Services, Adelaide, Australia; 2University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia; 3University of Adelaide, Discipline of Medicine, Adelaide, Australia; 4Royal Adelaide Hospital, Endocrine and Metabolic Unit, Adelaide, Australia; 5Flinders University, Department of Critical Care Medicine, Adelaide, Australia; 6Flinders Medical Centre, Department of Intensive Care Medicine, Adelaide, Australia; 7Lyell McEwin Hospital, Department of Intensive Care Medicine, Adelaide, Australia; 8Baker IDI Heart and Diabetes Institute, Melbourne, Australia Correspondence: Y. Ali Abdelhamid - University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia Introduction: Diabetes is a risk factor for the development and severity of critical illness, and the need for Intensive Care Unit (ICU) admission. Despite this, diabetes does not confer a greater risk of death for those with critical illness within the index hospital admission. However, long-term outcomes in patients with diabetes who survive ICU admission remain unknown.

ment and severity of critical illness, and the need for Intensive Care Unit (ICU) admission. Despite this, diabetes does not confer a greater risk of death for those with critical illness within the index hospital admission. However, long-term outcomes in patients with diabetes who survive ICU admission remain unknown. Objectives: Our objectives were to evaluate the effect of diabetes on both long-term survival rates and average years of life lost for patients admitted to ICU and who survived hospitalisation. Methods: We evaluated all adult patients in South Australia between 2004 and 2011 who survived hospitalisation requiring admission to a Public Hospital ICU. Demographic and admission data from the Australia and New Zealand Intensive Care Society Adult Patient Database were linked to population based datasets to match (i) International Classification of Disease (ICD-10) codes for diabetes through the Department of Health Integrated South Australian Activity Collection dataset, (ii) diabetes diagnosis in the national register (the National Diabetes Service Scheme), and (iii) mortality through the Australian National Death Index. Life years lost were calculated using age and sex specific life-tables from the Australian Bureau of Statistics.

egrated South Australian Activity Collection dataset, (ii) diabetes diagnosis in the national register (the National Diabetes Service Scheme), and (iii) mortality through the Australian National Death Index. Life years lost were calculated using age and sex specific life-tables from the Australian Bureau of Statistics. Data are presented as frequencies and proportions for categorical variables and mean (standard deviation) or median [interquartile range] for continuous variables. Between group comparisons were performed by t-test, Wilcoxon rank-sum or Chi-square tests. Longitudinal survival was analysed using Cox proportional hazards regression and average life years lost by linear regression. Results: 5451 patients with and 17022 patients without diabetes were included. Patients with diabetes were older (64.7 (14.9) vs 57.6 (19.7) years; p < 0.0001) and had higher illness severity on admission to ICU (APACHE III 62 [48, 80] vs 54 [38, 72]; p < 0.0001). Crude mortality rates [95%CI] were 105.5 [101.6-109.6] and 67.6 [65.9-69.3] per 1000 person years for those with and without diabetes respectively. Patients with diabetes were more likely to die after hospital discharge (unadjusted hazard ratio 1.52 [95 % CI 1.46-1.59]; p < 0.001) (Fig. 38).

s 54 [38, 72]; p < 0.0001). Crude mortality rates [95%CI] were 105.5 [101.6-109.6] and 67.6 [65.9-69.3] per 1000 person years for those with and without diabetes respectively. Patients with diabetes were more likely to die after hospital discharge (unadjusted hazard ratio 1.52 [95 % CI 1.46-1.59]; p < 0.001) (Fig. 38). Diabetes remained an independent risk factor for death when adjusted for age, sex, site, indigenous status, APACHE III score and admission diagnosis (hazard ratio 1.17 [95 % CI 1.12-1.22]; p < 0.01). Patients with diabetes also suffered a greater number of average life years lost and this was most marked in the 4th and 5th decades of life (Fig. 39). Conclusions: Mortality and number of life years lost for ICU survivors with diabetes is greater than for survivors without diabetes. Strategies to improve outcomes for these patients should be evaluated. GRANT ACKNOWLEDGEMENT Dr Ali Abdelhamid is supported by a Royal Adelaide Hospital Clarkson Scholarship. The study was supported by a Diabetes Australia Research Trust Project Grant.Fig. 38 (abstract A510). ᅟ Fig. 39 (abstract A510). ᅟ

Conclusions: Mortality and number of life years lost for ICU survivors with diabetes is greater than for survivors without diabetes. Strategies to improve outcomes for these patients should be evaluated. GRANT ACKNOWLEDGEMENT Dr Ali Abdelhamid is supported by a Royal Adelaide Hospital Clarkson Scholarship. The study was supported by a Diabetes Australia Research Trust Project Grant.Fig. 38 (abstract A510). ᅟ Fig. 39 (abstract A510). ᅟ A511 Glycated hemoglobin A1C on the day of emergency surgery is the marker of premorbid glycemic control M. Hokka, M. Egi, S. Mizobuchi Kobe University Hospital, Anesthesiology, Kobe, Japan Correspondence: M. Hokka - Kobe University Hospital, Anesthesiology, Kobe, Japan Introduction: Glycated hemoglobin A1c (HbA1c) is used to estimate chronic glycemic control. Recently, several studies suggested that the optimal glycemic control in acutely ill patients might be better being adjusted according to the premorbid chronic glycemic control (1–3). However, the majority of patients would not have been measured the HbA1c-level prior to the acute illness. As the HbA1c-level would be affected by the patients' conditions, HbA1c-levels during an acute illness may not reflect the actual premorbid glycemic status. Only limited research has been conducted so far to assess the value of HbA1c-levels during the state of an acute illness to estimate the premorbid glycemic control (4). Objectives: Assessment of the relationship between HbA1c during the state of an acute illness and HbA1c levels measured within 30 days before an acute illness.

A511 Glycated hemoglobin A1C on the day of emergency surgery is the marker of premorbid glycemic control M. Hokka, M. Egi, S. Mizobuchi Kobe University Hospital, Anesthesiology, Kobe, Japan Correspondence: M. Hokka - Kobe University Hospital, Anesthesiology, Kobe, Japan Introduction: Glycated hemoglobin A1c (HbA1c) is used to estimate chronic glycemic control. Recently, several studies suggested that the optimal glycemic control in acutely ill patients might be better being adjusted according to the premorbid chronic glycemic control (1–3). However, the majority of patients would not have been measured the HbA1c-level prior to the acute illness. As the HbA1c-level would be affected by the patients' conditions, HbA1c-levels during an acute illness may not reflect the actual premorbid glycemic status. Only limited research has been conducted so far to assess the value of HbA1c-levels during the state of an acute illness to estimate the premorbid glycemic control (4). Objectives: Assessment of the relationship between HbA1c during the state of an acute illness and HbA1c levels measured within 30 days before an acute illness. Methods: These are retrospective observational studies. The Ethics Committee of the Kobe University Hospital approved this investigation. The committee waived the need for informed consent concerning studies involving the use of databases. We screened patients' HbA1c levels that were measured within 30 days prior to an emergency surgery, requiring general anesthesia. Among them, we selected patients whose HbA1c levels were also measured on the day of the surgery. Thus we obtained HbA1c-levels measured within 30 days before the emergency surgery (preHbA1c) and on the day of the surgery (opeHbA1c). Those HbA1c-levels were measured in the same laboratory. The correlations of the two HbA1c levels were assessed, using Pearson's correlation coefficient. Agreement between the two HbA1c-levels was assessed using the Bland-Altman approach.

before the emergency surgery (preHbA1c) and on the day of the surgery (opeHbA1c). Those HbA1c-levels were measured in the same laboratory. The correlations of the two HbA1c levels were assessed, using Pearson's correlation coefficient. Agreement between the two HbA1c-levels was assessed using the Bland-Altman approach. Results: We studied 100 HbA1c-levels in 50 patients. The patients were 66.8 years old in average. 31 of them were male patients (62 %). 38 out of 50 patients (76 %) were required emergency cardiovascular surgery. PreHbA1c was measured 11.4 days before the emergency operation. The average preHbA1c was 6.3 % and the opeHbA1c was 6.2 %. There is a significant correlation between preHbA1c and opeHbA1c (r = 0.83, p < 0.001). The mean difference between preHbA1c and opeHbA1c was −0.1 % (95 % confidential interval; −1.2 % to 0.9 %). This difference of the two HbA1c did not correlate with the time between the two measurements (r = 0.17, p = 0.29). Conclusions: The HbA1c-level on the day of surgery is a useful marker of premorbid glycemic control in patients requiring emergency surgery. References 1) Crit Care Med 2011; 39:105–111 2) Intensive Care Med 2014; 40:973–980 3) Intensive Care Med. 2016;42(4):562–71 4) Crit Care Med.2016 [Epub ahead of print

Results: We studied 100 HbA1c-levels in 50 patients. The patients were 66.8 years old in average. 31 of them were male patients (62 %). 38 out of 50 patients (76 %) were required emergency cardiovascular surgery. PreHbA1c was measured 11.4 days before the emergency operation. The average preHbA1c was 6.3 % and the opeHbA1c was 6.2 %. There is a significant correlation between preHbA1c and opeHbA1c (r = 0.83, p < 0.001). The mean difference between preHbA1c and opeHbA1c was −0.1 % (95 % confidential interval; −1.2 % to 0.9 %). This difference of the two HbA1c did not correlate with the time between the two measurements (r = 0.17, p = 0.29). Conclusions: The HbA1c-level on the day of surgery is a useful marker of premorbid glycemic control in patients requiring emergency surgery. References 1) Crit Care Med 2011; 39:105–111 2) Intensive Care Med 2014; 40:973–980 3) Intensive Care Med. 2016;42(4):562–71 4) Crit Care Med.2016 [Epub ahead of print A512 Effects of antecedent hypoglycaemia on cardiac and gastric responses to subsequent hypoglycaemia in health P. Kar1,2, M. Plummer3, Y. Ali Abdelhamid1,2, E. Giersch1, M. Summers1, S. Hatzinikolas4, S. Heller5, M. Chapman1,2,4, K. Jones4,6, M. Horowitz4,6, A. Deane1,2,4 1Royal Adelaide Hospital, Intensive Care Unit, Adelaide, Australia; 2University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia; 3Addenbrooke's Hospital, Neurosciences Critical Care Unit, Cambridge, UK; 4University of Adelaide, NHMRC Centre for Research Excellence, Adelaide, Australia; 5University of Sheffield, Academic Unit of Diabetes, Endocrinology and Metabolism, Sheffield, UK; 6University of Adelaide, Discipline of Medicine, Adelaide, Australia Correspondence: P. Kar - University of Adelaide, Discipline of Acute Care Medicine, Adelaide, Australia Introduction: Hypoglycaemia during ICU admission is associated with adverse outcomes after discharge. In both health and patients with diabetes, acute hypoglycaemia triggers counter-regulatory responses increasing cardiac contractility and accelerating gastric emptying. However, antecedent hypoglycaemia attenuates endogenous catecholamine and vagal responses to subsequent hypoglycaemia1. Accordingly, patients who were hypoglycaemic during ICU may be unable to mount a physiological response to hypoglycaemia once discharged from ICU.

diac contractility and accelerating gastric emptying. However, antecedent hypoglycaemia attenuates endogenous catecholamine and vagal responses to subsequent hypoglycaemia1. Accordingly, patients who were hypoglycaemic during ICU may be unable to mount a physiological response to hypoglycaemia once discharged from ICU. Objectives: To determine the effects of antecedent hypoglycaemia on cardiac and gastric responses to subsequent hypoglycaemia in health. Methods: Ten healthy men (age 22.5 (1.0) years, BMI 23.8 (0.5) kg/m2) were studied on two occasions, lasting 30 hours, separated by ≥ 6 weeks and randomised to either 'control' (Day 1 (C1) - three euglycaemic clamps at a blood glucose of 6 mmol/L followed day 2 (C2) - one hypoglycaemic clamp at 2.8 mmol/L), or 'antecedent hypoglycaemia' (Day 1 (AH1) - three clamps at blood glucose of 2.8 mmol/L followed by day 2 (AH2) - one clamp at 2.8 mmol/L). Cardiac contractility was measured using fractional shortening (FS) via 2D echocardiography on days 1 and 2 of the control (FSC1 and FSC2) and antecedent hypoglycaemia (FSAH1 and FSAH2) periods. Gastric emptying (GE) was measured on day 1 and day 2, for both control (GEC1 and GEC2) and antecedent hypoglycaemia (GEAH1 and GEAH2) study periods, using scintigraphy, following ingestion of a standardised meal consisting of 100 g of minced beef labelled with 20 MBq 99mtechnetium-sulphur colloid. Radioisotopic data were acquired for 180 minutes. Data are mean (SE).

ay 2, for both control (GEC1 and GEC2) and antecedent hypoglycaemia (GEAH1 and GEAH2) study periods, using scintigraphy, following ingestion of a standardised meal consisting of 100 g of minced beef labelled with 20 MBq 99mtechnetium-sulphur colloid. Radioisotopic data were acquired for 180 minutes. Data are mean (SE). Results: Fractional shortening was greater (FSC2 vs. FSC1, P < 0.01) and gastric emptying faster (GEC2 AUC vs. GEC1 AUC, P = 0.01) during acute, single episode hypoglycaemia compared to euglycaemia. When testing the effect of antecedent hypoglycaemia compared to a single episode of hypoglycaemia, fractional shortening may be attenuated (FSAH2 vs. FSC2, P = 0.06), whereas gastric emptying was unaffected (GEAH2AUC vs. GEC2 AUC, P = 0.74) (Figs. 40 and 41). Conclusions: In health, the increase in cardiac contractility during acute hypoglycaemia may be affected by antecedent hypoglycaemia whereas acceleration of gastric emptying during acute hypoglycaemia does not appear to be affected by antecedent was unaffected. These data suggest that antecedent hypoglycaemia during ICU may have “downstream” effects after discharge. References 1. Heller SR, Cryer PE. Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in nondiabetic humans. Diabetes. Feb 1991;40(2):223–226 Grant acknowledgement Dr Kar is supported by a Royal Adelaide Hospital A.R. Clarkson Scholarship.Fig. 40 (abstract A512). Cardiac fractional shortening on day 1 and day 2 of control and antecedent hypoglycaemia study periods

