Browse the corpus

CHILDREN, TEENAGER AND FAMILIES IN THE ICU A793 Malnutrition and clinical outcomes in critically ill children T. Velasquez1, G. Mackey2,3, J. Lusk2,3, U.G. Kyle2,3, T. Fontenot2, P. Marshall2, L.S. Shekerdemian2,3, J.A. Coss-Bu2,3 1Texas Children's Hospital, Clinical Nutriiton, Houston, United States; 2Texas Children's Hospital, Intensive Care, Houston, United States; 3Baylor College of Medicine, Pediatrics, Houston, United States Correspondence: J.A. Coss-Bu – Texas Children's Hospital, Intensive Care, Houston, United States Introduction: Critically ill children in the pediatric intensive care unit (PICU) are at high risk for developing nutritional deficiencies and undernutrition is known to be a risk factor for morbidity and mortality. Malnutrition represents a continuous spectrum ranging from marginal nutrient status to severe metabolic and functional alterations and this in turn, affects clinical outcome. Objectives: The aim of the study was to assess nutritional status of critically ill children admitted to the PICU and its association to clinical outcomes.

CHILDREN, TEENAGER AND FAMILIES IN THE ICU A793 Malnutrition and clinical outcomes in critically ill children T. Velasquez1, G. Mackey2,3, J. Lusk2,3, U.G. Kyle2,3, T. Fontenot2, P. Marshall2, L.S. Shekerdemian2,3, J.A. Coss-Bu2,3 1Texas Children's Hospital, Clinical Nutriiton, Houston, United States; 2Texas Children's Hospital, Intensive Care, Houston, United States; 3Baylor College of Medicine, Pediatrics, Houston, United States Correspondence: J.A. Coss-Bu – Texas Children's Hospital, Intensive Care, Houston, United States Introduction: Critically ill children in the pediatric intensive care unit (PICU) are at high risk for developing nutritional deficiencies and undernutrition is known to be a risk factor for morbidity and mortality. Malnutrition represents a continuous spectrum ranging from marginal nutrient status to severe metabolic and functional alterations and this in turn, affects clinical outcome. Objectives: The aim of the study was to assess nutritional status of critically ill children admitted to the PICU and its association to clinical outcomes. Methods: Critically ill children age 6 months to 18 years were prospectively enrolled on PICU admission. Nutritional status was assessed by weight for age (WFA: underweight), weight for height (WFH: wasting), height for age (HFA: stunting) z-scores and mid upper arm circumference (MUAC: wasting) according to the WHO. (1,2) Malnutrition was defined as mild, moderate, and severe if z-scores were > −1, > − 2, and > −3, respectively. Hospital and PICU length of stay (LOS), duration of mechanical ventilation (MV), and risk of mortality (ROM) by the Pediatric Index of Mortality 2 (PIM2) were obtained. Sensitivity and specificity of the MUAC to identify children with wasting (WFH) were calculated.

if z-scores were > −1, > − 2, and > −3, respectively. Hospital and PICU length of stay (LOS), duration of mechanical ventilation (MV), and risk of mortality (ROM) by the Pediatric Index of Mortality 2 (PIM2) were obtained. Sensitivity and specificity of the MUAC to identify children with wasting (WFH) were calculated. Results: Two hundred and fifty children (136 males), aged 81 months (23–167; median (25-75th IQR)), were prospectively included in the study. The hospital LOS was 8 (4–16) days; PICU LOS: 2 (1–4) days; duration of MV, 0 (0–1.5) days; PIM2 ROM 2.61 ± 0.25 %. WFA, WFH, and HFA z-scores of −0.48 ± 0.14; 0.19 ± 0.13; and −0.95 ± 0.13 respectively; MUAC, 16.3 ± 0.18 cm (6 to 59 months, n = 108); 24.2 ± 0.46 cm (5 to 18 years, n = 142). The prevalence of underweight, wasting and stunting was 26.4 %, 19.6 %, and 44.4 % respectively. The sensitivity and specificity for MUAC vs. WFH to identify wasting was: 34.5 % (20.3-50.6; 95 % CI) and 95.5 % (91.8-97.9), respectively. Values are mean ± SE. Conclusions: Malnutrition in critically ill children is prevalent with half of the patients being stunted, reflecting the chronic nature of the disease process and its effects on the nutritional status. The performance of MUAC as a screening tool in this population was poor, but identified correctly almost all children with wasting. There was an association between nutritional status and length of stay and risk of mortality. References 1 WHO: Technical Report Series, No. 854, 1995 2 Bulletin of the WHO, 1997, 75:11–18 Grant acknowledgement Internal FUNDING

Conclusions: Malnutrition in critically ill children is prevalent with half of the patients being stunted, reflecting the chronic nature of the disease process and its effects on the nutritional status. The performance of MUAC as a screening tool in this population was poor, but identified correctly almost all children with wasting. There was an association between nutritional status and length of stay and risk of mortality. References 1 WHO: Technical Report Series, No. 854, 1995 2 Bulletin of the WHO, 1997, 75:11–18 Grant acknowledgement Internal FUNDING Texas Children´s HospitalTable 1 (abstract A793). Malnutriiton and Outcomes Underweight Wasting Stunting Odds Ratio 95 % C.I. p value Odds Ratio 95 % C.I. p value Odds Ratio 95 % C.I. p value Hospital LOS 2.40 1.16–4.99 0.019 2.26 1.17–4.36 0.015 1.05 0.64–1.73 0.854 ICU LOS 2.17 1.05–4.49 0.037 2.28 1.05–4.95 0.037 2.33 1.14–4.79 0.021 PIM2 ROM 1.16 1.05–1.28 0.004 1.08 1.01–1.17 0.034 1.12 1.01–1.23 0.024 LOS length of stay, PIM2 pediatric index of mortality, ROM risk of mortality

Underweight Wasting Stunting Odds Ratio 95 % C.I. p value Odds Ratio 95 % C.I. p value Odds Ratio 95 % C.I. p value Hospital LOS 2.40 1.16–4.99 0.019 2.26 1.17–4.36 0.015 1.05 0.64–1.73 0.854 ICU LOS 2.17 1.05–4.49 0.037 2.28 1.05–4.95 0.037 2.33 1.14–4.79 0.021 PIM2 ROM 1.16 1.05–1.28 0.004 1.08 1.01–1.17 0.034 1.12 1.01–1.23 0.024 LOS length of stay, PIM2 pediatric index of mortality, ROM risk of mortality A794 Retrospective analysis for predicting optimal tracheal tube size in pediatric patients A. Nishigaki, T. Yatabe, T. Tamura, K. Yamashita, M. Yokoyama Kochi Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan Correspondence: A. Nishigaki – Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan Introduction: There are several methods to estimate the optimal tracheal tube size in pediatric patients such as the Cole's formula (inner diameter (ID) = 4 + Age/4) [1]. However, these evaluation methods are made based on age in years (not months) and ID. Moreover, outer diameter (OD) may vary according to the type of the tracheal tube. Objectives: We hypothesized that prediction of OD for determining the optimal tracheal tube size in pediatric patients based on age in months is better than Cole's formula. Therefore, we conducted a retrospective analysis to investigate our hypothesis.

A794 Retrospective analysis for predicting optimal tracheal tube size in pediatric patients A. Nishigaki, T. Yatabe, T. Tamura, K. Yamashita, M. Yokoyama Kochi Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan Correspondence: A. Nishigaki – Medical School, Department of Anesthesiology and Intensive Care Medicine, Nankoku, Japan Introduction: There are several methods to estimate the optimal tracheal tube size in pediatric patients such as the Cole's formula (inner diameter (ID) = 4 + Age/4) [1]. However, these evaluation methods are made based on age in years (not months) and ID. Moreover, outer diameter (OD) may vary according to the type of the tracheal tube. Objectives: We hypothesized that prediction of OD for determining the optimal tracheal tube size in pediatric patients based on age in months is better than Cole's formula. Therefore, we conducted a retrospective analysis to investigate our hypothesis. Methods: The ethics committee of our hospital approved this retrospective study. We included consecutive patients aged < 6 years who underwent tracheal intubation under general anesthesia in our hospital from August 2013 to October 2015. We collected the following data from the anesthesia records: age in months, height, weight, type of a tracheal tube, and ID and OD of tracheal tube. Patients who were intubated using a cuffed tracheal tube or had incomplete data were excluded. We developed a regression formula for calculating ID and OD based on age in months and calculated the coefficient of determination R2 by using a regression analyses. A difference of 0.4 mm in the actual and predicted tube size was considered clinically permissible. Then, we compared the rate of a clinical permissible estimation of the Cole's formula and our new formulas used by multiple comparison analysis and a p value less than 0.05 was considered statistically significant.

yses. A difference of 0.4 mm in the actual and predicted tube size was considered clinically permissible. Then, we compared the rate of a clinical permissible estimation of the Cole's formula and our new formulas used by multiple comparison analysis and a p value less than 0.05 was considered statistically significant. Results: A total of 207 pediatric patients received general anesthesia during the study period. Of these, 67 patients were excluded because they did not meet the inclusion criteria. Finally, we included 140 patients for this analysis. The regression formula for predicting ID by based on age in months was ID = 0.019 × age in months + 3.48, and the coefficient of determination R2 was 0.54. The regression formula for predicting OD based on age in months was OD = 0.024 × age in months + 5.21, and coefficient of determination R2 was 0.558. The rate of a clinical permissible estimation of our ID and OD formulas were significantly higher than that of the Cole's formula (61 %, 69 % and 43 %, respectively; p < 0.01). Conclusions: Our results showed that the prediction of ID based on age in months is more useful than that using Cole's formula. In addition, estimation of OD based on age in months might be more rational because OD varies according to the type of the tracheal tube used. These results should be confirmed in a future prospective study. References [1] Cole F. Pediatric formulas for the anesthesiologist. AMA J Dis Child. 1957;94:672–3.

Conclusions: Our results showed that the prediction of ID based on age in months is more useful than that using Cole's formula. In addition, estimation of OD based on age in months might be more rational because OD varies according to the type of the tracheal tube used. These results should be confirmed in a future prospective study. References [1] Cole F. Pediatric formulas for the anesthesiologist. AMA J Dis Child. 1957;94:672–3. A795 Teenagers perception towards cardiopulmonary resuscitation J.-C. Ruiz-Rodriguez1, B. Encina1, R. Belmonte1, I. Troncoso1, P. Tormos2, M. Riveiro3, J. Baena3, A. Sanchez1, J. Bañeras4, J. Cordón1, N. Duran5, A. Ruiz1, J. Caballero1, X. Nuvials1, J. Riera1, J. Serra1 1Vall d' Hebron University Hospital, Critical Care Department, Barcelona, Spain; 2Vall d' Hebron University Hospital, Anesthesia & Reanimation Department, Barcelona, Spain; 3Vall d' Hebron University Hospital, Neurocritical Care Department, Barcelona, Spain; 4Vall d' Hebron University Hospital, Coronary Care Unit, Barcelona, Spain; 5Sagrat Cor University Hospital, Critical Care Department, Barcelona, Spain Correspondence: J.-C. Ruiz-Rodriguez – Vall d' Hebron University Hospital, Critical Care Department, Barcelona, Spain Introduction: Survival among out-of-hospital cardiac arrest (CA) relies primarily on bystanders and their knowledge of basic life support (BLS) manouvers [1]. Many medical societies and organizations recommend teaching BLS at schools as part of the educative program [2]; being this a reality in North European countries, but not yet an education standard issue in others including Spain. Moreover, less is written about the perception of CA and cardiopulmonary resuscitation (CPR) among the general population, and even less in school age.

LS at schools as part of the educative program [2]; being this a reality in North European countries, but not yet an education standard issue in others including Spain. Moreover, less is written about the perception of CA and cardiopulmonary resuscitation (CPR) among the general population, and even less in school age. Objectives: Describe the perception and knowledge about CA and CPR among a teenager school population in Barcelona, Spain. Methods: Prospective, descriptive study carried out between 2007–2009 and 2012–2015 among teenagers school population, based on surveys before and after BLS - CPR classes. During this period , 17th classes were held, in 3 different schools in Barcelona. Before attending the class , each pupil was asked to answer a survey with questions related to previous knowledge of sudden death, CA, and CPR, and their attitude towards them. The class consisted on a three - hour theorical and practical instruction based on the European Resuscitation Council guidelines, adapted for laypersons. Practices were held with an instructor (ratio instructor:pupil 1:6–8), with the Little Anne mannequins (Laerdal®). After the class, a new survey (post intervention) was distributed, with questions related to the new concepts and skills learnt, the attitude toward CA and CPR.

Methods: Prospective, descriptive study carried out between 2007–2009 and 2012–2015 among teenagers school population, based on surveys before and after BLS - CPR classes. During this period , 17th classes were held, in 3 different schools in Barcelona. Before attending the class , each pupil was asked to answer a survey with questions related to previous knowledge of sudden death, CA, and CPR, and their attitude towards them. The class consisted on a three - hour theorical and practical instruction based on the European Resuscitation Council guidelines, adapted for laypersons. Practices were held with an instructor (ratio instructor:pupil 1:6–8), with the Little Anne mannequins (Laerdal®). After the class, a new survey (post intervention) was distributed, with questions related to the new concepts and skills learnt, the attitude toward CA and CPR. Results: We have instructed 561 pupils (14.02 (±0.79) years, 48.2 % female). The 87.8 % had heard about sudden death and CA before the class. Regarding starting CPR: 40 % said they were not capable of doing it, and 51.2 % suggested they would be able to do CPR but in a wrong manner. In a CA scene 58.9 % would contact the emergency service and start CPR, 27.4 % would call and wait, and 11.4 % would only do CPR. After attending the classes 98.6 % declared had understood the theorical concepts and practical skilles taught; 95.4 % would changed positively their attitude towards CPR; and 97.4 % would be prone to start maneuvers.

contact the emergency service and start CPR, 27.4 % would call and wait, and 11.4 % would only do CPR. After attending the classes 98.6 % declared had understood the theorical concepts and practical skilles taught; 95.4 % would changed positively their attitude towards CPR; and 97.4 % would be prone to start maneuvers. Conclusions: CPR and CA remain a well known issue among teenager population in Barcelona, as long as being an interesting topic. Nevertheless they do not feel capable of starting maneuvers. The concepts taught during the class were easy to learn , and after the intervention the majority were prone to start CPR. This population is adequate to teach CPR. References 1. Hansen CM et al. The role of bystanders, first responders, and emergency medical service providers in timely defibrillation and related outcomes after out-of-hospital cardiac arrest: Results from a statewide registry. Resuscitation 2015;96:303–9. 2. Böttiger BW et al. Kids save lives--Training school children in cardiopulmonary resuscitation worldwide is now endorsed by the World Health Organization (WHO). .Resuscitation. 2015;94:A5-7.

1. Hansen CM et al. The role of bystanders, first responders, and emergency medical service providers in timely defibrillation and related outcomes after out-of-hospital cardiac arrest: Results from a statewide registry. Resuscitation 2015;96:303–9. 2. Böttiger BW et al. Kids save lives--Training school children in cardiopulmonary resuscitation worldwide is now endorsed by the World Health Organization (WHO). .Resuscitation. 2015;94:A5-7. A796 Family = children included , guidance of visiting children at an adult intensive care A.M.F. Rutten, S.N.M. van Ieperen, E.P.H.M. Der Kinderen, T. Van Logten St Elisabeth Twee Steden Hospital, ICU, Tilburg, Netherlands Correspondence: S.N.M. van Ieperen – St Elisabeth Twee Steden Hospital, ICU, Tilburg, Netherlands Introduction: To meet the need of patients family members and staff we started to guide visiting children at our adult ICU in the St Elisabeth hospital (EZ) in Tilburg 3 years ago. To do so we developed a guidance leaflet for parents with practical instructions and information. Additionally, practical advice is given, such as what to say to the child and what to expect when visiting. The leaflet is subdivided in developmental stages. Furthermore we developed a book “mees op bezoek”, in which a child visits his father at the ICU. Pictures show what children can expect, which helps prepare the child for visiting at home. An instruction box is present at the ICU with ICU materials such as an iv catheter, a pulse oximetry or a tracheal tube. These materials give children a tactile experience of the ICU. The box is divided in two parts; the second part contains guidance materials for when a patient may die. Pedagogical staff are available to support parents, children and staff. If there are more profound problems a referral to our children's psychologist is possible. We made some improvements to our waiting area to make it more appealing to children. We instructed and educated our nurses and doctors on how to use these materials and how to guide children.

support parents, children and staff. If there are more profound problems a referral to our children's psychologist is possible. We made some improvements to our waiting area to make it more appealing to children. We instructed and educated our nurses and doctors on how to use these materials and how to guide children. We recently merged with the Twee steden Hospital in Tilburg (TSZ), in this hospital there was no program to guide children. With the merger we also wanted to introduce our “Child as a visitor program” at the ICU on location TSZ. We wanted to know if there were differences of opinion between the nursing staff on guidance of children. Methods: We held a survey among our nursing staff. In TSZ we handed out surveys on paper during an obligatory education. In EZ the same survey was sent by email.

We recently merged with the Twee steden Hospital in Tilburg (TSZ), in this hospital there was no program to guide children. With the merger we also wanted to introduce our “Child as a visitor program” at the ICU on location TSZ. We wanted to know if there were differences of opinion between the nursing staff on guidance of children. Methods: We held a survey among our nursing staff. In TSZ we handed out surveys on paper during an obligatory education. In EZ the same survey was sent by email. Results: Response rate in EZ was 61 % (n = 127). Respons rate nursing staff in TSZ was 100 % (n = 33). Nearly all nurses share the view that children should be allowed to visit an ICU: EZ 97 % and TSZ 94 %. The appropriate age for children was deemed higher in TSZ with an average of 2,5 years, in EZ this was 0,74 years. 65 % of the nurse in EZ responded that children of all ages were welcome versus 33 % in TSZ. In EZ 44 % of the nurses didn't need any more support to guide children. In TSZ this was 3 %. 88 % of the nurses in TSZ wanted more education on the subject. In EZ there was still a great need: 56 % wanted this. More help from pedagogical staff was needed in 53 % of the nurses but in EZ this was 14 %. EZ 9 % needed more informational materials in TSZ this was 40 %.(see graph 1).

e children. In TSZ this was 3 %. 88 % of the nurses in TSZ wanted more education on the subject. In EZ there was still a great need: 56 % wanted this. More help from pedagogical staff was needed in 53 % of the nurses but in EZ this was 14 %. EZ 9 % needed more informational materials in TSZ this was 40 %.(see graph 1). Conclusion: When you allow children to visit your ICU, nurses want to be educated on the subject, they need practical aids and help from pedagogical staff. The need for more pedagogical help and practical aids are less with the nurses who have more experience. A need for education on the subject will remain. We are introducing the “Child as a visitor program” at the ICU on location TSZ and will expand education in EZ.Fig. 1 (abstract A796). Do you need more support to guide children in our ICU?

ogical help and practical aids are less with the nurses who have more experience. A need for education on the subject will remain. We are introducing the “Child as a visitor program” at the ICU on location TSZ and will expand education in EZ.Fig. 1 (abstract A796). Do you need more support to guide children in our ICU? A797 High-frequency chest wall oscillation therapy in pediatric cardiac intensive care unit L. Kovacikova1, P. Skrak2, M. Zahorec3 1National Institute of Cardiovascular Diseases, PCICU, Bratislava, Slovakia; 2National Institute of Cardiovascular Diseases, Bratislava, Slovakia; 3National Institute of Cardiovascular Diseases, Pediatric Cardiac Intensive Care Unit, Bratislava, Slovakia Correspondence: L. Kovacikova – National Institute of Cardiovascular Diseases, PCICU, Bratislava, Slovakia Introduction: In critically ill children with cardiac diseases lung complications are frequently highlighting atelectasis and pneumonia. Physiotherapy has an important role in the treatment of these complications. High Frequency Chest Wall Oscillation (HFCWO) has been shown to be effective in helping to clear secretions from the lungs of patients with cystic fibrosis, primary ciliary dyskinesia, bronchiectasis and others. However, the role of HFCWO in children with cardiac diseases has not been established. Objectives: This prospective observational study was conducted to determine if HFCWO treatment, as provided by The Vest™ Airway Clearance System (Hill-Rom, Saint Paul, MN), was safe and tolerated by these patients.

A797 High-frequency chest wall oscillation therapy in pediatric cardiac intensive care unit L. Kovacikova1, P. Skrak2, M. Zahorec3 1National Institute of Cardiovascular Diseases, PCICU, Bratislava, Slovakia; 2National Institute of Cardiovascular Diseases, Bratislava, Slovakia; 3National Institute of Cardiovascular Diseases, Pediatric Cardiac Intensive Care Unit, Bratislava, Slovakia Correspondence: L. Kovacikova – National Institute of Cardiovascular Diseases, PCICU, Bratislava, Slovakia Introduction: In critically ill children with cardiac diseases lung complications are frequently highlighting atelectasis and pneumonia. Physiotherapy has an important role in the treatment of these complications. High Frequency Chest Wall Oscillation (HFCWO) has been shown to be effective in helping to clear secretions from the lungs of patients with cystic fibrosis, primary ciliary dyskinesia, bronchiectasis and others. However, the role of HFCWO in children with cardiac diseases has not been established. Objectives: This prospective observational study was conducted to determine if HFCWO treatment, as provided by The Vest™ Airway Clearance System (Hill-Rom, Saint Paul, MN), was safe and tolerated by these patients. Methods: Eighty-five treatment courses were evaluated in 25 pediatric cardiac patients during the stay at intensive care unit. Median age of the patients was 2 months (range; 12 days - 7 years) and weight 4.2 kg (range; 2.4 - 54 kg). Twenty-three (92 %) patients were following cardiac surgery. Patients were receiving invasive or non-invasive mechanical ventilation (31 and 27 courses, respectively), or high-flow nasal cannula oxygen delivered by Vapotherm (27 courses). The main indication for HFCWO was atelectasis detected on a chest x-ray (84 % patients). Other indications included lack of cough reflex, arterial oxygen desaturations, and dyspnea. HFCWO was applied at 7 Hz (range; 5–15 Hz) and a pressure of 2 (range; 1–6 arbitrary units) for 10 minutes. Routine hemodynamic and pulse oximetry data, and qualitative data on patient tolerance were collected before, during, and after HFCWO.

ns included lack of cough reflex, arterial oxygen desaturations, and dyspnea. HFCWO was applied at 7 Hz (range; 5–15 Hz) and a pressure of 2 (range; 1–6 arbitrary units) for 10 minutes. Routine hemodynamic and pulse oximetry data, and qualitative data on patient tolerance were collected before, during, and after HFCWO. Results: Heart rate, systolic and diastolic blood pressure, and respiratory rate increased significantly during HFCWO courses and decreased significantly following therapy. Oxygen saturations significantly decreased during HFCWO and significantly increased after discontinuation of treatment. The differences between pre- and post- HFCWO data were not significant. Patients remained calm during 80 HFCWO courses, and became agitated during 5 courses. No chest tubes, intracardiac lines, or catheters were dislodged in association with HFCWO. No premature discontinuation of therapy was required because of intolerance. Conclusions: The study suggests that HFCWO therapy is safe and well tolerated in children with cardiac diseases in intensive care unit.