1. Heller SR, Cryer PE. Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in nondiabetic humans. Diabetes. Feb 1991;40(2):223–226 Grant acknowledgement Dr Kar is supported by a Royal Adelaide Hospital A.R. Clarkson Scholarship.Fig. 40 (abstract A512). Cardiac fractional shortening on day 1 and day 2 of control and antecedent hypoglycaemia study periods Fig. 41 (abstract A512). Gastric emptying curves on day 1 and day 2 of control and antecedent hypoglycaemia study periods

Dr Kar is supported by a Royal Adelaide Hospital A.R. Clarkson Scholarship.Fig. 40 (abstract A512). Cardiac fractional shortening on day 1 and day 2 of control and antecedent hypoglycaemia study periods Fig. 41 (abstract A512). Gastric emptying curves on day 1 and day 2 of control and antecedent hypoglycaemia study periods A513 Limits of the Stewart approach for plasmatic acid–base disturbance interpretation: an illustration with a mechanically ventilated porcin model of metabolic acidosis R. Schweizer1, M. Jacquet-Lagreze1, P. Portran1, S. Junot2, B. Allaouchiche3, J.-L. Fellahi1 1Anesthesiology and Critical Care Medicine, Hôpital Cardiologique, Hospices Civils de Lyon, Bron, France; 2VetAgro Sup, Ecole vétérinaire de Lyon, Marcy l'Etoile, France; 3Anesthesiology and Critical Care Medicine, Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Pierre Benite, France Correspondence: R. Schweizer - Anesthesiology and Critical Care Medicine, Hôpital Cardiologique, Hospices Civils de Lyon, Bron, France Introduction: Plasmatic acid–base equilibrium remains a confusing area more than a century after the Henderson-Hasselbalch equation publication. However, a consensus exists concerning arterial carbon dioxide pressure (pCO2) that would represent respiratory part of this equilibrium 1. Stewart described this parameter as "independent", in contrast to "dependent" parameters, e.g. pH and HCO3−2. Objectives: The primary objective of this experimental trial is to demonstrate that lactic acid infusion increase pCO2. This would be inconsistent with Stewart approach.

A513 Limits of the Stewart approach for plasmatic acid–base disturbance interpretation: an illustration with a mechanically ventilated porcin model of metabolic acidosis R. Schweizer1, M. Jacquet-Lagreze1, P. Portran1, S. Junot2, B. Allaouchiche3, J.-L. Fellahi1 1Anesthesiology and Critical Care Medicine, Hôpital Cardiologique, Hospices Civils de Lyon, Bron, France; 2VetAgro Sup, Ecole vétérinaire de Lyon, Marcy l'Etoile, France; 3Anesthesiology and Critical Care Medicine, Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Pierre Benite, France Correspondence: R. Schweizer - Anesthesiology and Critical Care Medicine, Hôpital Cardiologique, Hospices Civils de Lyon, Bron, France Introduction: Plasmatic acid–base equilibrium remains a confusing area more than a century after the Henderson-Hasselbalch equation publication. However, a consensus exists concerning arterial carbon dioxide pressure (pCO2) that would represent respiratory part of this equilibrium 1. Stewart described this parameter as "independent", in contrast to "dependent" parameters, e.g. pH and HCO3−2. Objectives: The primary objective of this experimental trial is to demonstrate that lactic acid infusion increase pCO2. This would be inconsistent with Stewart approach. Methods: After local ethical committee approval, 9 anesthetized, curarized and mechanically ventilated piglets were studied. Lactic acid infusion (0.33 mmol.kg−1.min−1) was administered. Each 3 minutes, an arterial gasometry was done. Each minute, the following parameters were collected: mean arterial pressure (MAP), cardiac output measured by transpulmonary thermodilution (CO), pulse oxymetry (SpO2) and end-tidal CO2 (EtCO2).

were studied. Lactic acid infusion (0.33 mmol.kg−1.min−1) was administered. Each 3 minutes, an arterial gasometry was done. Each minute, the following parameters were collected: mean arterial pressure (MAP), cardiac output measured by transpulmonary thermodilution (CO), pulse oxymetry (SpO2) and end-tidal CO2 (EtCO2). Bicarbonate (HCO3−) was calculated like this: HCO3− = 0.03*pCO2*10pH-6.1. Total CO2(tCO2) was calculated like this: tCO2 = HCO3− + 0.03*pCO2 The primary endpoints were pCO2 increase and tCO2 stability. To test the hypothesis of normal distribution, Shapiro Wilk test was applied. The p values were obtained from ANOVA analyses. p < 0.05 was considered as significant. Results: The results were: a decrease in pH (p < 0.001) (Fig. 42a), an increase in pCO2 (p = 0.002) (Fig. 42b), a decrease in HCO3− (p = 0.015) (Fig. 42d) without significant alteration in tCO2 (p = 0.08) (Fig. 42c). We also observed: an increase in EtCO2 (p = 0.001) (Fig. 42e) and lactatemia (p < 0.001) (Fig. 42f). CO (p = 0.052), MAP (p = 0.506) and SpO2 (p = 0.905) were not significantly altered. Conclusions: This in vivo metabolic acidosis model with stable minute ventilation shows an increase in pCO2 without significant alteration in tCO2. Although these results were biochemically predictable, they question one of the basements of Stewart approach, that considered pCO2 as an independent parameter. In our study, pCO2 seems a dependent parameter. tCO2, too often neglected parameter, seems an independent parameter. References (1) HJ Adrogué et al., Assessing acid–base disorders, Kidney Int, 2009; 76(12):1239–47

Conclusions: This in vivo metabolic acidosis model with stable minute ventilation shows an increase in pCO2 without significant alteration in tCO2. Although these results were biochemically predictable, they question one of the basements of Stewart approach, that considered pCO2 as an independent parameter. In our study, pCO2 seems a dependent parameter. tCO2, too often neglected parameter, seems an independent parameter. References (1) HJ Adrogué et al., Assessing acid–base disorders, Kidney Int, 2009; 76(12):1239–47 (2) PA Stewart, Modern quantitative acid–base chemistry, Can J Physiol Pharmacol, 1983; 61(12):1444–61Fig. 42 (abstract A513). ᅟ

Conclusions: This in vivo metabolic acidosis model with stable minute ventilation shows an increase in pCO2 without significant alteration in tCO2. Although these results were biochemically predictable, they question one of the basements of Stewart approach, that considered pCO2 as an independent parameter. In our study, pCO2 seems a dependent parameter. tCO2, too often neglected parameter, seems an independent parameter. References (1) HJ Adrogué et al., Assessing acid–base disorders, Kidney Int, 2009; 76(12):1239–47 (2) PA Stewart, Modern quantitative acid–base chemistry, Can J Physiol Pharmacol, 1983; 61(12):1444–61Fig. 42 (abstract A513). ᅟ A514 The role of bicarbonate precursors in balanced fluids during haemorrhagic shock with and without compromised liver function P. Guerci1,2,3, B. Ergin1, A. Kapucu4, C. Ince1 1Academic Medical Center, University of Amsterdam, Department of Translational Physiology, Amsterdam, Netherlands; 2University Hospital of Nancy, Departement of Anaesthesiology and Critical Care Medicine, Vandoeuvre-Les-Nancy, France; 3University of Lorraine, INSERM U1116, Vandoeuvre-Les-Nancy, France; 4Science Faculty, University of Istanbul, Department of Biology, Istanbul, Turkey Correspondence: P. Guerci - University of Lorraine, INSERM U1116, Vandoeuvre-Les-Nancy, France Introduction: Lactate, acetate and gluconate are anions used in balanced resuscitation fluids of which lactate and acetate are considered bicarbonate precursors. Several studies have demonstrated that balanced fluids are commonly used for volume expansion of critically ill patients who may also have impaired liver function (1,2). Little is known of the metabolic fate of this anions when the primary metabolising organ is compromised. Moreover, the role of gluconate embedded in PlasmaLyte in terms of acid–base control is unknown.

ids are commonly used for volume expansion of critically ill patients who may also have impaired liver function (1,2). Little is known of the metabolic fate of this anions when the primary metabolising organ is compromised. Moreover, the role of gluconate embedded in PlasmaLyte in terms of acid–base control is unknown. Objectives: This study investigated the role of the liver in the efficacy of balanced and unbalanced solutions to correct acid–base alterations and renal haemodynamics and microvascular oxygenation in a rat model of resuscitated hemorrhagic shock (HS).

ids are commonly used for volume expansion of critically ill patients who may also have impaired liver function (1,2). Little is known of the metabolic fate of this anions when the primary metabolising organ is compromised. Moreover, the role of gluconate embedded in PlasmaLyte in terms of acid–base control is unknown. Objectives: This study investigated the role of the liver in the efficacy of balanced and unbalanced solutions to correct acid–base alterations and renal haemodynamics and microvascular oxygenation in a rat model of resuscitated hemorrhagic shock (HS). Methods: Ringer's Lactate (RL), Ringer's Acetate (RA), Plasma-Lyte (PL) or normal saline (NS) were administered following HS in the presence or absence of a 70 % partial liver resection (PLR). Renal haemodynamics and microvascular oxygenation (by oxygen-dependent quenching of phosphorescence) were measured as well as levels of lactate, gluconate and acetate in plasma and urine. Kidney wet and dry weighing was also assessed. Results: PLR resulted in increased liver enzymes compared in control and HS groups (p < 0.01). HS decreased systemic and renal haemodynamics and reduced microvascular kidney oxygenation to 20 and 14 mmHg for cortex and medulla respectively, associated with lactic acidosis (p < 0.01). Resuscitation with balanced fluids did not fully restore renal oxygenation (p < 0.01). RA and PL increased bicarbonate levels (15.5 ± 1.1 and 16.4 ± 1.6 vs 13.4 ± 2.0 and 11.5 ± 2.5 mmol/L) and restored pH better than RL or NS (7.29 ± 0.09 and 7.30 ± 0.04 vs 7.22 ± 0.12 and 7.14 ± 0.09) respectively in the PLR experiment (p < 0.01). PLR caused an increase in plasma gluconate after PL resuscitation (10.4 ± 2.8 vs 13.8 ± 5.1 g/L p < 0.05).

vels (15.5 ± 1.1 and 16.4 ± 1.6 vs 13.4 ± 2.0 and 11.5 ± 2.5 mmol/L) and restored pH better than RL or NS (7.29 ± 0.09 and 7.30 ± 0.04 vs 7.22 ± 0.12 and 7.14 ± 0.09) respectively in the PLR experiment (p < 0.01). PLR caused an increase in plasma gluconate after PL resuscitation (10.4 ± 2.8 vs 13.8 ± 5.1 g/L p < 0.05). Conclusions: Acetate buffered balanced fluids show superior buffering effects than RL and NS. Gluconate is partially metabolized by the liver although it does not contribute to acid–base control because of large excretion in urine. Acetate is metabolized regardless of liver function and may be the most efficient bicarbonate precursor. Lactate infusion tends to overwhelm the metabolism capacities of the residual liver. NS seemed to be the most unsuitable fluid when compared to balanced fluids, even in cases of liver failure. References 1. Cecconi M et al. Fluid challenges in intensive care: the FENICE study: A global inception cohort study. Intensive Care Med 2015; 41: 1529–37 2. Boulain T et al. Volume expansion in the first 4 days of shock: a prospective multicentre study in 19 French intensive care units. Intensive Care Med 2015; 41: 248–56 Grant acknowledgement P. Guerci is supported by a grant from the Société Française d'Anesthésie-Réanimation (SFAR), France. This study has been funded by Baxter HealthCare.Fig. 43 (abstract A514). Plasma and urine gluconate levels after fluid

2. Boulain T et al. Volume expansion in the first 4 days of shock: a prospective multicentre study in 19 French intensive care units. Intensive Care Med 2015; 41: 248–56 Grant acknowledgement P. Guerci is supported by a grant from the Société Française d'Anesthésie-Réanimation (SFAR), France. This study has been funded by Baxter HealthCare.Fig. 43 (abstract A514). Plasma and urine gluconate levels after fluid A515 Prevalence of ketosis, ketonuria and ketoacidosis during permissive hyperglycemia in critically ill patients with diabetes L. Cioccari1,2, N. Luethi1, M. Crisman1,3, R. Bellomo1,4, J. Mårtensson1,5 1Austin Health, University of Melbourne, Department of Intensive Care, Heidelberg, Australia; 2Lucerne Cantonal Hospital, Department of Intensive Care, Lucerne, Switzerland; 3Cattinara Hospital, Trieste University School of Medicine, Department of Perioperative Medicine, Intensive Care and Emergency, Trieste, Italy; 4Australian and New Zealand Intensive Care Research Centre (ANZIC-RC), Monash University, Department of Epidemiology and Preventive Medicine, Melbourne, Australia; 5Karolinska Institutet, Section of Anaesthesia and Intensive Care Medicine, Department of Physiology and Pharmacology, Stockholm, Sweden Correspondence: L. Cioccari - Lucerne Cantonal Hospital, Department of Intensive Care, Lucerne, Switzerland Introduction: Recent evidence suggests that a more liberal glycemic control (target blood glucose level between 10 and 14 mmol/l) may be beneficial in patients with diabetes adapted to chronic hyperglycemia (glycated hemoglobin A1c [HbA1c] >7 %)1. However, whether such liberal strategies leads to relative insulin deficiency, accelerated ketone body production and ketoacidosis is uncertain. Accordingly, we conducted a prospective observational study to explore the incidence of ketosis, ketonuria and ketoacidosis during liberal glucose management of critically ill diabetic patients.

ch liberal strategies leads to relative insulin deficiency, accelerated ketone body production and ketoacidosis is uncertain. Accordingly, we conducted a prospective observational study to explore the incidence of ketosis, ketonuria and ketoacidosis during liberal glucose management of critically ill diabetic patients. Methods: We studied 60 critically ill diabetic patients treated according to a liberal glucose protocol targeting a blood glucose level (BGL) between 10-14 mmol/l. We performed daily measurement of bedside blood 3-beta-hydroxybutyrate (β-OHB) and semi-quantitative urine ketones on ICU admission and on the following mornings during the ICU stay, for a maximum of 10 consecutive days. Results: Median (IQR) blood ketone level on admission was 0.3 (0.1, 0.8) mmol/l. Ketoacidosis was rare (3 %), but some level of ketosis (β-OHB ≥ 0.6 mmol/l) was found in 38 patients (63 %) at some stage during their ICU stay. However, there was no significant difference in severity or prevalence of ketonemia and ketonuria among patients with BGL above (permissive hyperglycemia) or below 10 mmol/l. On multivariate linear regression there was no association between blood ketone levels and BGL, HbA1c, lactate levels or APACHE III score. Conclusions: Liberal glycemic control in critically ill, predominantly type 2, diabetic patients itself does not appear to lead to increased ketogenesis. We found no difference in prevalence and severity of ketonemia between patients with or without permissive hyperglycemia. Furthermore, severity of ketosis was unrelated to blood glucose levels. References

Conclusions: Liberal glycemic control in critically ill, predominantly type 2, diabetic patients itself does not appear to lead to increased ketogenesis. We found no difference in prevalence and severity of ketonemia between patients with or without permissive hyperglycemia. Furthermore, severity of ketosis was unrelated to blood glucose levels. References 1. Mårtensson J, Bellomo R. The Rationale for Permissive Hyperglycemia in Critically Ill Patients with Diabetes. In: Vincent J-L, editor. Annual Update in Intensive Care and Emergency Medicine 2016. Cham: Springer International Publishing; 2016. p. 365–372. A516 Evaluation of different glucose variability indices with continuous and intermittent glucose monitoring C. Righy Shinotsuka, D. Fagnoul, A. Brasseur, D. Orbegozo, J.-L. Vincent, J.-C. Preiser Université Libre de Bruxelles (ULB) - Hôpital Erasme, Intensive Care, Brussels, Belgium Correspondence: C. Righy Shinotsuka - Université Libre de Bruxelles (ULB) - Hôpital Erasme, Intensive Care, Brussels, Belgium Introduction: Reducing glucose variability (GV) is one of the aims of blood glucose (BG) control as it is an independent predictor of ICU and hospital mortality. Continuous glucose monitoring (CGM) might help to decrease GV by providing frequent measurements of glucose and information about glucose trend. However, the metrics of GV might differ according to the frequency of measurement. Objectives: To evaluate the correlation between GV indices derived from intermittent glucose values and from CGM.

A516 Evaluation of different glucose variability indices with continuous and intermittent glucose monitoring C. Righy Shinotsuka, D. Fagnoul, A. Brasseur, D. Orbegozo, J.-L. Vincent, J.-C. Preiser Université Libre de Bruxelles (ULB) - Hôpital Erasme, Intensive Care, Brussels, Belgium Correspondence: C. Righy Shinotsuka - Université Libre de Bruxelles (ULB) - Hôpital Erasme, Intensive Care, Brussels, Belgium Introduction: Reducing glucose variability (GV) is one of the aims of blood glucose (BG) control as it is an independent predictor of ICU and hospital mortality. Continuous glucose monitoring (CGM) might help to decrease GV by providing frequent measurements of glucose and information about glucose trend. However, the metrics of GV might differ according to the frequency of measurement. Objectives: To evaluate the correlation between GV indices derived from intermittent glucose values and from CGM. Methods: A correlation between GV indices derived from intermittent readings (blood gas analyser [BGA]) and those derived from CGM was searched from data recorded in the same patients over the same period of time.

Objectives: To evaluate the correlation between GV indices derived from intermittent glucose values and from CGM. Methods: A correlation between GV indices derived from intermittent readings (blood gas analyser [BGA]) and those derived from CGM was searched from data recorded in the same patients over the same period of time. Data from patients recruited for the MANAGE (Manual vs. Automated Monitoring Accuracy of Glucose) II trial between July 2012 and April 2014 were analysed. Inclusion criteria were age >18 years, expected ICU length of stay of ≥ 3 days, an Acute Physiology and Chronic Health Evaluation (APACHE) II ≥ 10, hyperglycemia >150 mg/dl at the time of admission and need for a central venous catheter, one lumen of which was connected to Optiscanner® (Optiscan, Hayward, CA), a validated mid-infrared spectroscopy method of glucose measurement. GV indices (standard deviation [SD], glucose variability index [GVI], glucose lability index [GLI], mean amplitude of glycemic excursions [MAGE], J-index and maximal glucose change [MGC]) were calculated from blood glucose (in mg/dl) measured intermittently by BGA and by CGM over the same period. Pearson´s correlation coefficient between intermittent and CGM values was calculated for each metric. A Bland-Altman plot with bias (mean difference between the device and BGA measurements) and limits of agreement (bias ± 1.96 x standard deviation of the bias) was used to analyse the agreement between the device and the BGA results.

elation coefficient between intermittent and CGM values was calculated for each metric. A Bland-Altman plot with bias (mean difference between the device and BGA measurements) and limits of agreement (bias ± 1.96 x standard deviation of the bias) was used to analyse the agreement between the device and the BGA results. Results: Data from 88 patients were analysed. For each index, there was a significant correlation (p < 0.0001) between values calculated from intermittent readings and CGM. However, correlations were higher for J-index (0.96), SD (0.95) and GVI (0.9) than for MGC (0.88), GLI (0.82) and MAGE (0.77). The mean difference and limits of agreement (lower limit, upper limit) observed in Bland-Altman plot were 0.01 (−0.10, 0.12) for GVI, −0.43 (−34.34, 33.47) for MGC, −0.57 (−11.90, 10.76) for SD, 0.76 (−6.23, 7.74) for MAGE, −5.59 (−110.82, 99.64) for GLI and −1034.74 (−13524.05, 11454.57) for J-index. Conclusions: The most reliable GV indices (highest correlation and lowest bias) are GVI and SD. These are the least influenced by the timing of measurement.

Results: Data from 88 patients were analysed. For each index, there was a significant correlation (p < 0.0001) between values calculated from intermittent readings and CGM. However, correlations were higher for J-index (0.96), SD (0.95) and GVI (0.9) than for MGC (0.88), GLI (0.82) and MAGE (0.77). The mean difference and limits of agreement (lower limit, upper limit) observed in Bland-Altman plot were 0.01 (−0.10, 0.12) for GVI, −0.43 (−34.34, 33.47) for MGC, −0.57 (−11.90, 10.76) for SD, 0.76 (−6.23, 7.74) for MAGE, −5.59 (−110.82, 99.64) for GLI and −1034.74 (−13524.05, 11454.57) for J-index. Conclusions: The most reliable GV indices (highest correlation and lowest bias) are GVI and SD. These are the least influenced by the timing of measurement. A517 Effects of near-continuous glucose monitoring as a guide for glycemic control: a cluster-randomized study J.-C. Preiser, O. Lheureux, A. Thooft, S. Brimioulle, J.-L. Vincent Erasme University Hospital, Université Libre de Bruxelles, Brussels, Belgium Correspondence: J.-C. Preiser - Erasme University Hospital, Université Libre de Bruxelles, Brussels, Belgium Introduction: Each of the three domains of dysglycemia, i.e. hyperglycemia, hypoglycemia and high glycemic variability is associated with poor outcome. The glycemic control with insulin to maintain blood glucose (BG) within a narrow range as currently recommended involves repeated checks of BG. Near-continuous intravascular monitoring (CGM) can represent a useful tool to facilitate and to improve glycemic control.

glycemic variability is associated with poor outcome. The glycemic control with insulin to maintain blood glucose (BG) within a narrow range as currently recommended involves repeated checks of BG. Near-continuous intravascular monitoring (CGM) can represent a useful tool to facilitate and to improve glycemic control. Objectives: To assess the effects of the use of CGM on the quality and safety of glycemic control. Methods: Adult critically ill patients expected to stay at least 3 days in the department of intensive care of the Erasme University Hospital in Brussels, Belgium and presenting an hyperglycemia (BG > 150 mg/dl) up to 6 hours after admission and / or ongoing insulin therapy with a patent access to a large peripheral vein were equipped with a catheter connected to a enzymatic CGM sensor (GlucoClear®, Edwards Lifesciences, Irvine, CA). Two of the 4 units of the department were randomized to adjust insulin infusion rate to keep BG between 90 and 150 mg/dl (dynamic scale) using the values displayed on the monitor (unblinded group, U). In the 2 other units (blind group (B)), no value was displayed on the screen. The quality of glycemic control assessed by the proportion of time in range (TIR) and the safety assessed by the percentage of time spent below 70 mg/dl (TB70) were calculated from the values recorded by the CGM for each day and for the whole study period and compared between the two groups by ANOVA.

yed on the screen. The quality of glycemic control assessed by the proportion of time in range (TIR) and the safety assessed by the percentage of time spent below 70 mg/dl (TB70) were calculated from the values recorded by the CGM for each day and for the whole study period and compared between the two groups by ANOVA. Results: Seventy-seven eligible patients (age 61 ± 8 years, 56 males, medical admission 74 %, APACHE II 24 ± 5, ICU mortality 23 %) consented to participate and were included in the U group (n = 39) or the B group (n = 38). 43107 BG values (22721 in the U group and 20386 in the B group) were recorded during 1 day (n = 77), 2 days (n = 64) or 3 days (n = 40). TIR did not differ between groups (70 ± 27 (U) vs 73 ± 23 % (B)), nor for any of the days of recording. However, the TB70 was lower in the U group than in the B group (0.4 ± 0.9 vs 1.6 ± 3.4 %, p < .05). Conclusions: The use of a CGM-based strategy improved the safety of glycemic control in a mixed population of ICU patients with stress hyperglycemia.

Results: Seventy-seven eligible patients (age 61 ± 8 years, 56 males, medical admission 74 %, APACHE II 24 ± 5, ICU mortality 23 %) consented to participate and were included in the U group (n = 39) or the B group (n = 38). 43107 BG values (22721 in the U group and 20386 in the B group) were recorded during 1 day (n = 77), 2 days (n = 64) or 3 days (n = 40). TIR did not differ between groups (70 ± 27 (U) vs 73 ± 23 % (B)), nor for any of the days of recording. However, the TB70 was lower in the U group than in the B group (0.4 ± 0.9 vs 1.6 ± 3.4 %, p < .05). Conclusions: The use of a CGM-based strategy improved the safety of glycemic control in a mixed population of ICU patients with stress hyperglycemia. A518 The effects of computer regulated continuous blood glucose management in diabetic patients underwent cardiac surgery H. Iwasaka1, S. Tahara2, M. Nagamine2, A. Ichigatani2 1Oita Almeida Hospital, Intensive Care Unit, Oita, Japan; 2Oita Almeida Hospital, Anesthesiology, Oita, Japan Correspondence: H. Iwasaka - Oita Almeida Hospital, Intensive Care Unit, Oita, Japan Introduction: It is well known that poor preoperative blood glucose control is associated with poor postoperative blood glucose control and postoperative complications. Postoperative hyperglycemia is common in patients underwent cardiac surgery, especially in those with a preoperative elevated HbA1c. The STG-55 (Nikkiso CO., Ltd., Tokyo, Japan) is a newly developed and commercialized computer regulated continuous blood glucose management apparatus.

rol and postoperative complications. Postoperative hyperglycemia is common in patients underwent cardiac surgery, especially in those with a preoperative elevated HbA1c. The STG-55 (Nikkiso CO., Ltd., Tokyo, Japan) is a newly developed and commercialized computer regulated continuous blood glucose management apparatus. Objectives: To compare the usefulness and the workload of ICU nurses of the STG-55 with the conventional sliding scale method in diabetic patients underwent cardiac surgery. Methods: The study was approved by the Institutional Review Board, and written, informed consent was obtained from all participants. This study included 28 patients underwent elective cardiac surgery with preoperative HbA1c > 6.0 %. The management of blood glucose was performed from immediately after admission to the ICU for two research days. The patients were randomly assigned to control blood glucose levels with the STG-55 group (n = 16) or sliding scale method, SS group (n = 12). The usefulness of blood glucose management defined the incidence of hyperglycemia (blood glucose > 180 mg/dL), hypoglycemia (<80 mg/dL), and the maximum glycemic variability (maximum blood glucose level minus minimum blood glucose level). The workload of ICU nurses defined the number of blood samplings for the management of blood glucose and the number of calls made to the physician.

f hyperglycemia (blood glucose > 180 mg/dL), hypoglycemia (<80 mg/dL), and the maximum glycemic variability (maximum blood glucose level minus minimum blood glucose level). The workload of ICU nurses defined the number of blood samplings for the management of blood glucose and the number of calls made to the physician. Results: The two groups were comparable with respect to age, gender, height, weight, duration of surgery, the amount of blood loss during surgery and the preoperative HbA1c. The blood glucose levels determined STG-55 and ABL3 acid–base laboratory analyzer were strongly correlated (R2 = 0.936), with nearly identical values. The incidence of hyperglycemia and hypoglycemia was significantly lower in the STG-55 group (1.9 ± 1.3 vs 6.7 ± 4.9 times/day, p < 0.001; 0.2 ± 0.1 vs 5.8 ± 6.0 times/day, p < 0.001) than SS group. The maximum glycemic variability was also significantly lower in the STG-55 group (46 ± 32 vs 283 ± 165 mg/dL, p < 0.01). The frequency of blood samplings (4.2 ± 3.7 vs 21.5 ± 14.1 times/day, p < 0.01), and the number of calls made to physician (2.1 ± 2.2 vs 6.9 ± 5.4 times/day, p < 0.05) were significantly lower in the STG-55 group than SS group. Conclusions: Use of the STG-55 in the ICU contributed to improved blood glucose management and reduced workload of ICU nurses compared to using the sliding scale method. References 1. Hasegawa A, et al. Surg Today 2001, 41: 1385–1390. Grant acknowledgement None.