Results: Heart rate, systolic and diastolic blood pressure, and respiratory rate increased significantly during HFCWO courses and decreased significantly following therapy. Oxygen saturations significantly decreased during HFCWO and significantly increased after discontinuation of treatment. The differences between pre- and post- HFCWO data were not significant. Patients remained calm during 80 HFCWO courses, and became agitated during 5 courses. No chest tubes, intracardiac lines, or catheters were dislodged in association with HFCWO. No premature discontinuation of therapy was required because of intolerance. Conclusions: The study suggests that HFCWO therapy is safe and well tolerated in children with cardiac diseases in intensive care unit. A798 Protein feeding in pediatric acute kidney injury does not delay renal recovery U.G. Kyle1,2, A. Akcan-Arikan1,2,3, J.C. Silva1,2, G. Mackey1,2, J. Lusk1,2, M. Goldsworthy1,2, L.S. Shekerdemian1,2, J.A. Coss-Bu1,2 1Texas Children's Hospital, Intensive Care, Houston, United States; 2Baylor College of Medicine, Pediatrics, Houston, United States; 3Texas Children's Hospital, Renal Service, Houston, United States Correspondence: J.A. Coss-Bu – Texas Children's Hospital, Intensive Care, Houston, United States Introduction: Critically ill children are underfed early in their Pediatric Intensive Care Unit (PICU) stay and this may contribute to worse outcomes. Acute Kidney Injury (AKI) occurs in 10 % of all PICU admissions and the risk of acute and chronic malnutrition is high in these patients with AKI, and the presence of malnutrition in the context of AKI has been associated with more severe clinical deterioration and organ dysfunction. Critically ill children with AKI are at high risk of underfeeding.

s in 10 % of all PICU admissions and the risk of acute and chronic malnutrition is high in these patients with AKI, and the presence of malnutrition in the context of AKI has been associated with more severe clinical deterioration and organ dysfunction. Critically ill children with AKI are at high risk of underfeeding. Objectives: To evaluate the effects of protein feeding on the resolution of AKI. Methods: This is a retrospective study of critically ill children admitted from 10/2012-12/2013 to the PICU. Patients with a diagnosis of end stage renal disease requiring renal replacement therapy or had received a kidney transplant were excluded. Nutritional status assessed by weight and height WHO z-scores after admission and caloric and protein intakes calculated from I.V. fluids and parenteral and enteral nutrition for the first 8 days of admission. Energy and protein needs estimated by Schofield and A.S.P.E.N., respectively. AKI was defined by pRIFLE (creatinine only) and persistent AKI was defined as patients who did not resolve their AKI during the first eight days of PICU stay.

I.V. fluids and parenteral and enteral nutrition for the first 8 days of admission. Energy and protein needs estimated by Schofield and A.S.P.E.N., respectively. AKI was defined by pRIFLE (creatinine only) and persistent AKI was defined as patients who did not resolve their AKI during the first eight days of PICU stay. Results: A total of 511 patients were included and 156 patients (30.5 %) had AKI. Patients with AKI vs. non-AKI had: age; 1.2 yrs (0.2-6) median (IQR) vs. 1.5 yrs. (0.4-6) (p = 0.10); height: 76 cm (58–110) vs. 81 cm (64–117) (p = 0.02), weight: 9.3 kg (5–21) vs. 11 kg (7–21) (p = 0.04) and mortality: 8.3 % vs. 4.2 % respectively. Forty-four (8.61 %) and 112 (21.9 %) patients had persistent and resolved AKI, respectively, and persistent AKI patients were more likely to have moderate/severe chronic malnutrition vs. non-AKI patients, odds ratio (95 % CI) 2.4 (1.2-4.6) (p = 0.014). Patients with no AKI, resolved AKI, and persistent AKI received in the first 8 days of PICU stay 73 %, 80 % and 80 % of recommended energy needs, and 39 %, 42 %, and 51 % of protein needs, respectively. Compared to 12 % of no-AKI patients, 17 % and 27 % of patients with resolved and persistent AKI, met ≥80 % of protein needs respectively, (p = 0.01) Although patients with persistent AKI received higher protein intake of ≥80 % of goal, was not independently associated with persistent AKI after adjustment for PRISM score (p = 0.13).

2 % of no-AKI patients, 17 % and 27 % of patients with resolved and persistent AKI, met ≥80 % of protein needs respectively, (p = 0.01) Although patients with persistent AKI received higher protein intake of ≥80 % of goal, was not independently associated with persistent AKI after adjustment for PRISM score (p = 0.13). Conclusions: Protein prescription is improved in children with AKI in our PICU, largely due to ongoing educational efforts. Higher protein intake was not associated with a delay in renal recovery in patients with AKI after adjustment for severity of illness. References: 1. Akcan-Arikan A, Kidney Int. 2007. 2. Schofield WN. Clin Nutr 1985. 3. Mehta NM. JPEN, 2009. Grant acknowledgement Internal Funding, Texas Children's Hospital

Conclusions: Protein prescription is improved in children with AKI in our PICU, largely due to ongoing educational efforts. Higher protein intake was not associated with a delay in renal recovery in patients with AKI after adjustment for severity of illness. References: 1. Akcan-Arikan A, Kidney Int. 2007. 2. Schofield WN. Clin Nutr 1985. 3. Mehta NM. JPEN, 2009. Grant acknowledgement Internal Funding, Texas Children's Hospital A799 Current intensive care management for adolescents in the United Kingdom: a retrospective cohort study D. Wood1, D. Harrison2, R. Parslow3, P. Davis1, J. Pappachan4, S. Goodwin1, P. Ramnarayan5 1Bristol Royal Hospital for Children, Paediatric Intensive Care Unit, Bristol, United Kingdom; 2Intensive Care National Audit & Research Centre (ICNARC), London, United Kingdom; 3University of Leeds, Leeds, United Kingdom; 4Southampton University Hospitals NHS Trust, Southampton, United Kingdom; 5Great Ormond Street Hospital, London, United Kingdom Correspondence: D. Wood – Bristol Royal Hospital for Children, Paediatric Intensive Care Unit, Bristol, United Kingdom Introduction: The transition between childhood and adulthood is a time of rapid physical, psychological and behavioural change. Adolescents (aged 12–19 years) requiring intensive care differ from both the typical paediatric intensive care (PICU) population, mainly infants and pre-school children, and from the typical adult intensive care unit (AICU) population of much older adults. Critically ill adolescents are distinct from the majority of patients treated in either adult or paediatric intensive care units (ICU). Little data exist to describe how best to meet their needs or those of their families.

children, and from the typical adult intensive care unit (AICU) population of much older adults. Critically ill adolescents are distinct from the majority of patients treated in either adult or paediatric intensive care units (ICU). Little data exist to describe how best to meet their needs or those of their families. Objectives: We describe the case mix, resource use, and outcomes for adolescents admitted to AICU and PICU in the UK. Methods: Analysis of national, prospectively collected data for adolescents (aged 12–19 years) admitted to ICUs in the UK between 2007 and 2014. Results: 37320 adolescents were admitted during the study period. Excluding admissions following elective surgery, there were a total of 27442 admissions; in this group ICU mortality was 6.0 % and 5.4 % for those admitted to PICU and AICU respectively, a non-significant difference. The most common diagnostic categories for adolescents in AICU were trauma and drug/alcohol-related; those admitted to PICU most commonly had respiratory diagnoses. Conclusions: ICU mortality was higher for those admitted to PICU than AICU, but this may represent a greater severity of acute illness or underlying burden of chronic illness for adolescents cared for in a PICU. There is increasing recognition that adolescents need special consideration when planning preventative health-care and access to health services. Identifying the appropriate setting for the provision of intensive care for this population may be equally important.

of chronic illness for adolescents cared for in a PICU. There is increasing recognition that adolescents need special consideration when planning preventative health-care and access to health services. Identifying the appropriate setting for the provision of intensive care for this population may be equally important. A800 Comparison of postoperative sedation and analgesia of neonates in cardiac surgery: dexmedetomidine vs standard regimen S. Chernyshuk, H. Yemets, V. Zhovnir UCCC, ICU, Kyiv, Ukraine Correspondence: H. Yemets – UCCC, ICU, Kyiv, Ukraine Introduction: Sedation and analgesia are important components of postoperative management of neonates who underwent cardiac surgery. Excessive or inadequate sedation may have a significant adverse effect on patient outcome. Objectives. We aimed to determine which drug regimen would be most effective with less side-effect and better outcome. Methods: From March 2012 till March 2014 we conducted a randomized controlled prospective study in 60 neonates with congenital heart disease who underwent Arterial Switch Operation in our clinic. Inclusion criteria: 1) gestational age more than 36 weeks, 2) birth weight over 2500 g, 3) age - up to 28 days, 4) absence of concomitant diseases and surgical complications. Patients were randomized into 2 cohorts: 30 patients (50 %) were given infusion of dexmedetomidine with morphine boluses (study group) and 30 patients (50 %) were randomized to the standard regimen - infusion of morphine with diazepam boluses (control group).

Methods: From March 2012 till March 2014 we conducted a randomized controlled prospective study in 60 neonates with congenital heart disease who underwent Arterial Switch Operation in our clinic. Inclusion criteria: 1) gestational age more than 36 weeks, 2) birth weight over 2500 g, 3) age - up to 28 days, 4) absence of concomitant diseases and surgical complications. Patients were randomized into 2 cohorts: 30 patients (50 %) were given infusion of dexmedetomidine with morphine boluses (study group) and 30 patients (50 %) were randomized to the standard regimen - infusion of morphine with diazepam boluses (control group). Results: In both groups there were no differences in pre- and intraoperative indexes, duration of mechanical ventilation, sympathomimetic support, and time of infusion of dexmedetomidine/morphine. In study group time of ICU stay - 93.5 h - was significantly shorter than in control group -120 h (p-0.02). Onset of peristalsis and start of feeding in study group was earlier than in control group - 1-st vs 2-d day (p- 0.007) and 2-d vs 2.5-day (p-0.035), respectively. In the control group there were more patients who had complicated feeding (start after 3-d day, bloating or vomiting) - 11(37 %) vs 3(10 %) in the study group. We did not observe any decrease of mean blood pressure and heart rate in the study group as it could be expected.

7) and 2-d vs 2.5-day (p-0.035), respectively. In the control group there were more patients who had complicated feeding (start after 3-d day, bloating or vomiting) - 11(37 %) vs 3(10 %) in the study group. We did not observe any decrease of mean blood pressure and heart rate in the study group as it could be expected. Conclusion: Use of dexmedetomidine with morphine hydrochloride boluses for postoperative sedation and analgesia is effective and facilitates feeding process in neonates, leads to earlier onset of peristalsis and start of feeding, decreasing ICU stay.

7) and 2-d vs 2.5-day (p-0.035), respectively. In the control group there were more patients who had complicated feeding (start after 3-d day, bloating or vomiting) - 11(37 %) vs 3(10 %) in the study group. We did not observe any decrease of mean blood pressure and heart rate in the study group as it could be expected. Conclusion: Use of dexmedetomidine with morphine hydrochloride boluses for postoperative sedation and analgesia is effective and facilitates feeding process in neonates, leads to earlier onset of peristalsis and start of feeding, decreasing ICU stay. A801 Impact of positive end expiratory pressure on cerebral hemodynamic in paediatric patients with post-traumatic brain swelling treated by surgical decompression S.M. Pulitano’1, S. De Rosa1,2, A. Mancino1, G. Villa2,3, F. Tosi1, P. Franchi1, G. Conti1 1Catholic University, Department of Anesthesia and Intensive Care, Rome, Italy; 2International Renal Research Institute of Vicenza (IRRIV), Department of Nephrology, Dialysis and Transplantation, Vicenza, Italy; 3University of Florence, Department of Health Science, Section of Anaesthesiology and Intensive Care, Florence, Italy Correspondence: P. Franchi – Catholic University, Department of Anesthesia and Intensive Care, Rome, Italy Introduction: Current Brain Trauma recommendations are based to early correction of hypoxemia and avoidance of hypocarbia after severe paediatric TBI. The use of positive end-expiratory pressure (PEEP) in this situation remains controversial. Positive end expiratory pressure (PEEP) may reduce ventilator-induced lung injury by avoiding cyclic recruitment/derecruitment and prevent lung collapse. The aim of this investigation is to evaluate the impact of different PEEP levels on cerebral hemodynamic, gas exchange and respiratory system mechanics in paediatric patients with a severe post-traumatic brain swelling treated with decompressive craniectomy (DC).

lic recruitment/derecruitment and prevent lung collapse. The aim of this investigation is to evaluate the impact of different PEEP levels on cerebral hemodynamic, gas exchange and respiratory system mechanics in paediatric patients with a severe post-traumatic brain swelling treated with decompressive craniectomy (DC). Objectives: The aim of this investigation is to evaluate the impact of different PEEP levels on cerebral hemodynamic, gas exchange and respiratory system mechanics in paediatric patients with a severe post-traumatic brain swelling treated with decompressive craniectomy (DC). Methods: A prospective physiologic study was carried out on 14 paediatric patients presenting with severe traumatic brain swelling and treated with DC. Intracranial pressure (ICP), and cerebral perfusion pressure (CPP), central venous pressure (CVP), arterial oxygen saturation and the middle cerebral artery mean velocity (Vmed) was determined. After assessment at 0 PEEP (ZEEP), PEEP 4 and PEEP 8 were applied: all parameters were recorded at each level.

reated with DC. Intracranial pressure (ICP), and cerebral perfusion pressure (CPP), central venous pressure (CVP), arterial oxygen saturation and the middle cerebral artery mean velocity (Vmed) was determined. After assessment at 0 PEEP (ZEEP), PEEP 4 and PEEP 8 were applied: all parameters were recorded at each level. Results: The application of PEEP (from ZEEP to PEEP 8) significantly increased compliance of the respiratory system indexed to the weight of the patients (CrsI) (P = 0.02) without ICP modifications. No significant variations were observed in values of arterial pressure (MAP), CPP, Vmed, total resistance of the respiratory system indexed to the weight of the patients (RRSmaxI), and ohmic resistance of the respiratory system indexed to the weight of the patients (RRSminI). CVP significantly increased between ZEEP and PEEP 8 (P = 0.005), and between PEEP 4 and PEEP 8 (P = 0.05). Conclusions: We describe cerebral hemodynamic responses to PEEP application in pediatrics. PEEP values up to 8 cm H2O seem to be safe in paediatric patients with a severe post-traumatic brain swelling treated with DC. References: 1. Bein T, Kuhr LP, Bele S, Ploner F, Keyl C, Taeger K. Lung recruitment maneuver in patients with cerebral injury: effects on intracranial pressure and cerebral metabolism. Intensive Care Med 2002;28:554–8 2. Bor-Seng-Shu E, Hirsch R, Teixeira MJ, De Andrade AF, Marino R Jr (2006) Cerebral hemodynamic changes gauged by transcranial Dopp- ler ultrasonography in patients with posttraumatic brain swelling treated by surgical decompression. J Neurosurg 104:93–100

1. Bein T, Kuhr LP, Bele S, Ploner F, Keyl C, Taeger K. Lung recruitment maneuver in patients with cerebral injury: effects on intracranial pressure and cerebral metabolism. Intensive Care Med 2002;28:554–8 2. Bor-Seng-Shu E, Hirsch R, Teixeira MJ, De Andrade AF, Marino R Jr (2006) Cerebral hemodynamic changes gauged by transcranial Dopp- ler ultrasonography in patients with posttraumatic brain swelling treated by surgical decompression. J Neurosurg 104:93–100 A802 Randomized clinical trial of high concentration oxygen versus titrated oxygen therapy in pediatric asthma exacerbation B. Patel1, H. Khine2, A. Shah2, D. Sung2, L. Singer2 1The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States; 2The Children's Hospital at Montefiore, Bronx, United States Correspondence: B. Patel – The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States Introduction: Asthma exacerbation is one of the most common diagnoses seen in the pediatric ED. Several adult randomized controlled trials have shown that administration of high concentration oxygen leads to rise in carbon dioxide and increases admission rates. However, there are no studies in the pediatric population comparing the effects of high concentration oxygen versus titrated oxygen therapy in asthma exacerbation. Objectives: We evaluated the effects of transcutaneous carbon dioxide (tPaCO2) in high concentration oxygen therapy versus titrated oxygen therapy to maintain saturation between 92 to 95 % in pediatric patients with acute asthma exacerbation in the ED.

A802 Randomized clinical trial of high concentration oxygen versus titrated oxygen therapy in pediatric asthma exacerbation B. Patel1, H. Khine2, A. Shah2, D. Sung2, L. Singer2 1The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States; 2The Children's Hospital at Montefiore, Bronx, United States Correspondence: B. Patel – The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States Introduction: Asthma exacerbation is one of the most common diagnoses seen in the pediatric ED. Several adult randomized controlled trials have shown that administration of high concentration oxygen leads to rise in carbon dioxide and increases admission rates. However, there are no studies in the pediatric population comparing the effects of high concentration oxygen versus titrated oxygen therapy in asthma exacerbation. Objectives: We evaluated the effects of transcutaneous carbon dioxide (tPaCO2) in high concentration oxygen therapy versus titrated oxygen therapy to maintain saturation between 92 to 95 % in pediatric patients with acute asthma exacerbation in the ED. Methods: Children 2 to 18 years with previously diagnosed asthma with moderate to severe asthma exacerbation (asthma score > 5) were randomized to high concentration oxygen therapy (100 % oxygen via face mask at >4 L/min.) or titrated oxygen therapy (titrated up from 21 % via a blender continuously) to maintain saturations between 92 to 95 % while receiving their nebulized treatments. Exclusion criteria included disorders with hypercapnic respiratory failure, unconscious patient, history of congenital heart disease, pregnancy, history of smoking or using sedatives and depressants. Asthma therapy was provided per the ED physician. Asthma score, tPaCO2, PEFR (age >7 years) were measured at the start of the study and every 20 minutes for the first hour then every 30 minutes until disposition decision. The primary outcome was increase in tPaCO2 with high concentration oxygen therapy. Secondary outcome included rate of admission to the hospital.

an. Asthma score, tPaCO2, PEFR (age >7 years) were measured at the start of the study and every 20 minutes for the first hour then every 30 minutes until disposition decision. The primary outcome was increase in tPaCO2 with high concentration oxygen therapy. Secondary outcome included rate of admission to the hospital. Results: 73 patients were enrolled with mean age of 8.6 years. 60 % were males and 72 % had poorly controlled asthma with mean asthma score of 7.6. There were 36 patients enrolled in the high concentration oxygen group (HCOT) and 37 patients in the titrated oxygen group (TOT). The 0 minute tPaCO2 were not statistically different(35.6 ± 3.8 HCOT v. 37.4 ± 4.4 TOT,p = 0.07); whereas, the 20 minutes tPaCO2 was statistically different(40 ± 3.8 HCOT v. 37.5 ± 5.1 TOT, p = 0.02). The 60 minutes tPaCO2 was 39.2 ± 4.6 HCOT v. 35.5 ± 4.3 TOT, p = 0.0009. At 20 minutes, 89 % of the patients had a rise in tPaCO2 in HCOT v. 30 % in the TOT(p = < 0.0001), and at 60 minutes 78 % had a rise in tPaCO2 in HCOT v. 16 % in the TOT(p = < 0.0001). The asthma score was similar in the two groups at 0 minute (7.8 ± 1.4 HCOT v. 7.4 ± 1.3 TOT, p = 0.23); whereas, the 60 minutes asthma score was lower in the TOT(4.7 ± 1.5 HCOT v. 3.7 ± 1.3 TOT, p = 0.002). The rate of admission to the hospital was 36.1 % in HCOT v. 24.3 % in the TOT. Conclusions: High concentration oxygen therapy in pediatric asthma exacerbation leads to significantly higher carbon dioxide levels. It also causes rise in carbon dioxide from the baseline which increases the asthma scores and rate of admission. References:

Results: 73 patients were enrolled with mean age of 8.6 years. 60 % were males and 72 % had poorly controlled asthma with mean asthma score of 7.6. There were 36 patients enrolled in the high concentration oxygen group (HCOT) and 37 patients in the titrated oxygen group (TOT). The 0 minute tPaCO2 were not statistically different(35.6 ± 3.8 HCOT v. 37.4 ± 4.4 TOT,p = 0.07); whereas, the 20 minutes tPaCO2 was statistically different(40 ± 3.8 HCOT v. 37.5 ± 5.1 TOT, p = 0.02). The 60 minutes tPaCO2 was 39.2 ± 4.6 HCOT v. 35.5 ± 4.3 TOT, p = 0.0009. At 20 minutes, 89 % of the patients had a rise in tPaCO2 in HCOT v. 30 % in the TOT(p = < 0.0001), and at 60 minutes 78 % had a rise in tPaCO2 in HCOT v. 16 % in the TOT(p = < 0.0001). The asthma score was similar in the two groups at 0 minute (7.8 ± 1.4 HCOT v. 7.4 ± 1.3 TOT, p = 0.23); whereas, the 60 minutes asthma score was lower in the TOT(4.7 ± 1.5 HCOT v. 3.7 ± 1.3 TOT, p = 0.002). The rate of admission to the hospital was 36.1 % in HCOT v. 24.3 % in the TOT. Conclusions: High concentration oxygen therapy in pediatric asthma exacerbation leads to significantly higher carbon dioxide levels. It also causes rise in carbon dioxide from the baseline which increases the asthma scores and rate of admission. References: 1. Chien J.Uncontrolled oxygen administration and respiratory failure in acute asthma.Chest 2000;117(3):728–733.

Conclusions: High concentration oxygen therapy in pediatric asthma exacerbation leads to significantly higher carbon dioxide levels. It also causes rise in carbon dioxide from the baseline which increases the asthma scores and rate of admission. References: 1. Chien J.Uncontrolled oxygen administration and respiratory failure in acute asthma.Chest 2000;117(3):728–733. A803 Treatment of refractory status epilepticus with thiopental versus propofol in children: a randomized trial S. Haghbin1, S. Inaloo2, Z. Serati2 1Shiraz University of Medical Sciences, Pediatric intensive Care, Shiraz, Islamic Republic of Iran; 2Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran Correspondence: S. Haghbin – Shiraz University of Medical Sciences, Pediatric intensive Care, Shiraz, Islamic Republic of Iran Introduction: Refractory status epilepticus (RSE) is a life-threatening condition in which seizures do not respond to first- and second-line anticonvulsant drug therapies and is associated with increased hospital length of stay, mortality and functional disability (1). Coma induction is advocated for its management by different agents (2). Objectives: We aimed to assess the effectiveness (RSE control, adverse events) of propofol versus thiopental infusion in the treatment of RSE.