Results: The two groups were comparable with respect to age, gender, height, weight, duration of surgery, the amount of blood loss during surgery and the preoperative HbA1c. The blood glucose levels determined STG-55 and ABL3 acid–base laboratory analyzer were strongly correlated (R2 = 0.936), with nearly identical values. The incidence of hyperglycemia and hypoglycemia was significantly lower in the STG-55 group (1.9 ± 1.3 vs 6.7 ± 4.9 times/day, p < 0.001; 0.2 ± 0.1 vs 5.8 ± 6.0 times/day, p < 0.001) than SS group. The maximum glycemic variability was also significantly lower in the STG-55 group (46 ± 32 vs 283 ± 165 mg/dL, p < 0.01). The frequency of blood samplings (4.2 ± 3.7 vs 21.5 ± 14.1 times/day, p < 0.01), and the number of calls made to physician (2.1 ± 2.2 vs 6.9 ± 5.4 times/day, p < 0.05) were significantly lower in the STG-55 group than SS group. Conclusions: Use of the STG-55 in the ICU contributed to improved blood glucose management and reduced workload of ICU nurses compared to using the sliding scale method. References 1. Hasegawa A, et al. Surg Today 2001, 41: 1385–1390. Grant acknowledgement None. A519 A physicochemical approach to acid–base balance in critically ill patients after infusion of seven different types of balanced fluids A. Rugerio Cabrera, E. Monares Zepeda, J. Franco Granillo, J.S. Aguirre Sánchez, A.A. Tanaka Montoya, A. Pedraza Montenegro, G.A. Gálvez Blanco, C.M. Coronado Robles The Americam British Cowdray Medical Center, Critical Care Department 'Dr. Mario Shapiro', Mexico City, Mexico Correspondence: A. Rugerio Cabrera - The Americam British Cowdray Medical Center, Critical Care Department 'Dr. Mario Shapiro', Mexico City, Mexico Introduction: Acid–base status in a body fluid is physically determined by several “independent variables”. These are: PCO2, the “strong ion difference” (SID) and all the strong anions (among them is Cl-), and concentrations of nonvolatile weak acids (Atot). Normal acid–base status is achieved when the independent variables have normal (empirically established) values. The Simplified Fencl-Stewart´s Method can be used at the bedside of the patient and is more accurate for the assessment of acid–base balance. Omron, E. developed a physicochemical model of the projected change in standard base excess (SBE) as a consequence of infused crystalloid solutions of common use (isotonic saline and balanced fluids); unfortunately this was a clinical simulation at standard physiological state. In addition, Kaplan, L. evaluated acid–base status after the administration of balanced fluids in trauma patients. Nevertheless, to our knowledge, there are no other clinical trials that evaluate de administration of other types of balanced fluids.

unately this was a clinical simulation at standard physiological state. In addition, Kaplan, L. evaluated acid–base status after the administration of balanced fluids in trauma patients. Nevertheless, to our knowledge, there are no other clinical trials that evaluate de administration of other types of balanced fluids. Objective: To assess acid–base status of critically ill patients after the infusion of seven different types of balanced solutions Methods: This was a retrospective, observational and descriptive study, conducted in an intensive care unit of a tertiary care hospital. We included all patients above 18 years old admitted to this department from January 2014 to December 2015. We evaluated the effects on acid–base balance after the infusion of seven different solutions: 1) Hartmann + 17.8 mEq/L sodium bicarbonate (NaHCO3) (SID 45.8), 2) Hartmann + 8.9 mEq/L NaHCO3 (SID 36.9), 3) Hartmann + 15 mEq/L NaHCO3 (SID 43), 4) Hartmann + 25 mEq/L NaHCO3 (SID 53), 5) Hartmann (SID 28), 6) normal saline 0.45 % + 77 mEq/L NaHCO3 (SID 75), and 7) dextrose solution 5 % + 154 mEq/L NaHCO3 (SID 154). Arterial blood gases, serum electrolytes, and proteins were measured in the same blood sample. Also SIDa, SEDe, SIG, ATOT, pCO2, change in standard base excess (SBE), pH, [HCO3], [Na]p and SOFA were calculated. pH, SBE and PCO2 were estimated with the ABL8000 FLEX blood gas analyzer. Data are mean ± SD or percents. We used data analysis package SSPS.

trolytes, and proteins were measured in the same blood sample. Also SIDa, SEDe, SIG, ATOT, pCO2, change in standard base excess (SBE), pH, [HCO3], [Na]p and SOFA were calculated. pH, SBE and PCO2 were estimated with the ABL8000 FLEX blood gas analyzer. Data are mean ± SD or percents. We used data analysis package SSPS. Results: Ninety-nine patients were included. Of these, 54 % were women and 45 % men. The most used solutions were Hartmann (25 %), Hartmann + 8.9 mEq/L NaHCO3 (21 %), and Hartmann + 25 mEq/l NaHCO3 (18 %). Before the infusion, SIDe was under 30 mEq/L in 30 % of patients and above in 23 % of them. The effect on the SIDe was significant before the infusion of different solutions (p 0.01), SIDe > 30 ± 8 mEq/L. No metabolic alkalosis or greater decrease of SIDa/SIDe was observed. Conclusions: This study assesses additional varieties of fluids that have a different SID in the clinical setting. No major acid–base disturbances were observed. References 1. Vladimir, F. jabor, A. et al. Diagnosis of metabolic acid–base disturbances in critically ill patients. Am J Respir Crit Care Med. 2000; 162:2246–2251.

Conclusions: This study assesses additional varieties of fluids that have a different SID in the clinical setting. No major acid–base disturbances were observed. References 1. Vladimir, F. jabor, A. et al. Diagnosis of metabolic acid–base disturbances in critically ill patients. Am J Respir Crit Care Med. 2000; 162:2246–2251. A520 Acid–base disturbances and their clinical relevance in critically ill patients with acute-on-chronic liver failure - a preliminary analysis A. Drolz1,2, T. Horvatits1,2, K. Roedl1,2, K. Rutter1,2, S. Kluge2, G.C. Funk3, B. Schneeweiss1, V. Fuhrmann1,2 1Medical University of Vienna, Internal Medicine III, Gastroenterology and Hepatology, Vienna, Austria; 2Medical University Center Hamburg-Eppendorf, Intensive Care Medicine, Hamburg, Germany; 3Otto-Wagner-Spital, Respiratory and Critical Care Medicine, Vienna, Austria Correspondence: A. Drolz - Medical University Center Hamburg-Eppendorf, Intensive Care Medicine, Hamburg, Germany Introduction: Severe acid–base balance abnormalities are frequently observed in critically ill patients at the intensive care unit (ICU) and can be present in different patterns. Acid–base profiles of critically ill cirrhotic patients with acute-on-chronic liver failure (ACLF) have not been investigated. Objectives: To assess disturbances of acid–base status in critically ill patients with ACLF in comparison to patients without.

A520 Acid–base disturbances and their clinical relevance in critically ill patients with acute-on-chronic liver failure - a preliminary analysis A. Drolz1,2, T. Horvatits1,2, K. Roedl1,2, K. Rutter1,2, S. Kluge2, G.C. Funk3, B. Schneeweiss1, V. Fuhrmann1,2 1Medical University of Vienna, Internal Medicine III, Gastroenterology and Hepatology, Vienna, Austria; 2Medical University Center Hamburg-Eppendorf, Intensive Care Medicine, Hamburg, Germany; 3Otto-Wagner-Spital, Respiratory and Critical Care Medicine, Vienna, Austria Correspondence: A. Drolz - Medical University Center Hamburg-Eppendorf, Intensive Care Medicine, Hamburg, Germany Introduction: Severe acid–base balance abnormalities are frequently observed in critically ill patients at the intensive care unit (ICU) and can be present in different patterns. Acid–base profiles of critically ill cirrhotic patients with acute-on-chronic liver failure (ACLF) have not been investigated. Objectives: To assess disturbances of acid–base status in critically ill patients with ACLF in comparison to patients without. Methods: Preliminary analysis of patients admitted to medical ICU at the Medical University of Vienna. Presence of cirrhosis and ACLF, respectively, was assessed, and blood for laboratory- and blood gas analysis was drawn from an arterial line. Reference values were obtained from blood samples of healthy volunteers (n = 14). Mortality was assessed on site.

nts admitted to medical ICU at the Medical University of Vienna. Presence of cirrhosis and ACLF, respectively, was assessed, and blood for laboratory- and blood gas analysis was drawn from an arterial line. Reference values were obtained from blood samples of healthy volunteers (n = 14). Mortality was assessed on site. Results: A total of 615 patients were studied, 142 of these patients had underlying liver cirrhosis. One hundred twenty-eight patients fulfilled criteria of ACLF. Median SOFA score of the total cohort was 10 (IQR 6–13), median SAPS II score 53 (IQR 38–70) and median age 60 (IQR 49–69) years. Patients with liver cirrhosis showed a marked metabolic acidosis compared to critically ill patients without (Base excess (BE) -6.8 (IQR −12.3 to 0.8) mmol/l vs. -1.5 (IQR −5.2 to 2.7) mmol/l, p < 0.001). Analysis of the BE subcomponents revealed that this acidosis was primarily attributable to hyperchloremia (BECl = −4.6 (IQR −7.5 to 0.8) mmol/l), lactate (BELactate = −2.7 (IQR −6.1 to −1.0) mmol/l), and unmeasured anions (BEUMA = −1.6 (IQR −6.2 to 1.8) mmol/l). Interestingly, hyperchloremic acidosis antagonized by hypoalbuminemic alkalosis was found in critically ill patients with and without liver disease. In patients with cirrhosis, we observed respiratory alkalosis (partial pressure of arterial carbon dioxide (PaCO2) = 36.4 (IQR 28.5 - 44.5) mmHg) counteracting metabolic acidosis.

ingly, hyperchloremic acidosis antagonized by hypoalbuminemic alkalosis was found in critically ill patients with and without liver disease. In patients with cirrhosis, we observed respiratory alkalosis (partial pressure of arterial carbon dioxide (PaCO2) = 36.4 (IQR 28.5 - 44.5) mmHg) counteracting metabolic acidosis. With progression to ACLF, the compensatory mechanisms (hyperventilation and hypoalbuminemia) became insufficient resulting in progression of metabolic acidosis. Accordingly, BE in patients with ACLF grade 3 was −10.1 (IQR −17.4 to −3.6) mmol/l with a resulting pH of 7.29 (IQR 7.14 to 7.41). Lactate (BELactate = −3.6 (IQR −8.3 to −1.5) mmol/l) and unmeasured anions (BEUMA = −3.9 (IQR −8.9 to 0.9) mmol/l) were main contributors to metabolic acidosis in ACLF grade 3. The extent of metabolic derangement attributable to lactate and unmeasured anions, respectively, in patients with cirrhosis was associated with 28-day-mortality (Fig. 44). Conclusions: Metabolic acidosis is a common feature of cirrhosis and ACLF at the ICU and is more severe compared to patients without cirrhosis. Acidosis in ACLF is primarily attributable to lactate and unmeasured anions. Although hyperchloremic acidosis did contribute to metabolic acidosis in critically ill patients, we found no clinically relevant difference in chloride-related acidosis between patients with and without liver disease.Fig. 44 (abstract A520). Kaplan Meier Plot: UMA means unmeasured anions

lactate and unmeasured anions. Although hyperchloremic acidosis did contribute to metabolic acidosis in critically ill patients, we found no clinically relevant difference in chloride-related acidosis between patients with and without liver disease.Fig. 44 (abstract A520). Kaplan Meier Plot: UMA means unmeasured anions A521 Unmeasured anions in deceased donor: can they predict liver transplantation outcome? G. Sabetian1, F. Pooresmaeel2, F. Zand3, S. Ghaffaripour2, A. Farbod2, H. Tabei2 1Shiraz University of Medical Sciences, Trauma Research Center, Shiraz, Islamic Republic of Iran; 2Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran; 3Shiraz University of Medical Sciences, Anesthesiology and Critical Care Research Center, Shiraz, Islamic Republic of Iran Correspondence: G. Sabetian - Shiraz University of Medical Sciences, Trauma Research Center, Shiraz, Islamic Republic of Iran Introduction: Liver transplantation is the treatment of choice in end stage liver disease, but significant mortality rate and organ shortage has resulted in multiple attempts to chose good donors and appropriate recipients. (1) The Stewart´s approach is a new concept in critical ill patients and can identify occult acid–base disturbances. (2,) Although study on Stewart´s approach were increased in recent years but its diagnostic and prognostic advantages remain contraversial. (3)

e attempts to chose good donors and appropriate recipients. (1) The Stewart´s approach is a new concept in critical ill patients and can identify occult acid–base disturbances. (2,) Although study on Stewart´s approach were increased in recent years but its diagnostic and prognostic advantages remain contraversial. (3) Objectives: Our primary objective was correlation between SIG (strong ion difference) in donor and graft function. Secondary objective was comparison between SIG and marginal graft criteria as a predictor for liver transplantation outcome. Methods: In 51 donors, marginal graft criteria were determined. SIG in all cases were calculated and we followed mortality and complications in recepiants at 72 hr,1 week, 1 month after transplantation. Results: 54.9 % of livers were marginal. Mean MELD Score in recipients was 21.25 ± 5.95 . According to calculated SIG value, recipients were divided into SIG > 10 and SIG < 10. This study showed that no difference was seen in these two groups(pvalue; 0.154) .Also no difference was seen in two groups.(marginal and non-marginal)(Pvalue = 0.245)Conclusions: In this study, SIG was not a good predictor for liver transplantation outcome. References 1)Shaked A,Nones F,Olthoff KM,Lucey MR.Assessment of liver function:Pre and peritransplant evaluation.Clin Chem 1997; 43: 1539–1545. 2)Ali Y,Abouelnaga S,Khalaf H,Kamel Y.Physical chemical approach versus traditional technique in analyzing blood gases and electrolytes during liver transplant surgery.Transplantation Proceedings2010;42:861–864.