A803 Treatment of refractory status epilepticus with thiopental versus propofol in children: a randomized trial S. Haghbin1, S. Inaloo2, Z. Serati2 1Shiraz University of Medical Sciences, Pediatric intensive Care, Shiraz, Islamic Republic of Iran; 2Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran Correspondence: S. Haghbin – Shiraz University of Medical Sciences, Pediatric intensive Care, Shiraz, Islamic Republic of Iran Introduction: Refractory status epilepticus (RSE) is a life-threatening condition in which seizures do not respond to first- and second-line anticonvulsant drug therapies and is associated with increased hospital length of stay, mortality and functional disability (1). Coma induction is advocated for its management by different agents (2). Objectives: We aimed to assess the effectiveness (RSE control, adverse events) of propofol versus thiopental infusion in the treatment of RSE. Methods: In this randomized, single blind studying children aged 2 months- 18 years with RSE not due to cerebral ischemia were included. Medications were increased toward the EEG burst-suppression or to maximum limit of medication, and then progressively weaned. The primary endpoint was the proportion of patients with RSE controlled after a first course of study medication; secondary endpoints included clinical outcomes measures.

ischemia were included. Medications were increased toward the EEG burst-suppression or to maximum limit of medication, and then progressively weaned. The primary endpoint was the proportion of patients with RSE controlled after a first course of study medication; secondary endpoints included clinical outcomes measures. Results: In this study, 40 patients were included, 18 received propofol and 22 thiopental. RSE was generalized in 32 patients and focal in 8.The primary endpoint was reached in 72 % with propofol versus 54 % with thiopental (P = 0.33). However, mean duration of treatment with propofol was 50 hrs (range 12–94), and with thiopental was 10 days. While mortality (44 % vs23% P = 0.18), infection and systemic hypotension were similar in both groups, thiopental use was associated with longer mechanical ventilator (P = 0.02). More patients returned to basic condition at discharge with propofol (P = 0.04). Treatment failure was seen in 7/8 patients with focal convulsion. Two patients died due to propofol infusion syndrome with dose of 8 and 7 mg/kg/hr, so the maximum dose of propofol decreased to 6 afterward. Five patients died due to complications of thiopental infusion.

condition at discharge with propofol (P = 0.04). Treatment failure was seen in 7/8 patients with focal convulsion. Two patients died due to propofol infusion syndrome with dose of 8 and 7 mg/kg/hr, so the maximum dose of propofol decreased to 6 afterward. Five patients died due to complications of thiopental infusion. Conclusions: Although this study showed no significant difference between two groups regarding effectiveness, adverse effects and mortality, patients on propofol obtained quicker convulsion control and better return to baseline condition. A previous study did not disclose any difference between these two agents (3). However, care must be taken when it is used in longer than 24 hours with higher dose than 6 mg/kg/hr . Trial Registration: IRCT.IR IRCT138707231349N1 References 1. Abend NS, Dlugos DJ. Treatment of refractory status epilepticus: literature review and a proposed protocol. Pediatr Neurol. 2008,38(6): 37790. 2. Brophy GM, Bell R, Claassen J, et al. Guidelines for the evaluation and management of status epilepticus. Neurocrit Care 2012; 17:3. 3. Rossetti AO, Milligan TA, Vulliémoz S, Michaelides C, Bertschi M, Lee JW. A randomized trial for the treatment of refractory status epilepticus. Neurocrit Care. 2011 Feb;14(1):4–10.

1. Abend NS, Dlugos DJ. Treatment of refractory status epilepticus: literature review and a proposed protocol. Pediatr Neurol. 2008,38(6): 37790. 2. Brophy GM, Bell R, Claassen J, et al. Guidelines for the evaluation and management of status epilepticus. Neurocrit Care 2012; 17:3. 3. Rossetti AO, Milligan TA, Vulliémoz S, Michaelides C, Bertschi M, Lee JW. A randomized trial for the treatment of refractory status epilepticus. Neurocrit Care. 2011 Feb;14(1):4–10. A804 An evaluation and accuracy of new zero-heat-flux thermometer (3 M SpotOn) in pediatric intensive care patients M. Idei1, T. Nomura2, N. Yamamoto1, Y. Sakai1, T. Yoshida1, Y. Matsuda1, Y. Yamaguchi1, S. Takaki1, O. Yamaguchi1, T. Goto2 1Yokohama City University Hospital, Intensive Care Unit, Yokohama, Japan; 2Yokohama City University Hospital, Department of Anesthesiology, Yokohama, Japan Correspondence: M. Idei – Yokohama City University Hospital, Intensive Care Unit, Yokohama, Japan Introduction: In critically ill patients, temperature measurement is a routine important care task and can lead to important decisions. Rectal temperature and bladder temperature are now used as a continuous body temperature measuring method in the pediatric intensive care, but these practices have several disadvantages including the patient´s discomfort, the risk of organ injury and the inaccurate measuring caused by the sensor position. A new temperature monitoring system 3MTM SpotOnTM (SpotOn) is a non-invasive zero-heat-flux thermometer designed to estimate core body temperature from the skin surface. Although the usefulness and accuracy of SpotOn system in adult patients have been demonstrated, there are no reports on pediatric intensive care patients.

w temperature monitoring system 3MTM SpotOnTM (SpotOn) is a non-invasive zero-heat-flux thermometer designed to estimate core body temperature from the skin surface. Although the usefulness and accuracy of SpotOn system in adult patients have been demonstrated, there are no reports on pediatric intensive care patients. Objectives: The aim of this study was to evaluate the effectiveness of a new temperature measurement system attached to the forehead, and compare it to rectal temperature sensors in terms of correlation and accuracy. Methods: Pediatric patients weighing less than 10Kg, who were managed in our ICU during the period from February 2015 to March 2016, were enrolled in this study. Core temperature was measured and recorded at every minute from the both thermistor of a rectal thermal probe and with SpotOn in these patients. The data when the forehead sensor or rectal probe was taken out for nursing care was excluded from statistical analysis. Results: 53495 sets of data of 26 children (Mean BW 5630 g) were examined retrospectively. In all patients, SpotOn showed higher than the rectal temperatures. The SpotOn temperature was analyzed to be 0.82 degrees (95 % limits of agreement of ± 0.51) higher temperature than the rectal one with a moderate correlation(r = 0.73).

53495 sets of data of 26 children (Mean BW 5630 g) were examined retrospectively. In all patients, SpotOn showed higher than the rectal temperatures. The SpotOn temperature was analyzed to be 0.82 degrees (95 % limits of agreement of ± 0.51) higher temperature than the rectal one with a moderate correlation(r = 0.73). Discussion and conclusion: Rectal temperature measurement is the gold standard method for pediatric patients in ICU despite several complications of rectal injury. Our children´s study demonstrated the slightly higher temperature in the SpotOn than rectal temperature with a substantial correlation. One possible explanation could be that the abundance of brain blood flow of children affected the results. Our study concluded that SpotOn system could be used as a highly reliable noninvasive core body temperature measurement for small pediatric patients. References 1. Eshraghi Y, Nasr V, Parra-Sanchez I et al. An evaluation of a zero-heat-flux cutaneous thermometer in cardiac surgical patients. Anesth Analg. 2014 Sep;119(3):543–9. 2. Hebbar K, Fortenberry JD, Rogers K, et al. Comparison of temporal artery thermometer to standard temperature measurements in pediatric intensive care unit patients. Pediatr Crit Care Med. 2005; 6: 557–561. Grant acknowledgement None

1. Eshraghi Y, Nasr V, Parra-Sanchez I et al. An evaluation of a zero-heat-flux cutaneous thermometer in cardiac surgical patients. Anesth Analg. 2014 Sep;119(3):543–9. 2. Hebbar K, Fortenberry JD, Rogers K, et al. Comparison of temporal artery thermometer to standard temperature measurements in pediatric intensive care unit patients. Pediatr Crit Care Med. 2005; 6: 557–561. Grant acknowledgement None A805 Viral bronchiolitis in pediatric acute respiratory distress syndrome N. Longani, S. Medar The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States Correspondence: N. Longani – The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States Introduction: Viral bronchiolitis (VB) remains one of the leading causes of hospitalization in early childhood. Despite the heavy burden of VB on the healthcare system, little is known about the incidence of Acute Respiratory Distress Syndrome (ARDS) in this cohort of patients. In 2015, the Pediatric Acute Lung Injury Consensus Conference (PALICC) published guidelines for the definition, management and research in pediatric ARDS (PARDS) (1). Objectives: To study the incidence and prevalence of PARDS in VB and to study the association between PARDS and specific PICU outcomes such as incidence of mechanical ventilation, noninvasive ventilator settings length of PICU stay in this group of patients.

A805 Viral bronchiolitis in pediatric acute respiratory distress syndrome N. Longani, S. Medar The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States Correspondence: N. Longani – The Children's Hospital at Montefiore, Pediatric Critical Care, Bronx, United States Introduction: Viral bronchiolitis (VB) remains one of the leading causes of hospitalization in early childhood. Despite the heavy burden of VB on the healthcare system, little is known about the incidence of Acute Respiratory Distress Syndrome (ARDS) in this cohort of patients. In 2015, the Pediatric Acute Lung Injury Consensus Conference (PALICC) published guidelines for the definition, management and research in pediatric ARDS (PARDS) (1). Objectives: To study the incidence and prevalence of PARDS in VB and to study the association between PARDS and specific PICU outcomes such as incidence of mechanical ventilation, noninvasive ventilator settings length of PICU stay in this group of patients. Methods: This is a retrospective single center observational cohort study that examined children 0–2 years of age admitted to the PICU with VB and respiratory failure (RF) from 2011–2014. PALICC criteria were applied to define PARDS. Clinical and demographic data was collected. Patients with a diagnosis of congenital heart disease or pre-existing chronic lung disease were excluded. Data was expressed as median with IQR ranges. Test of bivariate association were performed using Mann Whitney U test and chi square test. A two tailed p value of ≤ 0.05 was used to denote statistical significance.

Patients with a diagnosis of congenital heart disease or pre-existing chronic lung disease were excluded. Data was expressed as median with IQR ranges. Test of bivariate association were performed using Mann Whitney U test and chi square test. A two tailed p value of ≤ 0.05 was used to denote statistical significance. Results: Out of 1700 patients with RF, 330 with VB met study criteria. Eighty of these 330 (24 %) patients admitted for VB met the criteria for PARDS or at risk for PARDS. Out of these 80 patients, 25 (31 %) met criteria for PARDS and 55 (69 %) met criteria for “at risk of PARDS”. Median age was 5 (2,11) months and the median weight was 6.9 (5.3, 9.5) kgs. Most common etiology for VB was respiratory syncytial virus (RSV) 68 % followed by Rhinovirus (20 %). There was no statistically significant difference in age, weight, and etiology of VB in patients with PARDS and those “at risk of PARDS.” Patients with PARDS had longer hospital and PICU length of stay (LOS) and more likely to receive diuretics compared to those “at risk for PARDS” (16 (10, 21) Vs 8 (6, 10.5), p = 0.0001; 10 (7, 13) Vs 3 (2, 4.5), p < 0.0001; and 66 % vs 33 %, p = 0.02 respectively). Nineteen (19/25, 76 %) patients with PARDS received invasive mechanical ventilation with a median duration of ventilation of 6 (1, 10) days.

to receive diuretics compared to those “at risk for PARDS” (16 (10, 21) Vs 8 (6, 10.5), p = 0.0001; 10 (7, 13) Vs 3 (2, 4.5), p < 0.0001; and 66 % vs 33 %, p = 0.02 respectively). Nineteen (19/25, 76 %) patients with PARDS received invasive mechanical ventilation with a median duration of ventilation of 6 (1, 10) days. Conclusions: Almost a quarter of children with VB developed PARDS or were at risk of PARDS. The presence of PARDS in children with VB was significantly associated with longer PICU and Hospital LOS compared to those “at risk of PARDS”. Children with VB are a high risk group for the development of PARDS. References 1) Pediatric Acute Lung Injury Consensus Conference Group, et al.Pediatric acute respiratory distress syndrome:consensus recommendations from the Pediatric Acute Lung Injury Consensus Conference. Pediatric Crit Care Med 2015 Jun; 5: 428–439 2) Zorc J, Hall C. Bronchiolitis: recent evidence on diagnosis and management. Pediatrics 2010; 25:342–349

1) Pediatric Acute Lung Injury Consensus Conference Group, et al.Pediatric acute respiratory distress syndrome:consensus recommendations from the Pediatric Acute Lung Injury Consensus Conference. Pediatric Crit Care Med 2015 Jun; 5: 428–439 2) Zorc J, Hall C. Bronchiolitis: recent evidence on diagnosis and management. Pediatrics 2010; 25:342–349 A806 Mean platelet volume dynamics and platelet count as prognostic indicators in pediatric surgical intensive care: a descriptive observational study I.R. Abdel-Aal1, A.S. El Adawy2, H.M.E.-H. Mohammed2, A.N. Mohamed2 1Cairo University/Kasr Alainy Medical School, Anesthesia, Pain and Surgical ICU, Cairo, Egypt; 2Cairo University/Kasr Alainy Medical School, Cairo, Egypt Correspondence: I.R. Abdel-Aal – Cairo University/Kasr Alainy Medical School, Anesthesia, Pain and Surgical ICU, Cairo, Egypt Introduction: Mean platelet volume(MPV) seems to be a marker of platelet activation and may be related to severity of illness.1Changes in MPV and platelet count(PLC)could be used for disease prognosis and mortality in ICU patients.2We hypothesized that MPV changes and PLC could be used as prognostic tools in pediatric surgical intensive care units(PSICU). Objectives: To study the association between MPV changes and mortality and morbidity in PSICU. Also to study the relation between PLC and PSICU mortality and morbidity.

A806 Mean platelet volume dynamics and platelet count as prognostic indicators in pediatric surgical intensive care: a descriptive observational study I.R. Abdel-Aal1, A.S. El Adawy2, H.M.E.-H. Mohammed2, A.N. Mohamed2 1Cairo University/Kasr Alainy Medical School, Anesthesia, Pain and Surgical ICU, Cairo, Egypt; 2Cairo University/Kasr Alainy Medical School, Cairo, Egypt Correspondence: I.R. Abdel-Aal – Cairo University/Kasr Alainy Medical School, Anesthesia, Pain and Surgical ICU, Cairo, Egypt Introduction: Mean platelet volume(MPV) seems to be a marker of platelet activation and may be related to severity of illness.1Changes in MPV and platelet count(PLC)could be used for disease prognosis and mortality in ICU patients.2We hypothesized that MPV changes and PLC could be used as prognostic tools in pediatric surgical intensive care units(PSICU). Objectives: To study the association between MPV changes and mortality and morbidity in PSICU. Also to study the relation between PLC and PSICU mortality and morbidity. Methods: This descriptive observational study was conducted on consecutive 100 pediatric surgical patients who admitted to PSICUs at Cairo University Hospitals starting from 1/6-1/12/2015.After approval by research ethics committee,informed consents were obtained from parents and pediatric cases aged from 1 month-18 years and stayed for > 48 h were enrolled.MPV and PLC were obtained and recorded at baseline(pre-operative values),on the day of ICU admission(day 0),1st,2nd,3rd,5th and 7th days.To measure daily MPV changes; (ΔMPV) was constructed and computed where ΔMPV = ([MPVday(X) − MPVday (0)]/MPVday(0) × 100 %. Pediatric Index of Mortality(PIM)score was calculated on day 0 and the Pediatric Logistic Organ Dysfunction(PELOD)Score was recorded daily.

day of ICU admission(day 0),1st,2nd,3rd,5th and 7th days.To measure daily MPV changes; (ΔMPV) was constructed and computed where ΔMPV = ([MPVday(X) − MPVday (0)]/MPVday(0) × 100 %. Pediatric Index of Mortality(PIM)score was calculated on day 0 and the Pediatric Logistic Organ Dysfunction(PELOD)Score was recorded daily. Results: Patients who developed ICU complications (fever, sepsis, pneumonia, required mechanical ventilation, needed vasopressors or blood transfusion); showed higher ΔMPV compared to non complicated cases (Fig. 2). This association was statistically significant on days 2 (p value = 0.035),3(p value < 0.001), 5 (p value < 0.001) and 7(p value = 0.017) of ICU stay but it´s insignificant on day1(p value =0.691).According to receiver operating characteristics(ROC) curve analysis, the sensitivity of ΔMPV to detect complications on day 2 was 57.2 % but its specificity on day 2 was 76.6 %.Patients who developed ICU complications showed lower PLC compared to non complicated cases(Fig. 3).This association was statistically significant on days1(p value < 0.001),2(p value < 0.001) and 3(p value < 0.001) but it was insignificant on day 0(p value =0.237 ),5(p value =0.861) and 7(p value =0.247). On other hand, the sensitivity of PLC to detect complications day1 was 81.4 % but the specificity was 71.9 %, while the sensitivity of PLC to detect complications day 2 was 81.1 % but the specificity day 2 was 100 %.

ue < 0.001) but it was insignificant on day 0(p value =0.237 ),5(p value =0.861) and 7(p value =0.247). On other hand, the sensitivity of PLC to detect complications day1 was 81.4 % but the specificity was 71.9 %, while the sensitivity of PLC to detect complications day 2 was 81.1 % but the specificity day 2 was 100 %. Conclusions: MPV dynamics and PLC have prognostic roles and could be used in determining several complications in critically ill pediatric surgical patients. PLC is a more specific and sensitive tool to detect complications than mean MPV dynamics. References 1- Cekmez F et al. Mean platelet volume in very preterm infants: a predictor of morbidities.Eur Rev Med Pharmacol Sci. 2013; 17: 134–137. 2- Cengizhan S, et al. Alterations in platelet count and mean platelet volume as predictors of patient outcome in the respiratory intensive care unit. Clin Respir J. 2014;5:35–40.Fig. 2 (abstract A806). Percentage changes in MPV (Delta MPV) among ICU co. Fig. 3 (abstract A806). Receiver operating characteristics (ROC) curve for Fig. 4 (abstract A806). Platelets count (PLC) among intensive care unit.

2- Cengizhan S, et al. Alterations in platelet count and mean platelet volume as predictors of patient outcome in the respiratory intensive care unit. Clin Respir J. 2014;5:35–40.Fig. 2 (abstract A806). Percentage changes in MPV (Delta MPV) among ICU co. Fig. 3 (abstract A806). Receiver operating characteristics (ROC) curve for Fig. 4 (abstract A806). Platelets count (PLC) among intensive care unit. REHABILITATION & RECOVERY FROM ICU A807 Physical function in critical care (Pacific): a multi-centre observational study S.M. Parry1, L.D. Knight2, L. Denehy1, N. De Morton3, C.E. Baldwin4,5, D. Sani6, G. Kayambu6, V.Z.M. da Silva7,8, P. Phongpagdi1, Z.A. Puthucheary9,10,11, C.L. Granger1,2 1The University of Melbourne, Department of Physiotherapy, Melbourne, Australia; 2Royal Melbourne Hospital, Department of Physiotherapy, Melbourne, Australia; 3Peter MacCallum Cancer Centre, Department of Physiotherapy, Melbourne, Australia; 4Flinders Medical Centre, Department of Physiotherapy, Adelaide, Australia; 5University of South Australia, Member of the International Centre for Allied Health Evidence (iCAHE) and the Sansom Institute, Adelaide, Australia; 6National University Hospital, Department of Rehabilitation, Singapore, Singapore; 7Escola Superior da Saude, Health Sciences Program, Brasilia, Brazil; 8Hospital de Base do Distrito Federal, Brasilia, Brazil; 9University College London Hospitals, Division of Critical Care, London, United Kingdom; 10University College London Hospitals, Institute of Sports and Exercise Health, London, United Kingdom; 11National University Hospital, Division of Respiratory and Critical Care, Singapore, Singapore Correspondence: S.M. Parry – The University of Melbourne, Department of Physiotherapy, Melbourne, Australia Introduction: Impairment in physical function is a significant problem for survivors of critical illness [1,2]. There is growing urgency to develop a core set of outcome measures which can be adopted in clinical and research practice to evaluate efficacy in response to interventions such as rehabilitation. There is currently not a single outcome measure which can be used across the continuum from ICU admission to hospital discharge for individuals with critical illness [3].

of outcome measures which can be adopted in clinical and research practice to evaluate efficacy in response to interventions such as rehabilitation. There is currently not a single outcome measure which can be used across the continuum from ICU admission to hospital discharge for individuals with critical illness [3]. Objectives: (1) To determine the clinical utility of two physical function measures: De Morton Mobility Index (DEMMI) and Physical Function in Intensive Care test-scored (PFIT-s) when used in isolation across the hospital admission; and (2) To transform the (15-item) DEMMI and (4-item) PFIT-s into a single measure to evaluate function in ICU survivors using rasch analytical principles. Methods: Multi-centre prospective observational study conducted across four sites internationally. Consecutive eligible participants were recruited who met inclusion criteria; ; Adults > 18 years of age whom were mechanically ventilated > 48 hours and were ambulant at least 10 metres independently prior to their ICU admission. Physical function was evaluated at ICU awakening, and both ICU and hospital discharge using the PFIT-s and DEMMI, administered in a randomised sequence using concealed allocation on each measurement occasion to minimise bias in testing order.

ction: Survivors of critical illness experience a range of impairments after intensive care, including physical, cognitive and psychological compromise. The provision of information using a diary to describe the intensive care unit (ICU) experience is one strategy that has been proposed to improve psychological health. Objectives: The purpose of this study was to explore similarities and differences in patients' and relatives' perceptions of information containing strategies, including ICU diaries, to assist recovery after critical illness. Methods: An exploratory mixed-methods study was undertaken in an Australian tertiary hospital with general ICU patients admitted for ≥3 days and their relatives. Semi-structured interviews were conducted 3–5 months after ICU discharge. Transcripts were analysed using content analysis.

were ambulant at least 10 metres independently prior to their ICU admission. Physical function was evaluated at ICU awakening, and both ICU and hospital discharge using the PFIT-s and DEMMI, administered in a randomised sequence using concealed allocation on each measurement occasion to minimise bias in testing order. Results: 128 participants have been recruited into the study to date across the four sites. 61 % were male (n = 78) with median age of 65 [53–73]; and moderate severity of illness (median [IQR] APACHE II: 22 [17–27]). Median [IQR] ICU and hospital LOS were 9 [5–14] and 21 [13–37] days respectively. The incidence of ICU-acquired weakness was 50 % (n = 67). Aim 1: On awakening mean ± SD PFIT-s was 4.9 ± 2.5 (out of 10) and DEMMI was 19 ± 21 (out of 100). In isolation the PFIT-s had a floor effect of 9 % (n = 11) at ICU awakening, and 1 % (n = 1) at both ICU and hospital discharge; and a large ceiling effect at hospital discharge of 42 % (n = 40). The DEMMI in isolation had a large floor effect in the ICU of 23 % at awakening, and a small ceiling effect at hospital discharge of 14 % (n = 14). Both the PFIT-s and DEMMI were demonstrated to be highly responsive to change in functional recovery over the acute hospitalisation period (p < 0.005). Aim 2: Preliminary exploration of a subgroup with complete data at hospital discharge (n = 73) was evaluated.The data fit the Rasch model Chi squared =10.4, df = 24, p = 0.99 with no item misfit or differential item functioning based on age, gender, BMI, severity of illness (APACHE II) or comorbidity. A new single measure (12-items) has been proposed combining the DEMMI and PFIT-s.

at hospital discharge (n = 73) was evaluated.The data fit the Rasch model Chi squared =10.4, df = 24, p = 0.99 with no item misfit or differential item functioning based on age, gender, BMI, severity of illness (APACHE II) or comorbidity. A new single measure (12-items) has been proposed combining the DEMMI and PFIT-s. Conclusions: The PFIT-s and DEMMI have limitations when used in isolation. A new transformed scale based on rasch analytical principles is promising combining features of both tools for evaluation of functional recovery of critically ill. A808 Functional status at ICU admission, physical therapy treatment and critical care outcomes J.E. Rydingsward1, C.M. Horkan2, K.B. Christopher3,4 1Brigham and Women's Hospital, Department of Rehabilitation, Boston, United States; 2Brigham and Women's Hospital, Department of Medicine, Boston, United States; 3Brigham and Women's Hospital, Renal Division, Boston, United States; 4Brigham and Women's Hospital, Channing Division of Network Medicine, Boston, United States Correspondence: J.E. Rydingsward – Brigham and Women's Hospital, Department of Rehabilitation, Boston, United States Introduction: Limited information exists regarding the association between functional status at ICU admission at and outcomes. Objectives: We hypothesized that initial functional status assessment as well the amount of physical therapy delivered would be associated with outcomes in ICU survivors.