1)Shaked A,Nones F,Olthoff KM,Lucey MR.Assessment of liver function:Pre and peritransplant evaluation.Clin Chem 1997; 43: 1539–1545. 2)Ali Y,Abouelnaga S,Khalaf H,Kamel Y.Physical chemical approach versus traditional technique in analyzing blood gases and electrolytes during liver transplant surgery.Transplantation Proceedings2010;42:861–864. 3)Dubin A,Menises MM,Masevicius FD,Moseinco MC,Daniela KDM,Ventrice E,et al. Comparison of three different methods of evaluation of metabolic acid–base disorder.Crit Care Med 2007;35(27):1264–1270. Grant acknowledgement Financial support was exclusively provided by Shiraz University of Medical Sciences. Ethics of end-of-life-care A522 End of life care in haematology patients: a job for critical care outreach services? L. Taheri, R. Anandanadesan, V. Metaxa 1King's College Hospital (Denmark Hill), Intensive Care Medicine, London, UK Correspondence: L. Taheri - King's College Hospital (Denmark Hill), Intensive Care Medicine, London, UK Introduction: Despite recent medical advances and improved survival, patients with haematological malignancies (HM) still have a high mortality, and end-of-life (EoL) care has become an integral part of their treatment.1 Evidence suggests less palliative care involvement and limited advanced care planning in patients with HM, probably because of the unclear transition between curative and palliative phases of their disease.1,2 Critical Care Outreach Services (CCOS) have been shown to participate in more than 50 % of EoL planning as part of their everyday workload.3

ive care involvement and limited advanced care planning in patients with HM, probably because of the unclear transition between curative and palliative phases of their disease.1,2 Critical Care Outreach Services (CCOS) have been shown to participate in more than 50 % of EoL planning as part of their everyday workload.3 Objectives: To evaluate the role and input of CCOS in EoL care of patients with HM. Methods: We retrospectively reviewed the records of all patients with HM from January 2014 to October 2015 who were referred to CCOS in a London specialist hospital. Variables analysed included age, diagnosis, ICU admission, time spent by CCOS, interventions provided and in-hospital mortality. Results: There were 145 patients who were reviewed on 257 different occasions. Their age ranged from 18 to 84 years (median 56) and their diagnoses are shown in Fig. 45. Of those, 16/145 (11 %) patients were identified as palliative; 81 % of them received EoL care on the ward. National Early Warning Scores at referral ranged from 2 to 10 (median 7). A total of 723 days was spent by CCOS in reviewing patients with HM, with 146 days (20 % of their clinical time) spent on palliative patients alone (median time 5.5d vs. 3d for non-palliative patients). Overall, in-hospital mortality among these patients was 81 %.

ning Scores at referral ranged from 2 to 10 (median 7). A total of 723 days was spent by CCOS in reviewing patients with HM, with 146 days (20 % of their clinical time) spent on palliative patients alone (median time 5.5d vs. 3d for non-palliative patients). Overall, in-hospital mortality among these patients was 81 %. The services provided by CCOS to palliative patients were mainly facilitation of symptom control (67 %) and/or support of ward teams in making treatment limitation decisions (78 %). High Flow Nasal Cannula Oxygen (HFNC) was initiated in 38 % of patients for symptom control, which was deemed successful in all but 2 patients. Conclusions: A considerable part of the CCOS workload is spent on supporting the management of palliative haematology patients. Involvement in alleviating symptoms and initiating EoL discussions features strongly in the interventions requested by the ward teams. The significant use of HFNC therapy for symptom control that was observed in this population also warrants further investigation. References 1. Howell DA et al. Haematological malignancy: are patients appropriately referred for specialist palliative and hospice care? A systematic review and meta-analysis of published data. Palliative Medicine 2010;25:630–41 2. National Institute for Clinical Excellence. Improving outcomes in haematological cancers: the manual. London: National Institute for clinical Excellence. October 2003

1. Howell DA et al. Haematological malignancy: are patients appropriately referred for specialist palliative and hospice care? A systematic review and meta-analysis of published data. Palliative Medicine 2010;25:630–41 2. National Institute for Clinical Excellence. Improving outcomes in haematological cancers: the manual. London: National Institute for clinical Excellence. October 2003 3. Pattison N et al. Negotiating Transitions: Involvement of Critical Care Outreach Teams in End-of-Life Decision Making. Am J Crit Care 2015;24:232–40Fig. 45 (abstract A522). ᅟ A523 Contact with death, ethical decisions, and communication of bad news in intensive care and palliative units: results from a mixed-methods study C. Teixeira1,2,3, S.M. Pereira1, P. Hernández-Marrero1,4, A.S. Carvalho1 1Universidade Católica Portuguesa, Instituto de Bioética, Porto, Portugal; 2Centro Hospitalar do Porto, Hospital de Santo António, UCIP-Departamento de Anestesia e Cuidados Intensivos, Porto, Portugal; 3Universidade do Porto, Instituto de Ciências Biomédicas Dr. Abel Salazar, Porto, Portugal; 4Servicio Canario de Salud, Porto, Portugal Correspondence: C. Teixeira - Universidade do Porto, Instituto de Ciências Biomédicas Dr. Abel Salazar, Porto, Portugal Introduction: Professionals working in intensive care and palliative units (ICUs/PCUs) care for patients with life-threatening diseases, make ethical decisions, and provide end-of-life care. However, while palliative care aims to reduce suffering, intensive care has a major focus on saving lives.

lazar, Porto, Portugal Introduction: Professionals working in intensive care and palliative units (ICUs/PCUs) care for patients with life-threatening diseases, make ethical decisions, and provide end-of-life care. However, while palliative care aims to reduce suffering, intensive care has a major focus on saving lives. Objectives: To identify and compare the experiences of ICU and PCU healthcare professionals related to: contact with dying and death, making of ethical decisions and communication and delivery of bad news. Methods: Mixed approach, combining quantitative (questionnaire on experiences in the work context) and qualitative ones (interviews with doctors and nurses). 10 ICU and 9 PCU participated in this study. 392 professionals completed the survey; 28 were interviewed. A descriptive quantitative analysis was performed; the chi-square test was used to analyse the association between variables (significance level of p < .05). Interviews were subject to content analysis.

nd nurses). 10 ICU and 9 PCU participated in this study. 392 professionals completed the survey; 28 were interviewed. A descriptive quantitative analysis was performed; the chi-square test was used to analyse the association between variables (significance level of p < .05). Interviews were subject to content analysis. Results: In the week prior to survey completion, more professionals working in ICUs reported a patient's death; this was not statistically significant. The experience most mentioned by the professionals of both types of units during interviews was caring for patients nearing death. In the week before completing the questionnaire, the most common ethical decision was palliative/terminal sedation; this was more, often in ICUs (27 % vs. 12 %; p = .004). In the day of questionnaire completion, the most frequent ethical decision was also palliative sedation. Though this decision was more frequent in ICU, statistical significance was not reached (p = .440). The communication of the diagnosis/prognosis to the patient, either in the week before or in the day of questionnaire completion was more frequent in PCUs (45 % vs. 29 %, p = .005; 22 % vs. 12 %, p = .026, respectively). Communication about the diagnosis/prognosis with the family in the week before survey completion was held with equal frequency by professionals from both contexts (58 % of professionals). Although not reaching statistical significance (p = .303), more professionals from PCU proceeded to communication with family about the diagnosis and prognosis (32 % vs. 26 %) in the survey day. From the analysis of the interviews, it was denoted that it were mainly professionals of PCU who referred to the communication on the diagnosis/prognosis, both with the patient as with the family.

ionals from PCU proceeded to communication with family about the diagnosis and prognosis (32 % vs. 26 %) in the survey day. From the analysis of the interviews, it was denoted that it were mainly professionals of PCU who referred to the communication on the diagnosis/prognosis, both with the patient as with the family. Conclusions: The workplace experiences in ICU and PCU are, despite some differences, guided by similarities. Caring for patients with life-threatening situations and imminent death and the need to make ethical decisions occur frequently in both contexts. The communication about the diagnosis/prognosis occurs more often in PCU. This highlights the need for integrating communication strategies of palliative care, in intensive care. References 1. Curtis JR, Vincent JL. Ethics and end-of-life care on the adult intensive care unit. Lancet. 2010 Oct 16;376(9749):1347–53. Epub 2010 Oct 11.

Conclusions: The workplace experiences in ICU and PCU are, despite some differences, guided by similarities. Caring for patients with life-threatening situations and imminent death and the need to make ethical decisions occur frequently in both contexts. The communication about the diagnosis/prognosis occurs more often in PCU. This highlights the need for integrating communication strategies of palliative care, in intensive care. References 1. Curtis JR, Vincent JL. Ethics and end-of-life care on the adult intensive care unit. Lancet. 2010 Oct 16;376(9749):1347–53. Epub 2010 Oct 11. A524 An opportunity for advance decision-making: pre-operative risk stratification for adverse ICU outcomes in elective surgical patients M. Beckmann1, C.S. Hartog2, D. Schwarzkopf2, A. Raadts1 1Jena University Hospital, Anaesthesiologyy and Intensive Care, Jena, Germany; 2Jena University Hospital, Center for Sepsis Control and Care, Jena, Germany Correspondence: M. Beckmann - Jena University Hospital, Anaesthesiologyy and Intensive Care, Jena, Germany Introduction: Complex elective surgery is increasingly provided to an aging population with heightened risk for prolonged ICU stay or death. While risks of surgical and anaesthetic procedures are routinely discussed preoperatively in the process of obtaining informed consent, the patients' preferences in case of unfavourable ICU outcomes are rarely discussed and most critically ill patients do not have advance directives. If patients at risk of a prolonged ICU stay or death could be identified preoperatively, this would provide an opportunity to make their preferences known in a timely and adequate manner.

eferences in case of unfavourable ICU outcomes are rarely discussed and most critically ill patients do not have advance directives. If patients at risk of a prolonged ICU stay or death could be identified preoperatively, this would provide an opportunity to make their preferences known in a timely and adequate manner. Objectives: To identify the predictive validity of routine pre-operative risk assessment for an unfavourable ICU course (ICU stay > 24 hours or death), and to assess the prevalence of advance directives in these patients. Methods: Among all 13437 elective adult surgical cases seen in a tertiary university hospital´s pre-operative anaesthesiology clinic in 2014, 1832 consecutive cases were drawn. Data were extracted from hospital and ICU databases, including patient demographics, pre-operative American Society of Anesthesiologists (ASA) classification, length of ICU stay, mortality and presence of advance directives. A receiver operating characteristic analysis was conducted to test the predictive validity of ASA for a) having an ICU stay > 24 h and b) dying in the ICU or afterwards. Optimal cut-offs were identified by maximum Youden's J statistic.

A) classification, length of ICU stay, mortality and presence of advance directives. A receiver operating characteristic analysis was conducted to test the predictive validity of ASA for a) having an ICU stay > 24 h and b) dying in the ICU or afterwards. Optimal cut-offs were identified by maximum Youden's J statistic. Results: Among 1832 patients, 937 (51 %) were male, median age was 63 years (interquartile range 49–74), planned procedures were mainly from General, Trauma, Cardiothoracic, Eye, Gynecological, Urological or ENT Surgery. Pre-operatively, patients were classified into ASA risk classes (15 %, 41 %, 40 % and 4 % into ASA 1–4, respectively). Postoperatively, 504 (28 %) patients were admitted to the ICU. Of these, 373 (74 %) had an ICU LOS > 24 hours and 68 (13 %) died in the ICU or afterwards. Among patients with an ICU stay > 24 hours, presence of an AD was documented in 49 (15 %) and power of attorney for a legal proxy in 71 (22 %). Pre-operative ASA classification predicted an ICU stay >24 h with an area under the curve (AUC) of 0.79, and death in the ICU or afterwards with an AUC of 0.85. The optimal cut-off for both adverse outcomes was ASA ≥ 3. For ICU stay >24 h sensitivity was 0.85, specificity was 0.67; for death in or after ICU sensitivity was 0.93 and specificity was 0.58.

edicted an ICU stay >24 h with an area under the curve (AUC) of 0.79, and death in the ICU or afterwards with an AUC of 0.85. The optimal cut-off for both adverse outcomes was ASA ≥ 3. For ICU stay >24 h sensitivity was 0.85, specificity was 0.67; for death in or after ICU sensitivity was 0.93 and specificity was 0.58. Conclusions: Preoperative ASA classification with a cut-off of ≥ 3 has a good predictive validity to identify the risk of prolonged ICU stay or death in patients undergoing elective surgery. Such risk stratification could be useful to initiate advance decision-making at the time of the pre-operative work-up of these patients and thus increase the prevalence of adequate ADs in the ICU. Grant acknowledgement Funded partially by the Federal Ministry of Education and Research (BMBF), Germany, grant no: 01EO1002. A525 Treatment-limiting-decisions in patients with severe traumatic brain injury in a Norwegian trauma hospital A. Robertsen1, R. Førde2, N.-O. Skaga1, E. Helseth3 1Oslo University Hospital, Anesthesiology and critical care, Oslo, Norway; 2University of Oslo, Center of medical ethics, Oslo, Norway; 3Oslo University Hospital, Neurosurgery, Oslo, Norway Correspondence: A. Robertsen - Oslo University Hospital, Anesthesiology and critical care, Oslo, Norway Introduction: Studies have shown variations in practice across hospitals regarding treatment-limitations and mortality for brain-injured patients.

, Oslo, Norway; 3Oslo University Hospital, Neurosurgery, Oslo, Norway Correspondence: A. Robertsen - Oslo University Hospital, Anesthesiology and critical care, Oslo, Norway Introduction: Studies have shown variations in practice across hospitals regarding treatment-limitations and mortality for brain-injured patients. Objectives: To study treatment-limitations and associated mortality in a Norwegian trauma hospital and documentation of ethical aspects such as presence of advanced directives, dialogue with families, multi-team discussions, reasons and considerations behind decisions, conflicts and involvement of clinical ethics committees. Methods: A retrospective study of a 2-year cohort of severe head injured patients admitted 2011–12 to Oslo university hospital, Norway. Trauma registry data were combined with data from medical records. For data validation a definition guide for study variables was developed. Adults with abbreviated injury score head 4,5,6 were included (n = 579).

y of a 2-year cohort of severe head injured patients admitted 2011–12 to Oslo university hospital, Norway. Trauma registry data were combined with data from medical records. For data validation a definition guide for study variables was developed. Adults with abbreviated injury score head 4,5,6 were included (n = 579). Results: Eighty-five % of all patients were admitted to ICU. Treatment limitations were identified in 17 % of cases (101 patients). Decisions were: Withholding organ support (12 cases), withholding surgery (52 cases), withdrawing intracranial pressure-targeted therapy (23 cases), DNR-orders (44 cases), no-escalation of treatment (19 cases) or withdrawing organ support (44 cases). For some patients initial decisions were changed (19 cases) or revoked (3 cases) along the dynamic treatment trajectory. Twenty-six patients with devastating brain injury progressed to brain death. No patients had advanced directives. Dialogue with family was documented in most cases (98). No major conflict between families and treatment team was identified and there was no involvement by CEC. Rationale behind decisions was identified as medical only in 80 % of cases. Treatment-limitations followed situations categorized as futile (59 cases) or “potentially inappropriate treatment” (42 cases) (1). The overall 30-day mortality was 16 % (35 % for patient with GCS < 9 and 7,5 % for patients with GCS >8). Treatment-limitations were identified in 93 % of cases of in-hospital death. In-hospital mortality was 73 %, 30-day mortality 82 % and 2-year mortality 93 % in the treatment-limiting group (n = 101 patients, 25 patients were transferred to other facilities with limitations). In-hospital mortality was 1 %, 30-day mortality 2 % and 2-year mortality 8 % for patients without limitations made at the trauma hospital (n = 478).

3 %, 30-day mortality 82 % and 2-year mortality 93 % in the treatment-limiting group (n = 101 patients, 25 patients were transferred to other facilities with limitations). In-hospital mortality was 1 %, 30-day mortality 2 % and 2-year mortality 8 % for patients without limitations made at the trauma hospital (n = 478). Conclusions: Treatment limitations are common in patients with traumatic brain injury and were closely associated with in-hospital death. Withholding or withdrawing life-sustaining therapy in this early phase of hospitalization was primarily based on the medical situation and at the discretion of the physician. Whether patients' values and preferences were adequately addressed or had an impact on decisions (when appropriate) remains unclear. References 1. Rubin, M. A. and J. Bonomo (2016). "Neurocritical Care Society Views on "Potentially Inappropriate Treatments in Intensive Care Units"." American Journal of Respiratory and Critical Care Medicine 193(4): 466–467. A526 Ethical considerations when performing life saving but non restorative neurosurgical intervention. The oracle stroke study S. Honeybul1, K. Ho2 1Sir Charles Gairdner Hospital, Neurosurgery, Perth, Australia; 2Royal Perth Hospital, Perth, Australia Correspondence: S. Honeybul - Sir Charles Gairdner Hospital, Neurosurgery, Perth, Australia Introduction: This study assessed opinion of healthcare workers about consent and acceptable outcome in the context of decompressive hemicraniectomy for 'malignant cerebral artery infarction”

yal Perth Hospital, Perth, Australia Correspondence: S. Honeybul - Sir Charles Gairdner Hospital, Neurosurgery, Perth, Australia Introduction: This study assessed opinion of healthcare workers about consent and acceptable outcome in the context of decompressive hemicraniectomy for 'malignant cerebral artery infarction” Method: Seven Hundred and seventy three healthcare workers at the two major public neurosurgical centres in Western Australia participated in this study. Participants were asked to record their opinion regarding consent and acceptable outcome based on the Modified Rankin Score (mRS). They were then given a detailed analysis of the evidence for clinical efficacy of the procedure and an explanation of the “disability paradox”. They were then asked to reconsider their initial responses. Results: Of the 773 participants included in the study 407 (52.7 %) initially felt that they would provide consent for a decompressive craniectomy as a life-saving procedure but only a minority of them considered mRS 4 or 5 as an acceptable outcome (for mRS ≤ 4: n = 67, 8.7 %; for mRS = 4: n = 57, 7.4 %). After introducing the concept of the disability paradox and the evidence for the clinical efficacy of decompressive craniectomy, more participants were unwilling to accept decompressive craniectomy (18.1 % vs. 37.8 %) but, at the same time, more were willing to accept mRS ≤ 4 as an acceptable outcome (for mRS ≤4: n = 92, 11.9 %; for mRS = 4: n = 79, 10.2 %).

ability paradox and the evidence for the clinical efficacy of decompressive craniectomy, more participants were unwilling to accept decompressive craniectomy (18.1 % vs. 37.8 %) but, at the same time, more were willing to accept mRS ≤ 4 as an acceptable outcome (for mRS ≤4: n = 92, 11.9 %; for mRS = 4: n = 79, 10.2 %). Conclusion: Most participants felt that survival with dependency to be unacceptable. However, many would appear to be willing to provide consent for surgery in the hope that they may survive with some degree of independence. A527 Analysis of patients assessed but not admitted to ICU of a secondary hospital in Malaga P. Martinez Lopez, M. Nieto Gonzalez, P. Nuevo Ortega, E. Camara Sola, T. Spasova, M.V. de la Torre-Prados Virgen de la Victoria Hospital, Intensive Care, Málaga, Spain Correspondence: M. Nieto Gonzalez - Virgen de la Victoria Hospital, Intensive Care, Málaga, Spain Objective: To analyze why patients who are assessed by an intensivist physician for some reasons are not admitted to ICU. Also we want to find out what happened to this patients after intensivist evaluation. Design: Observational and prospective study during 12 consecutive months. Scope: Second level Hospital with 550 beds and 18 ICU boxes.

A527 Analysis of patients assessed but not admitted to ICU of a secondary hospital in Malaga P. Martinez Lopez, M. Nieto Gonzalez, P. Nuevo Ortega, E. Camara Sola, T. Spasova, M.V. de la Torre-Prados Virgen de la Victoria Hospital, Intensive Care, Málaga, Spain Correspondence: M. Nieto Gonzalez - Virgen de la Victoria Hospital, Intensive Care, Málaga, Spain Objective: To analyze why patients who are assessed by an intensivist physician for some reasons are not admitted to ICU. Also we want to find out what happened to this patients after intensivist evaluation. Design: Observational and prospective study during 12 consecutive months. Scope: Second level Hospital with 550 beds and 18 ICU boxes. Method: We searched daily for patients who were assessed but not admitted to ICU during 12 consecutive months. We write down medical record number, which of the hospital departments asked for that ICU evaluation and the reason why the patient was not finally admitted to ICU (low severity, life-sustaining treatment limitation or not ICU bed availability). We follow up those patients until hospital discharged, finding either decease, hospital discharge or later ICU admission.

he hospital departments asked for that ICU evaluation and the reason why the patient was not finally admitted to ICU (low severity, life-sustaining treatment limitation or not ICU bed availability). We follow up those patients until hospital discharged, finding either decease, hospital discharge or later ICU admission. Results: 1297 patients were assessed but not admitted to ICU. The reasons for not admission were low severity in 736 patients (56.8 %), life-sustaining treatment limitation in 459 patients (35.5 %) and no ICU bed availability in 102 situations (7.9 %). Main hospital departments that asked for assessment without consequent ICU admission were Emergency (N = 899, 69 %), Internal Medicine (N = 140, 10.8 %) and General Surgery (N = 62, 4.8 %). Life-sustaining treatment limitation decision was mostly taken in Emergency Department (N = 253) and Internal Medicine (N = 83), 166 cases survived and were discharged from hospital (36 %). 736 patients were dismissed because of low severity, but 91 were finally admitted to ICU later (12.3 %) and 125 patients died (16.9 %). 102 patients were not admitted because of no ICU bed availability, majority were treated in Observation Room in Emergency Department (N = 94, 92.2 %); 29 of them were admitted to ICU later (28.4 %) and 20 died (19.6 %).

of low severity, but 91 were finally admitted to ICU later (12.3 %) and 125 patients died (16.9 %). 102 patients were not admitted because of no ICU bed availability, majority were treated in Observation Room in Emergency Department (N = 94, 92.2 %); 29 of them were admitted to ICU later (28.4 %) and 20 died (19.6 %). Conclusions: From these results we conclude that maybe the lack of beds in ICU can generate a belated admission to ICU and higher mortality rate in recoverable patients. One third of patients with a life-sustaining treatment limitation decision finally survive the hospitalization. Also we conclude that a considerable amount of intensivist clinical assistance during the shifts takes place outside de ICU.

erate a belated admission to ICU and higher mortality rate in recoverable patients. One third of patients with a life-sustaining treatment limitation decision finally survive the hospitalization. Also we conclude that a considerable amount of intensivist clinical assistance during the shifts takes place outside de ICU. A528 Oddicus - how much care is at odds with patient's values and preferences at the end of life? O. Kopecky1, K. Rusinova1, P. Waldauf2, Z. Cepeplikova1, M. Balik1 1General University Hospital, 1st Faculty of Medicine, Charles University in Prague, Department of Anaesthesia and Intensive Care Medicine, Prague, Czech Republic; 23rd Medical Faculty, Charles University, Department of Anaesthesia and Intensive Care Medicine, Prague, Czech Republic Correspondence: K. Rusinova - General University Hospital, 1st Faculty of Medicine, Charles University in Prague, Department of Anaesthesia and Intensive Care Medicine, Prague, Czech Republic Introduction: Addressing the quality of end of life care has become an important focus of research and patient management in the ICU. Although quality of care and support for patients who die in the ICU and their families are well described, a little is known about the decision making process and the quality of care provided for patients dying outside the ICU. Objectives: To characterize the care for patients dying during the last hospitalization in terms of circumstances of death, provided palliative care, medical decision making, nurses perceptions and family experience.

A528 Oddicus - how much care is at odds with patient's values and preferences at the end of life? O. Kopecky1, K. Rusinova1, P. Waldauf2, Z. Cepeplikova1, M. Balik1 1General University Hospital, 1st Faculty of Medicine, Charles University in Prague, Department of Anaesthesia and Intensive Care Medicine, Prague, Czech Republic; 23rd Medical Faculty, Charles University, Department of Anaesthesia and Intensive Care Medicine, Prague, Czech Republic Correspondence: K. Rusinova - General University Hospital, 1st Faculty of Medicine, Charles University in Prague, Department of Anaesthesia and Intensive Care Medicine, Prague, Czech Republic Introduction: Addressing the quality of end of life care has become an important focus of research and patient management in the ICU. Although quality of care and support for patients who die in the ICU and their families are well described, a little is known about the decision making process and the quality of care provided for patients dying outside the ICU. Objectives: To characterize the care for patients dying during the last hospitalization in terms of circumstances of death, provided palliative care, medical decision making, nurses perceptions and family experience. Methods: A prospective observational cohort study in two tertiary university centers in Prague, Czech Republic. Data from the patient documentation and semi-structured questionnaire with the attending physician and the nurse present at the moment of death were collected. A semi-structured phone interview with the family member was held 3–6 months after the death. The study has received an IRB approval.

in Prague, Czech Republic. Data from the patient documentation and semi-structured questionnaire with the attending physician and the nurse present at the moment of death were collected. A semi-structured phone interview with the family member was held 3–6 months after the death. The study has received an IRB approval. Results: Among 44236 patients hospitalized between February 2015 and February 2016, 1097 has died, of them 926 patients (86,4 %) were enrolled in the study (42.1 % in the ICU and 57.9 % at wards). The basic patients´ characteristic is presented in the Table 20. The majority of deaths in the hospitals were expected and hence not preceded by a cardiopulmonary resuscitation. The mention of a poor prognosis was documented in a majority of cases, the DNR and DNI being the most frequent type of treatment limitation both in ICU and wards. The treatment limitation was communicated with the family fare more often than to the patient (81 % and 31 %, respectively). Patients did not have an advance directives document, nor were their treatment preferences documented during the last hospitalization (0,5 and 1,62 %, respectively). The various aspects of treatment limitations in the ICU vs. wards are presented in Table 21.

more often than to the patient (81 % and 31 %, respectively). Patients did not have an advance directives document, nor were their treatment preferences documented during the last hospitalization (0,5 and 1,62 %, respectively). The various aspects of treatment limitations in the ICU vs. wards are presented in Table 21. Conclusions: This is the first large prospective study comparing the decision making process and the perceptions of clinicians and family members of patients dying in a tertiary care hospital. The study shows an insufficient interdisciplinary and intra-team communication with discrepancies between the documented and the agreed treatment limitations. Furthermore, a difficult communication with patients and their families in final phases of patients´ lives suggests a gap between the care provided and the patients preferences regarding the end-of-life, although these are known to be key aspects of good end-of-life care. References 1. Piers RD, Azoulay E, et al. Inappropriate care in European ICUs: confronting views from nurses and junior and senior physicians. Chest. 2014;146(2):267–75. 2. Rusinova K, Kukal J et al. Limited family members/staff communication in ICUs in the Czech and Slovak Republics considerably increases anxiety in patients´ relatives. BMC Psychiatry. 2014:27;14–21 Grant acknowledgement This study was supported by AVAST foundation.Table 20 (abstract A528). Patient characteristics