A808 Functional status at ICU admission, physical therapy treatment and critical care outcomes J.E. Rydingsward1, C.M. Horkan2, K.B. Christopher3,4 1Brigham and Women's Hospital, Department of Rehabilitation, Boston, United States; 2Brigham and Women's Hospital, Department of Medicine, Boston, United States; 3Brigham and Women's Hospital, Renal Division, Boston, United States; 4Brigham and Women's Hospital, Channing Division of Network Medicine, Boston, United States Correspondence: J.E. Rydingsward – Brigham and Women's Hospital, Department of Rehabilitation, Boston, United States Introduction: Limited information exists regarding the association between functional status at ICU admission at and outcomes. Objectives: We hypothesized that initial functional status assessment as well the amount of physical therapy delivered would be associated with outcomes in ICU survivors. Methods: We performed a retrospective cohort study in one Boston teaching hospital on 2,828 adults who received critical care from 1997 to 2011 and survived hospitalization. All patients had a formal evaluation by a physical therapist in the week prior to ICU admission and at hospital discharge. The exposure of interest was functional status determined by a licensed physical therapist based on the functional mobility sub scales of the Functional Independence Measure. All patients received physical therapy to improve functional performance. The primary outcome was 90-day all-cause mortality. We used logistic regression to describe how 90-day mortality differed with functional status at ICU admission. Negative binomial regression was utilized to describe how functional status at hospital discharge differed with functional status at ICU admission, the extent of physical therapy received and hospital length of stay.

logistic regression to describe how 90-day mortality differed with functional status at ICU admission. Negative binomial regression was utilized to describe how functional status at hospital discharge differed with functional status at ICU admission, the extent of physical therapy received and hospital length of stay. Results: The cohort was 52 % male, 22 % non-white and had a mean age of 64.1 years. 10 % of the cohort had sepsis, 7 % had acute kidney injury, 32 % had respiratory failure and 53 % were surgical cases. The median [IQR] hospital length of stay was 8 [4, 14] days. The 90-day mortality rate was 14.6 %. Functional status at ICU admission was robustly associated with 90-day mortality. In a logistic regression model adjusted for age, gender, race, surgical patient type, Deyo-Charlson index, acute organ failure, sepsis, length of stay and the extent of physical therapy received, the second lowest and lowest quartiles of functional status at ICU admission was associated with a 1.8 and 2.3 fold increased odds of 90-day mortality respectively, compared to patients with the highest quartile of functional status [OR = 1.80(95%CI 1.26-2.57) and OR = 2.34(95%CI 1.63-3.36)]. Every 15 minute increment in physical therapy completed was associated with a decrease in the adjusted odds of 90-day mortality [OR = 0.60 (95%CI 0.53-0.68)]. Further, in survivors of hospitalization (n = 2,364), patients with the second lowest and lowest quartiles of functional status at ICU admission had a 2.7 and 3.4-fold lower functional status assessed at hospital discharge following adjustment, compared to patients with the highest quartile of functional status [IRR 2.74 (95%CI 2.50-3.01) and IRR 3.42 (95%CI 3.10-3.77)] respectively.

Objectives: The purpose of this study was to explore similarities and differences in patients' and relatives' perceptions of information containing strategies, including ICU diaries, to assist recovery after critical illness. Methods: An exploratory mixed-methods study was undertaken in an Australian tertiary hospital with general ICU patients admitted for ≥3 days and their relatives. Semi-structured interviews were conducted 3–5 months after ICU discharge. Transcripts were analysed using content analysis. Results: Twenty-two patients and 19 relatives consented to participation and completed interviews prior to reaching data saturation. Patients were usually male (63 %) and aged 52 ± 14 years. Patients raised similar themes to relatives, although with diverse opinions. Themes of wanting to have a diary kept and considering they would find a diary helpful were consistent across a majority of participants, although with a minority expressing a desire to 'move on' and would not have liked a diary kept. Differences between patients and relatives arose in the areas of the purpose, content, ownership and timing of delivery of a diary. Patients viewed the diary as a therapeutic tool while relatives considered it as an information sharing mechanism, including as a mechanism to demonstrate to the patient 'how sick he really was' and 'what he put us through'. Possibly as a result of these differences, patients considered that ownership of the diary rested with them while some relatives envisaged shared ownership. Patients were more likely to note that the diary should not be provided to them until some weeks after ICU while relatives considered an early time point soon after ICU discharge to be appropriate. Patients were more likely to raise concerns about the potential negative impact of information sharing strategies including diaries and return visits to the ICU.

h the second lowest and lowest quartiles of functional status at ICU admission had a 2.7 and 3.4-fold lower functional status assessed at hospital discharge following adjustment, compared to patients with the highest quartile of functional status [IRR 2.74 (95%CI 2.50-3.01) and IRR 3.42 (95%CI 3.10-3.77)] respectively. Conclusions: In critically ill patients, decreased functional status at ICU admission is associated with increased 90-day mortality. Increased intensity of physical therapy is associated with improved mortality outcomes. Both functional status at ICU admission and the intensity of physical therapy contribute to functional status determined at hospital discharge. A809 Does enhanced physiotherapy and early mobilisation reduce the degree of muscle loss for patients admitted to critical care? D. McWilliams, C. Jones, E. Reeves, G. Atkins, C. Snelson Queen Elizabeth Hospital NHS FT, Birmingham, United Kingdom Correspondence: D. McWilliams – Queen Elizabeth Hospital NHS FT, Birmingham, United Kingdom Introduction: Patients admitted to critical care are shown to lose significant muscle mass, with the degree of muscle loss as high as 20 % in the first week for those in multi organ failure (Puthucheary, 2013). Early rehabilitation has been demonstrated as a safe and effective method of improving functional status at the point of critical care discharge and reducing both ICU and hospital length of stay (McWilliams et al., 2015), although the specific impact of this on muscle mass has not been reported.

re (Puthucheary, 2013). Early rehabilitation has been demonstrated as a safe and effective method of improving functional status at the point of critical care discharge and reducing both ICU and hospital length of stay (McWilliams et al., 2015), although the specific impact of this on muscle mass has not been reported. Objectives: This study aimed to analyse the impact of enhanced physiotherapy incorporating early mobilisation on the rate of muscle decline for patients admitted to critical care. Methods: Patients admitted to a large UK teaching hospital during the trial period and ventilated for ≥ 5 days were included in the study. Patients were randomised to either enhanced physiotherapy or standard care groups as part of a larger RCT. Baseline measurements were taken on the day of recruitment and then repeated at critical care discharge. Muscle mass was measured using ultrasound to calculate cross sectional area of quadriceps and biceps. To ensure validity , 2 measures were taken and the average of these taken as the final value. All scans were reviewed for agreement by 2 therapists trained in muscle ultrasound.

hen repeated at critical care discharge. Muscle mass was measured using ultrasound to calculate cross sectional area of quadriceps and biceps. To ensure validity , 2 measures were taken and the average of these taken as the final value. All scans were reviewed for agreement by 2 therapists trained in muscle ultrasound. Results: 40 patients were included in the analysis. Patients in the enhanced physiotherapy group mobilised earlier and achieved a higher level of mobility at the point of critical care discharge (see Table 2.) All subjects demonstrated a reduction in muscle mass of both quadriceps and biceps over the course of their critical care stay. The extent of muscle loss was however lower in those receiving the enhanced physiotherapy, although this did not reach statistical significance (Quads 3 % vs 13 %. p = 0.14317; Biceps 6 % vs 13 %, p = 0.12033).Table 2 (abstract A809). Demongraphics Enhanced (n = 20) Control (n = 20) Median Age (years) 62.5 60 Sex (male) 14 (70 %) 10 (50 %) Median Time to Mobilise 8 days 9.5 days Median MMS at critical care discharge 7 5.5 Median Critical care length of stay 16.5 days 18 days Table 3 (abstract A809). USS measurements Baseline Critical care discharge Diff Control Quads 2.65 2.30 −0.35 (13 %) Enhanced Quads 2.55 2.48 −0.07 (3 %) Control Biceps 2.66 2.32 −0.34 (13 %) Enhanced Biceps 2.71 2.55 −0.16 (6 %) Conclusions: A programme of enhanced physiotherapy appeared to be associated with a lower rate of muscle loss in both biceps and quadriceps in comparison to standard care. An appropriately powered RCT is required to assess these findings.

Baseline Critical care discharge Diff Control Quads 2.65 2.30 −0.35 (13 %) Enhanced Quads 2.55 2.48 −0.07 (3 %) Control Biceps 2.66 2.32 −0.34 (13 %) Enhanced Biceps 2.71 2.55 −0.16 (6 %) Conclusions: A programme of enhanced physiotherapy appeared to be associated with a lower rate of muscle loss in both biceps and quadriceps in comparison to standard care. An appropriately powered RCT is required to assess these findings. References 1. Puthucheary ZA, Rawal J, McPhail M et al. (2013) Acute Skeletal Muscle Wasting in Critical Illness. JAMA. 2013;310(15):1591–1600 2. McWilliams D, Weblin J, Atkins G et al. (2014) Enhancing rehabilitation of mechanically ventilated patients in the intensive care unit: A quality improvement project. Journal of critical care. http://dx.doi.org/10.1016/j.jcrc.2014.09.18

1. Puthucheary ZA, Rawal J, McPhail M et al. (2013) Acute Skeletal Muscle Wasting in Critical Illness. JAMA. 2013;310(15):1591–1600 2. McWilliams D, Weblin J, Atkins G et al. (2014) Enhancing rehabilitation of mechanically ventilated patients in the intensive care unit: A quality improvement project. Journal of critical care. http://dx.doi.org/10.1016/j.jcrc.2014.09.18 A810 Similarities and differences in patients' and relatives' views of information provision after ICU L.M. Aitken1,2,3, J. Rattray4, J. Kenardy5, A.M. Hull6, A. Ullman2, R. Le Brocque5, M. Mitchell2,3, C. Davis3, B. Macfarlane2,3 1City University London, School of Health Sciences, London, United Kingdom; 2Griffith University, NHMRC Centre of Research Excellence in Nursing, Menzies Health Institute Queensland, Brisbane, Australia; 3Princess Alexandra Hospital, Intensive Care Unit, Brisbane, Australia; 4University of Dundee, Dundee, United Kingdom; 5University of Queensland, Brisbane, Australia, 6NHS Tayside, Dundee, United Kingdom Correspondence: L.M. Aitken – City University London, School of Health Sciences, London, United Kingdom Introduction: Survivors of critical illness experience a range of impairments after intensive care, including physical, cognitive and psychological compromise. The provision of information using a diary to describe the intensive care unit (ICU) experience is one strategy that has been proposed to improve psychological health.

the diary should not be provided to them until some weeks after ICU while relatives considered an early time point soon after ICU discharge to be appropriate. Patients were more likely to raise concerns about the potential negative impact of information sharing strategies including diaries and return visits to the ICU. Conclusions: Patients and relatives expressed common themes related to information sharing strategies after ICU, but with some important differences. Differences in purpose, content, ownership and timing of delivery of a diary suggest there is a need to consider whether the same intervention meets the needs of both groups of stakeholders. Grant acknowledgement: This project was funded by the NHMRC Centre of Research Excellence in Nursing, Menzies Health Institute Queensland, Griffith University, Australia. A811 PDCA for increasing effective use of CAM-ICU for delirium screening by critical care nurses J.C. Azevedo, L.L. Rocha, F.F.M. De Freitas, A.M. Cavalheiro, N.M. Lucinio, M.S. Lobato Hospital Israelita Albert Einstein, São Paulo, Brazil Correspondence: M.S. Lobato – Hospital Israelita Albert Einstein, São Paulo, Brazil Introduction: Confusion assessment method (CAM-ICU) is routinely used for delirium screening in ICU. In research environment, this tool has a very high sensitivity and specificity (1). However, in clinical settings, it may not be reproducible mainly because of inadequate training in CAM-ICU by bedside nurses (2). Objectives: Develop a PDCA (plan-do-check-act) to train bedside critical care nurses in CAM-ICU application.