1. Piers RD, Azoulay E, et al. Inappropriate care in European ICUs: confronting views from nurses and junior and senior physicians. Chest. 2014;146(2):267–75. 2. Rusinova K, Kukal J et al. Limited family members/staff communication in ICUs in the Czech and Slovak Republics considerably increases anxiety in patients´ relatives. BMC Psychiatry. 2014:27;14–21 Grant acknowledgement This study was supported by AVAST foundation.Table 20 (abstract A528). Patient characteristics Characteristic of the patients N=926 Age (mean; ±SD) 75±13.8 Charleson comorbidity scale (mean;±SD) 6.9 ±2.7 Performance status at hospital admission on the scale from 0 to 4 0–1.24% 1–4.8% 2–9.45% 3–30.3% 4–54.2% Advance directives (%) 0.5 Length of the last hospitalization (median days; IQR) 8 (2–19) Table 21 (abstract A528). Comparison of treatment limitations: ICU vs. wards

Charleson comorbidity scale (mean;±SD) 6.9 ±2.7 Performance status at hospital admission on the scale from 0 to 4 0–1.24% 1–4.8% 2–9.45% 3–30.3% 4–54.2% Advance directives (%) 0.5 Length of the last hospitalization (median days; IQR) 8 (2–19) Table 21 (abstract A528). Comparison of treatment limitations: ICU vs. wards ICU (N= 390) Wards (N=536) Logistic regression, (OR, 95% CI), p Documented poor prognosis 90,7% 88.3% 1.3 (0.8;2.1), p=0.3 Death expected during current hospitalization (perception of the physician) 89,3% 84.6% 1.5 (1.01;2.3), p=0.044 Palliative care provided (perception of the physician) 41,9% 62.8% 0.5 (0.4;0.7), p<0.001 Documented treatment limitation DNR: 88.4%; DNI: 47.6%; no RRT: 31.1%; no catecholamines: 25,8% DNR: 72.9%; DNI: 56,8%; No ICU transfer 19.4%; No antibiotics 6.5% 2.8 (1.8;4.4), p<0.001; 0.7 (0.5;0.95), p=0.024; NA; NA Limitation of treatment never discussed among staff 31,5% 36.6% 0.8 (0.6;1.05), p=0.112 CPR 24h prior death 13,8% 3,7% 3.9 (2.3;6.6), p<0.001 Nurse involved in decision 4.5 4,5% 3.2% 1.4 (0.6;3.1), p=0.385 Info about the treatment limit: − to the family - to the nurse - to the patient 76.3%; 98.3%; 40.0% 81.4%; 96.0%; 31.1% 0.7 (0.5;1.1), p=0.114; 2.4 (0.9;6.9), p=0.098; 1.5 (0.5;4.6), p=0.5 Family present at the moment of death 15,8% 10,7% 1.6 (1.06;2.3), p=0.024

involved in decision 4.5 4,5% 3.2% 1.4 (0.6;3.1), p=0.385 Info about the treatment limit: − to the family - to the nurse - to the patient 76.3%; 98.3%; 40.0% 81.4%; 96.0%; 31.1% 0.7 (0.5;1.1), p=0.114; 2.4 (0.9;6.9), p=0.098; 1.5 (0.5;4.6), p=0.5 Family present at the moment of death 15,8% 10,7% 1.6 (1.06;2.3), p=0.024 A529 Analgesia and sedation after limitation of life sustaining treatment thru terminal extubation. A comparison between controlled cardiac death donors and non-donors J. Palamidessi Domínguez, P. Matia Almudevar, S. Alcántara Carmona, J.J. Rubio Muñoz, D. Palacios Castañeda, A. Naharro Abellán, P. Rodríguez Villamizar, J. Veganzones Ramos, L. Pérez Pérez, A. Pérez Lucendo, M. Camós Ejarque Hospital Universitario Puerta de Hierro Majadahonda, Madrid, Spain Correspondence: J. Palamidessi Domínguez - Hospital Universitario Puerta de Hierro Majadahonda, Madrid, Spain Objective: To compare the different strategies of analgesia and sedation applied in control cardiac death donors (DCD) versus non-donors after limitation of life sustaining treatment (LLST) thru terminal extubation (TE).

ence: J. Palamidessi Domínguez - Hospital Universitario Puerta de Hierro Majadahonda, Madrid, Spain Objective: To compare the different strategies of analgesia and sedation applied in control cardiac death donors (DCD) versus non-donors after limitation of life sustaining treatment (LLST) thru terminal extubation (TE). Methods: Retrospective study (January 2012 - March 2016) in a Spanish tertiary hospital. We studied all patients that underwent LLST thru TE. Demographic factors, general patient characteristics (comorbidities, diagnosis upon ICU admission, APACHE II on admission and SOFA in the previous 24 h before TE), ICU length of stay and timing from TE to cardiac arrest together with their donor/non-donor condition were recorded. We also analyzed the different types and doses of analgesic and sedation drugs (morphine, fentanyl, remifentanyl, propofol and midazolam) and the different strategies used (bolus administration, increase of the infusion rate or both). In order to simplify dosage analysis we used the intervals previously described by other authors1. Statistical analysis with c2 and Mann–Whitney U test. Statistical significance p < 0,05. Results: During the period studied we recovered a total of 68 patients with LLST thru TE that were divided in two groups: DCD (group A, n = 26) and non-donors (group B, n = 42). Demographic data and general characteristic are described in Table 22.

Methods: Retrospective study (January 2012 - March 2016) in a Spanish tertiary hospital. We studied all patients that underwent LLST thru TE. Demographic factors, general patient characteristics (comorbidities, diagnosis upon ICU admission, APACHE II on admission and SOFA in the previous 24 h before TE), ICU length of stay and timing from TE to cardiac arrest together with their donor/non-donor condition were recorded. We also analyzed the different types and doses of analgesic and sedation drugs (morphine, fentanyl, remifentanyl, propofol and midazolam) and the different strategies used (bolus administration, increase of the infusion rate or both). In order to simplify dosage analysis we used the intervals previously described by other authors1. Statistical analysis with c2 and Mann–Whitney U test. Statistical significance p < 0,05. Results: During the period studied we recovered a total of 68 patients with LLST thru TE that were divided in two groups: DCD (group A, n = 26) and non-donors (group B, n = 42). Demographic data and general characteristic are described in Table 22. There were no statistical differences regarding the different types and doses of analgesia and sedation, and the different strategies used in group A vs group B (Table 23 and Fig. 46). Although none of the following reached statistical significance we found that the number of patients receiving doses described as superior to standard accounted for an important percentage in both groups, and that there was a higher rate of bolus administration followed by an increase in infusion rate un group B (53,3 %) when compared to group A (29,4 %).

d statistical significance we found that the number of patients receiving doses described as superior to standard accounted for an important percentage in both groups, and that there was a higher rate of bolus administration followed by an increase in infusion rate un group B (53,3 %) when compared to group A (29,4 %). Conclusion: In our study we found no differences in dose, type and strategy applied for analgesia and sedation after TE regardless of whether the patient was a potential organ donor or not. References 1. Troug RD et al. Recommendations for end-of-life care in the intensive care unit: A consensus statement by the American College of Critical Care Medicine. Crit Care Med 2008 Vol. 36, No. 3Table 22 (abstract A529). Demographic data and general characteristics Group A Group B Age (mean) 53±12 years 66±15 years* Presence of comorbidity 73% 90% (p 0,058) Specific comorbidities: − Cardiovascular risk factors − Heart disease − All other comorbidities without differences between groups 46% 4% (1 patient) 76% 31% Main diagnosis upon ICU admission (without differences between groups) 1*) Cardiac arrest; 42,3% 1*) Cardiac arrest; 38,1% 2*) Ischemic stroke and respiratory pathology: 15,4% 2*) Hemorrhagic stroke; 26,2% 3*) Hemorrhagic stroke: 7,7% 3*) Ischemic stroke: 9,5% APACHE II on admission (mean) 24±6 points 27±8 points SOFA 24 hours before TE (mean) 7±2 points 9±3 points* ICU length of stay (mean) 12±11 days 5±5 days* Timing from TE to cardiac arrest (median) 11 (9 to 17) minutes 18 (10 to 105) minutes* *p<0,05 Table 23 (abstract A529). Analgesia and sedation practices

Group A Group B Age (mean) 53±12 years 66±15 years* Presence of comorbidity 73% 90% (p 0,058) Specific comorbidities: − Cardiovascular risk factors − Heart disease − All other comorbidities without differences between groups 46% 4% (1 patient) 76% 31% Main diagnosis upon ICU admission (without differences between groups) 1*) Cardiac arrest; 42,3% 1*) Cardiac arrest; 38,1% 2*) Ischemic stroke and respiratory pathology: 15,4% 2*) Hemorrhagic stroke; 26,2% 3*) Hemorrhagic stroke: 7,7% 3*) Ischemic stroke: 9,5% APACHE II on admission (mean) 24±6 points 27±8 points SOFA 24 hours before TE (mean) 7±2 points 9±3 points* ICU length of stay (mean) 12±11 days 5±5 days* Timing from TE to cardiac arrest (median) 11 (9 to 17) minutes 18 (10 to 105) minutes* *p<0,05 Table 23 (abstract A529). Analgesia and sedation practices Group A Group B Standard dose 15,8% 16,7%* High dose 84,2% 83,3%* Bolus OR increase of perfusion rate 70,6% 46,7%* Bolus AND increase of perfusion rate 29,4% 53,3%* *p>0,05 Fig. 46 (abstract A529). ᅟ

Group A Group B Age (mean) 53±12 years 66±15 years* Presence of comorbidity 73% 90% (p 0,058) Specific comorbidities: − Cardiovascular risk factors − Heart disease − All other comorbidities without differences between groups 46% 4% (1 patient) 76% 31% Main diagnosis upon ICU admission (without differences between groups) 1*) Cardiac arrest; 42,3% 1*) Cardiac arrest; 38,1% 2*) Ischemic stroke and respiratory pathology: 15,4% 2*) Hemorrhagic stroke; 26,2% 3*) Hemorrhagic stroke: 7,7% 3*) Ischemic stroke: 9,5% APACHE II on admission (mean) 24±6 points 27±8 points SOFA 24 hours before TE (mean) 7±2 points 9±3 points* ICU length of stay (mean) 12±11 days 5±5 days* Timing from TE to cardiac arrest (median) 11 (9 to 17) minutes 18 (10 to 105) minutes* *p<0,05 Table 23 (abstract A529). Analgesia and sedation practices Group A Group B Standard dose 15,8% 16,7%* High dose 84,2% 83,3%* Bolus OR increase of perfusion rate 70,6% 46,7%* Bolus AND increase of perfusion rate 29,4% 53,3%* *p>0,05 Fig. 46 (abstract A529). ᅟ A530 National survey about bioethics quality indicators in spanish intensive care units A. Estella1, V. Lopez Camps2, M.C. Martín3, N. Masnou4, Bioethics work group of SEMICYUC. 1Hospital del SAS de Jerez, Intensive Care Unit, Jerez de la Frontera, Spain; 2Hospital de Sagunto, Intensive Care Unit, Sagunto, Spain; 3Hospital Universitario Torrejón de Ardoz, Intensive Care Unit, Madrid, Spain; 4Hospital Universitario de Girona, Intensive Care Unit, Girona, Spain Correspondence: A. Estella - Hospital del SAS de Jerez, Intensive Care Unit, Jerez de la Frontera, Spain Introduction: Quality of care is a key aspect of intensive care medicine. Spanish Society of Intensive and Critical Care and Coronary Units (SEMICYUC) published in 2011 quality indicators in critically ill patients.

Correspondence: A. Estella - Hospital del SAS de Jerez, Intensive Care Unit, Jerez de la Frontera, Spain Introduction: Quality of care is a key aspect of intensive care medicine. Spanish Society of Intensive and Critical Care and Coronary Units (SEMICYUC) published in 2011 quality indicators in critically ill patients. Objectives: The aim of the present study was to analyze the degree of compliance of the six bioethics quality indicators in Spanish ICU. Methods: Multicenter survey in Spanish ICU. During a time of 30 days a questionnaire was sent to local investigators of ICU. Question were related with the six bioethics quality indicators published by SEMICYUC: Indicator 96: Adequacy of care at the end of the life. Indicator 97: Information to patients and their relatives in ICU. Indicator 98: advance directives in decision makings. Indicator 99: Informed written consent forms corrected filled out . Indicator 100: Limitation life support. Indicator 101: Restraints applications. The data were analyzed using SPSS version 15 for Windows.

ator 97: Information to patients and their relatives in ICU. Indicator 98: advance directives in decision makings. Indicator 99: Informed written consent forms corrected filled out . Indicator 100: Limitation life support. Indicator 101: Restraints applications. The data were analyzed using SPSS version 15 for Windows. Results: 68 ICU participate. 50 belongs to public health services and 14 with other financial management. 70,6 % were university hospitals. Indicator 96: 44 % have end of life protocol available. Consensus decisions with participation of nurses in 79,6 %. Indicator 97: Daily information in 97 %, 82 % have specific room for information, in 92 % information is given by the doctor and in 61 % information is written in the medical record. Indicator 98: Advance directives 53 %. Indicator 99: percentage of informed written consent forms correctly filled out was variable according type of procedure. Indicator 100: 48 % have a limitation life support written form. Indicator 101: 40 % have a restraint application protocol. Conclusions: Best compliance was observed in the indicator related with the information in ICU. We must to improve variability observed about written inform consents and to promote advance directives in clinical decision making. Bioethics quality of care in ICU take on great importance. References 1. Quality indicators in critically ill patients. Madrid (Spain): Spanish Society of Intensive and Critical Care and Units Coronary (SEMICYUC);2011. 185p.

We must to improve variability observed about written inform consents and to promote advance directives in clinical decision making. Bioethics quality of care in ICU take on great importance. References 1. Quality indicators in critically ill patients. Madrid (Spain): Spanish Society of Intensive and Critical Care and Units Coronary (SEMICYUC);2011. 185p. A531 Religious beliefs may be associated with poor comprehension of end-of-life decisions S. Barbosa1, A. Varela1, I. Palma1, L. Cristina1, E. Nunes1, I. Pereira1, G. Campello2, C. Granja1,3,4 1Algarve Hospital Center, Faro Unit, Emergency and Intensive Care Department, Faro, Portugal; 2Algarve Hospital Center, Portimão Unit, Emergency and Intensive Care Department, Portimão, Portugal; 3University of Algarve, Department of Biomedical Sciences and Medicine and CINTESIS-UALG, Faro, Portugal; 4Faculty of Medicine of Porto, CINTESIS, Porto, Portugal Correspondence: S. Barbosa - Algarve Hospital Center, Faro Unit, Emergency and Intensive Care Department, Faro, Portugal Introduction: Despite technological advances that allow the support of organ and prolongs life, death in intensive care units is frequent. However, the success of care should not only be measured by survival, but should include the quality of life of each patient who survives and the comfort of the terminally ill. Objectives: To evaluate the comprehension of nurses and physicians concerning end-of-life decisions.

A531 Religious beliefs may be associated with poor comprehension of end-of-life decisions S. Barbosa1, A. Varela1, I. Palma1, L. Cristina1, E. Nunes1, I. Pereira1, G. Campello2, C. Granja1,3,4 1Algarve Hospital Center, Faro Unit, Emergency and Intensive Care Department, Faro, Portugal; 2Algarve Hospital Center, Portimão Unit, Emergency and Intensive Care Department, Portimão, Portugal; 3University of Algarve, Department of Biomedical Sciences and Medicine and CINTESIS-UALG, Faro, Portugal; 4Faculty of Medicine of Porto, CINTESIS, Porto, Portugal Correspondence: S. Barbosa - Algarve Hospital Center, Faro Unit, Emergency and Intensive Care Department, Faro, Portugal Introduction: Despite technological advances that allow the support of organ and prolongs life, death in intensive care units is frequent. However, the success of care should not only be measured by survival, but should include the quality of life of each patient who survives and the comfort of the terminally ill. Objectives: To evaluate the comprehension of nurses and physicians concerning end-of-life decisions. Methods: The study consisted of a cross-sectional survey with a self-completed questionnaire applied to nurses and physicians on a voluntary and anonymous basis. The questionnaire was constructed around the meaning of end-of-life decisions: withdrawal and withhold life-sustaining treatments and do not resuscitate decisions.

ds: The study consisted of a cross-sectional survey with a self-completed questionnaire applied to nurses and physicians on a voluntary and anonymous basis. The questionnaire was constructed around the meaning of end-of-life decisions: withdrawal and withhold life-sustaining treatments and do not resuscitate decisions. Results: A total of 174 questionnaires were delivered and 109 were returned (response rate of 63 %). However, two of the questionnaires were excluded by multiple responses to each question. Seventy-one percent were nurses, 29 % were physicians and 59 % were female; the median age was 35 years-old and the median years of practicing was 11; 64 % had religious beliefs (53 % of the physicians and 68 % of the nurses). The comprehension of end-of-life decisions was noted in 86 % of the physicians and nurses: 93 % comprehended the meaning of withdrawal and withhold life-sustaining treatments and 97 % comprehended the meaning of do not resuscitate decision. There were no significant statistical differences in the comprehension of end-of-life decisions between physicians and nurses. The religious belief was a factor significantly associated with poor comprehension of end-of-life decisions (p 0,044), namely withhold life-sustaining treatment.

of do not resuscitate decision. There were no significant statistical differences in the comprehension of end-of-life decisions between physicians and nurses. The religious belief was a factor significantly associated with poor comprehension of end-of-life decisions (p 0,044), namely withhold life-sustaining treatment. Conclusions: In this study, we found that not all physicians and nurses of an emergency and intensive care department comprehend the meaning of end-of-life decisions. Religious belief was significantly associated with poor comprehension of en-of-life decisions, namely withhold life-sustaining treatment. This study draws our attention to the need to improve training and education about end-of-life decisions in the emergency and intensive care units and to conduct further studies to evaluate the comprehension of the nurses and physicians on other departments of the hospital. References 1- Granja C, Teixeira-Pinto A, Costa-Pereira A. Attitudes towards do-not-resuscitate decisions: differences among health professionals in a Portuguese hospital. Intensive Care Med 2001; 27:555–558. 2- Carlet J, Thijs L, et al. Challenges in end-of-life care in the ICU. Statement of the 5th International consensus conference in critical care: Brussels, Belgium, April 2003. Intensive Care Med 2004 30:770–784.

1- Granja C, Teixeira-Pinto A, Costa-Pereira A. Attitudes towards do-not-resuscitate decisions: differences among health professionals in a Portuguese hospital. Intensive Care Med 2001; 27:555–558. 2- Carlet J, Thijs L, et al. Challenges in end-of-life care in the ICU. Statement of the 5th International consensus conference in critical care: Brussels, Belgium, April 2003. Intensive Care Med 2004 30:770–784. A532 Family participation in end of life care discussions in an Indian hospital R. Pande1, M. Pandey2, S. Varghese1, M. Chanu1 1BLK Superspeciality Hospital, Critical Care Medicine, New Delhi, India; 2Lady Hardinge Medical College, Anaesthesiology, New Delhi, India Correspondence: R. Pande - BLK Superspeciality Hospital, Critical Care Medicine, New Delhi, India Introduction: Withholding or withdrawal has been reported in 19-50 % of deaths in Indian ICUs, withdrawal being limited to only 8 % of cases1. Various reasons have been identified as barriers to end of life care in India2. A recent physician interview based Asian multinational study has highlighted significant difference in such practices3. Objectives: This study was undertaken to look at the attitude of Indian families towards end of life care (EOL) issues.

A532 Family participation in end of life care discussions in an Indian hospital R. Pande1, M. Pandey2, S. Varghese1, M. Chanu1 1BLK Superspeciality Hospital, Critical Care Medicine, New Delhi, India; 2Lady Hardinge Medical College, Anaesthesiology, New Delhi, India Correspondence: R. Pande - BLK Superspeciality Hospital, Critical Care Medicine, New Delhi, India Introduction: Withholding or withdrawal has been reported in 19-50 % of deaths in Indian ICUs, withdrawal being limited to only 8 % of cases1. Various reasons have been identified as barriers to end of life care in India2. A recent physician interview based Asian multinational study has highlighted significant difference in such practices3. Objectives: This study was undertaken to look at the attitude of Indian families towards end of life care (EOL) issues. Methods: A standard protocol for family discussion was developed. Relevant end of life care issues was discussed in presence of admitting physician, ICU nurse, medical administrator and members of ICU team. The filled up EOL family discussions forms related to 44 critically ill adult patients admitted to ICU between March 2010 & March 2011 were retrospectively analyzed. The discussions held during non-regular hours were excluded.

in presence of admitting physician, ICU nurse, medical administrator and members of ICU team. The filled up EOL family discussions forms related to 44 critically ill adult patients admitted to ICU between March 2010 & March 2011 were retrospectively analyzed. The discussions held during non-regular hours were excluded. Results: The data included 26 male and 18 female patients, admitted with a medical diagnosis (n = 18), surgical (n = 4), trauma (n = 6) or cancer (n = 16). The age of patients varied from 20–40 years (22.7 %, n = 10), 40–60 yrs. (18 %, n = 8), 60–80 yrs. (36.36 %, n = 16) and 10 patients (22.7 %) were >80 years of age. In majority of the discussions more than 3 members of the family participated (n = 29), 2 members (n = 12) and 1member in 3 patients. All families participated in end of life care discussions. They understood and agreed on a situation of medical futility and irreversibility in their loved ones and 79.54 % (n = 35) families requested for no further escalation in treatment. 43 % (n = 19) families agreed for No CPR orders, whereas 20.45 % (n = 9) agreed for Do not intubate and do not ventilate orders. Only 4.54 % (n = 2) families agreed for withdrawal of ventilator, 9 % agrees for withdrawal of vasopressors and only 1 family (2.27 %) agreed for organ donation. Majority of the families (90.9 %, n = 40) tried to understand the measures given to alleviate pain and the feeling of dyspnoea related to ventilator withholding or withdrawal.

2) families agreed for withdrawal of ventilator, 9 % agrees for withdrawal of vasopressors and only 1 family (2.27 %) agreed for organ donation. Majority of the families (90.9 %, n = 40) tried to understand the measures given to alleviate pain and the feeling of dyspnoea related to ventilator withholding or withdrawal. Conclusions: Our study shows willingness of families to participate in end of life care discussions and accepting medical futility and irreversibility of such situations. Withholding life care was more acceptable than withdrawing. References 1. Kapadia F, Singh M, Divatia J et al. Limitation and withdrawal of intensive therapy at end of life: practices in intensive care units in Mumbai, India. Crit Care Med. 2005; 33(6):1272–5. 2. Mani RK. End-of-life care in India. Intensive Care Med. 2006;32(7):1066–8. 3. Phua J, Joynt GM, Nishimura M et al. Withholding and withdrawal of life-sustaining treatments in low-middle-income versus high-income Asian countries and regions. Intensive Care Med 2016. DOI 10.1007/s00134-016-4347-y

1. Kapadia F, Singh M, Divatia J et al. Limitation and withdrawal of intensive therapy at end of life: practices in intensive care units in Mumbai, India. Crit Care Med. 2005; 33(6):1272–5. 2. Mani RK. End-of-life care in India. Intensive Care Med. 2006;32(7):1066–8. 3. Phua J, Joynt GM, Nishimura M et al. Withholding and withdrawal of life-sustaining treatments in low-middle-income versus high-income Asian countries and regions. Intensive Care Med 2016. DOI 10.1007/s00134-016-4347-y A533 Development and evaluation of an e-learning to enhance advanced care planning discussion by residents in a university medical center, a pilot study M.J. Van Dam1,2, E.W.M.T. Ter Braak2,3 1University Medical Center Utrecht, IC Center, Utrecht, Netherlands; 2University Medical Center Utrecht, Educational Center, Utrecht, Netherlands; 3University Medical Center Utrecht, Internal Medicine, Utrecht, Netherlands Correspondence: M.J. Van Dam - University Medical Center Utrecht, Educational Center, Utrecht, Netherlands Introduction: Advance care planning (ACP) is the process in which the patient makes decisions in consultation with health care givers about future health care once they become incapable of deliberate medical treatment decisions. ACP is increasingly recognized as area of concern respecting quality of care in the ICU and other wards in the hospital. However, research shows that ACP discussions (ACPd) are still rare in hospital settings: clinicians underestimate the number of patients willing to discuss ACP and the feeling being unskilled, inadequately trained or inexperienced may hamper a meaningful patient-doctor discussion.

e ICU and other wards in the hospital. However, research shows that ACP discussions (ACPd) are still rare in hospital settings: clinicians underestimate the number of patients willing to discuss ACP and the feeling being unskilled, inadequately trained or inexperienced may hamper a meaningful patient-doctor discussion. Objectives: To develop and evaluate an e-module preparing residents to perform effective and timely ACPd. Methods: Development of an e-module aiming to equip residents with knowledge, skills and appropriate attitudes for ACPd, preferably in a simulated situation before real life practice in the workplace. The design of this e-module is inspired by the 4C/ID-model. First, experts explain the ethical, legal and social issues concerning ACP that constitute the learning goals of the e-module. Subsequently, five cases are presented in videos illustrating appropriate performance. Each case is provided with comments from experts and residents and their advice based on personal experience. Cases are presented in increasing complexity sequence. Roles of patients are performed by trained simulation patients. Residents, senior staff and nurses demonstrate lifelike ACPd designed to illustrate preset educational goals.

ded with comments from experts and residents and their advice based on personal experience. Cases are presented in increasing complexity sequence. Roles of patients are performed by trained simulation patients. Residents, senior staff and nurses demonstrate lifelike ACPd designed to illustrate preset educational goals. Documenting ACP agreements in the electronic health record is shown using screen shots. Finally, background information by experts and useful links are offered. Learners work through the e-module at their own pace with the option to interrupt and looking back. The e-module was piloted in a group of residents purposefully sampled from a variety of specialties with a web-based 5 point Likert scale survey with room for narrative explanation to evaluate satisfaction of participants. Results: In this pilot 18 residents participated, mean age 30.4 (+/− SD 2.6) year, 33 % male, in training for neurology (n = 5), internal medicine (n = 5), surgery (n = 3), anesthesiology (n = 2), cardiology (n = 1) and undisclosed (n = 2), ranging from 1st to 6th year of training (median 3rd). 83 % of participants (strongly) agreed with the statement that the e-module is well-arranged. 89 % (n = 16) review the e-module as giving sufficient insight in the necessity of ACPd. Ten residents (55 %) appraise the e-module offering guidance for performing ACPd themselves in clinical practice. Overall satisfaction was 7/10 points (median, range 6–8).

reed with the statement that the e-module is well-arranged. 89 % (n = 16) review the e-module as giving sufficient insight in the necessity of ACPd. Ten residents (55 %) appraise the e-module offering guidance for performing ACPd themselves in clinical practice. Overall satisfaction was 7/10 points (median, range 6–8). Conclusions: This pilot suggests s that an e-module showing appropriate behaviors based on clarified skills and attitudes by role models in lifelike situations may help to engage residents in learning activities and actual practice to master ACPd, thus contributing to quality of care. A534 Limitation of life sustaining therapy in patients died early in ICU A. Estella1, M. Gracia1, R. Viciana2, M. Recuerda1, L. Perez Fontaiña1 1Hospital del SAS de Jerez, Intensive Care Unit, Jerez de la Frontera, Spain; 2Hospital del SAS de Jerez, Jerez de la Frontera, Spain Correspondence: A. Estella - Hospital del SAS de Jerez, Intensive Care Unit, Jerez de la Frontera, Spain Introduction: Early mortality in ICU may be related to the reason for admission, comorbidity, severity and initial therapy. Objectives: To describe severity and comorbidity relationship with early mortality in patients admitted in ICU.To analyze differences in patients early died in ICU according limitation of life-sustaining therapy (LST) decisions.