A811 PDCA for increasing effective use of CAM-ICU for delirium screening by critical care nurses J.C. Azevedo, L.L. Rocha, F.F.M. De Freitas, A.M. Cavalheiro, N.M. Lucinio, M.S. Lobato Hospital Israelita Albert Einstein, São Paulo, Brazil Correspondence: M.S. Lobato – Hospital Israelita Albert Einstein, São Paulo, Brazil Introduction: Confusion assessment method (CAM-ICU) is routinely used for delirium screening in ICU. In research environment, this tool has a very high sensitivity and specificity (1). However, in clinical settings, it may not be reproducible mainly because of inadequate training in CAM-ICU by bedside nurses (2). Objectives: Develop a PDCA (plan-do-check-act) to train bedside critical care nurses in CAM-ICU application. Methods: The study was conducted in a 600 bed tertiary private hospital in Sao Paulo, Brazil. A pre-training questionnaire to test bedside nurses about their knowledge of the correct application of CAM-ICU was applied. Later on, training sessions consisting of video lessons and practical mentored application of CAM-ICU were developed. Also, internal campaigns were developed to increase awareness about CAM-ICU between nurses. A post-training test was applied. Those who had a final test score higher than 90 % was approved. Those who had a final test score was ≤90 % were submitted to another round of training sections and another post-training test was applied until necessary score obtained. The instructors audited all in-training nurses CAM-ICU applications three months after training.

ad a final test score higher than 90 % was approved. Those who had a final test score was ≤90 % were submitted to another round of training sections and another post-training test was applied until necessary score obtained. The instructors audited all in-training nurses CAM-ICU applications three months after training. Results: A total of 50 nurses participated in the training, with a mean graduation time of seven years and mean hospital admission time of five years. In the pre-training questionnaire, about 83,3 % of bedside critical care nurses correctly answered questions about CAM-ICU. However when bedside application was checked, around 80 % of the nurses applied the tool correctly. The main identified causes for this were high patient turnover, demanding families, and lack of practice in CAM-ICU. After theoretical and practical training sections, all of the bedside nurses correctly answered the post-training test. In the audit period (three months after training), around 97 % of the nurses correctly applied the CAM-ICU. Conclusions: An educational program enhances the correct application of CAM-ICU by bedside critical care nurses. References 1. Ely EW et al. JAMA 2001;286:2703–10. 2. Van Eijik MM et al. Am J Resp Crit Care Med 2011;184:340–44. Grant acknowledgement None.

Results: A total of 50 nurses participated in the training, with a mean graduation time of seven years and mean hospital admission time of five years. In the pre-training questionnaire, about 83,3 % of bedside critical care nurses correctly answered questions about CAM-ICU. However when bedside application was checked, around 80 % of the nurses applied the tool correctly. The main identified causes for this were high patient turnover, demanding families, and lack of practice in CAM-ICU. After theoretical and practical training sections, all of the bedside nurses correctly answered the post-training test. In the audit period (three months after training), around 97 % of the nurses correctly applied the CAM-ICU. Conclusions: An educational program enhances the correct application of CAM-ICU by bedside critical care nurses. References 1. Ely EW et al. JAMA 2001;286:2703–10. 2. Van Eijik MM et al. Am J Resp Crit Care Med 2011;184:340–44. Grant acknowledgement None. A812 Rocking chair mobilization therapy for critically ill patients in the intensive care unit G. Ebeling1, A. Kraegpoeth1, E. Laerkner2 1Odense University Hospital, Anesthesiology and Intensive Care Medicine, Odense, Denmark; 2Odense University Hospital, University of Southern Denmark, Odense, Denmark Correspondence: G. Ebeling – Odense University Hospital, Anesthesiology and Intensive Care Medicine, Odense, Denmark Introduction: In the Intensive Care Unit (ICU) several patients are disturbed in their cerebral function due to their critical illness and medication, leading to discomfort, agitation, restlessness, pain and delirium.

: G. Ebeling – Odense University Hospital, Anesthesiology and Intensive Care Medicine, Odense, Denmark Introduction: In the Intensive Care Unit (ICU) several patients are disturbed in their cerebral function due to their critical illness and medication, leading to discomfort, agitation, restlessness, pain and delirium. Rocking Chair Mobilization Therapy (RCMT) is a chair with good seating comfort which gives rhythmic movements. Rocking chair studies have shown concrete results to improve patient satisfaction, balance and well-being in patients who suffered from dementia (1). However, no studies have evaluated the value and the effect of RCMT for critically ill patients in the ICU. Objectives: The purpose of the study was to evaluate whether RCMT could be used in the rehabilitation of critically ill patients in the intensive care. The focus was to explore the impact of RCMT on critically ill patients comfort, pain, agitation and delirium. Methods: The evaluation took place in a medical/surgical ICU in Denmark in the period from May to July 2015. Patients ≥ 18 years, who were physically stable and had the ability to be mobilized to chair could participate in the evaluation. The RCMT session lasted 20 minutes.Each session with RCMT was evaluated by registration of patient consciousness (Richmond Agitation and Sedation Scale (RASS)), pain (numeric rating scale (NRS) 0–10 or by Critical-Care Pain Observation Tool (CPOT)), delirium (CAM-ICU) before and after the session. Patient comfort was assessed by the patients as well as by the nurses during the session.

ted by registration of patient consciousness (Richmond Agitation and Sedation Scale (RASS)), pain (numeric rating scale (NRS) 0–10 or by Critical-Care Pain Observation Tool (CPOT)), delirium (CAM-ICU) before and after the session. Patient comfort was assessed by the patients as well as by the nurses during the session. Results: 47 sessions with RCMT were evaluated. 24 males and 23 females, age between 49 and 88 years, participated in the evaluation. The results showed a decrease in patient agitation level and an increase in patient consiousness. Patients´ with RASS > 0 decreased from 18 before the session to 6 after the session. Patients with RASS ≤ −1 decreased from 5 before the session to 3 after the session. A decrease in delirium where 11 patients were assessed CAM-ICU positive before the session and 4 patients after the session. A decrease in pain where six patients scored NRS > 3 before the session compared to one patient after the session and 9 patients had CPOT scores > 2 before the session compared to 5 patients after the session. Assessment and evaluation of comfort by patients themselves and by the nurses, who cared for the particular patient, showed that RCMT was associated with a high degree of patient relaxation and comfort. Conclusions: Promising results gives reason to recommend RCMT for critically ill patients in the ICU, as an alternative holistic non-pharmacological intervention to stimulate patients´ bodily awareness and enhance patient comfort and rehabilitation. References:

Results: 47 sessions with RCMT were evaluated. 24 males and 23 females, age between 49 and 88 years, participated in the evaluation. The results showed a decrease in patient agitation level and an increase in patient consiousness. Patients´ with RASS > 0 decreased from 18 before the session to 6 after the session. Patients with RASS ≤ −1 decreased from 5 before the session to 3 after the session. A decrease in delirium where 11 patients were assessed CAM-ICU positive before the session and 4 patients after the session. A decrease in pain where six patients scored NRS > 3 before the session compared to one patient after the session and 9 patients had CPOT scores > 2 before the session compared to 5 patients after the session. Assessment and evaluation of comfort by patients themselves and by the nurses, who cared for the particular patient, showed that RCMT was associated with a high degree of patient relaxation and comfort. Conclusions: Promising results gives reason to recommend RCMT for critically ill patients in the ICU, as an alternative holistic non-pharmacological intervention to stimulate patients´ bodily awareness and enhance patient comfort and rehabilitation. References: 1. Watson, Nancy M.; Wells, Thelma J.; Cox Christopher (1998) Rocking chair therapy for dementia patients: Its effect on psychosocial well -being and balance. American Journal of Alzheimer´s Disease. pp. 296–308 Grant acknowledgement None

Conclusions: Promising results gives reason to recommend RCMT for critically ill patients in the ICU, as an alternative holistic non-pharmacological intervention to stimulate patients´ bodily awareness and enhance patient comfort and rehabilitation. References: 1. Watson, Nancy M.; Wells, Thelma J.; Cox Christopher (1998) Rocking chair therapy for dementia patients: Its effect on psychosocial well -being and balance. American Journal of Alzheimer´s Disease. pp. 296–308 Grant acknowledgement None A814 Nutritional intake and physical functioning after ICU discharge I. De Brito-Ashurst Royal Brompton and Harefield NHS Foundation Trust, Rehabilitation and Therapies, London, United Kingdom Introduction: Critical illness and immobility in the Intensive care unit (ICU) lead to a loss of muscle mass and reduced exercise capacity for many years following hospital discharge.[1] Nutritional management of the critically ill is challenging and most nutritional studies are focused in this period. Nutritional recommendations are for a high protein diet to minimise muscle breakdown and support protein synthesis during rehabilitation. Nevertheless, during the rehabilitation period little is known of patients' protein intake and physical functioning. Objective: To investigate physical functioning, frailty and dietary protein intake after 6 months of ICU discharge.

A814 Nutritional intake and physical functioning after ICU discharge I. De Brito-Ashurst Royal Brompton and Harefield NHS Foundation Trust, Rehabilitation and Therapies, London, United Kingdom Introduction: Critical illness and immobility in the Intensive care unit (ICU) lead to a loss of muscle mass and reduced exercise capacity for many years following hospital discharge.[1] Nutritional management of the critically ill is challenging and most nutritional studies are focused in this period. Nutritional recommendations are for a high protein diet to minimise muscle breakdown and support protein synthesis during rehabilitation. Nevertheless, during the rehabilitation period little is known of patients' protein intake and physical functioning. Objective: To investigate physical functioning, frailty and dietary protein intake after 6 months of ICU discharge. Method: Our ICU is recognised as a therapy rehabilitation centre and the only ICU member of the UK Rehabilitation Outcomes (UK-ROC). Patients cognitive and physical functioning is assessed as part of their rehabilitation therapy with the Functional Independence Measure (FIM) score[2]. The FIM contains 18 items on motor (13) and cognitive (5) functions that are scored on a 7-point ordinal scale based on the amount of assistance a person requires to perform specific activities. The FIM scores on ICU discharge and also on return to the rehabilitation clinic after 6 months were assessed. In addition, frailty was assessed based on a scale ranging from very fit to very severely frail, terminally ill[3] and patients were asked to complete a protein food frequency questionnaire.

pecific activities. The FIM scores on ICU discharge and also on return to the rehabilitation clinic after 6 months were assessed. In addition, frailty was assessed based on a scale ranging from very fit to very severely frail, terminally ill[3] and patients were asked to complete a protein food frequency questionnaire. Results: Twenty patients were assessed. Data are reported as mean and (standard deviation). Patients were male 66 % and 53.4 years (33.5). Paired t tests of the changes in FIM scores from discharge showed significant increments; 9.7 ± 11.4 (P = 0.05) and 24.7 ± 23.4 (P < 0.04) for motor and cognitive scales respectively. Nevertheless, patients reported that they were “vulnerable to moderately frail” in the frailty scale. Dietary intake was also inadequate with a protein intake of 0.83 g/kg (1.15). Conclusion: There was improvement in FIM score after discharge but that was mainly from cognitive function. A lower improvement was observed in motor functioning supporting the vulnerable to moderately frail scale and a reduced protein intake. References [1] Herridge, M.S., et al. 2011. Functional disability 5 years after acute respiratory distress syndrome. N.Engl.J.Med., 364(14) 1293–1304. [2] Granger CV, Hamilton BB, Linacre JM, et al. Performance profiles of the functional independence measure. Am J Phys Med Rehabil. 1993; 72:84–89. [3] McDermid RC, Stelfox HT, Bagshaw SM. Frailty in the critically ill: a novel concept. Critical Care. 2011; 15:301.

[1] Herridge, M.S., et al. 2011. Functional disability 5 years after acute respiratory distress syndrome. N.Engl.J.Med., 364(14) 1293–1304. [2] Granger CV, Hamilton BB, Linacre JM, et al. Performance profiles of the functional independence measure. Am J Phys Med Rehabil. 1993; 72:84–89. [3] McDermid RC, Stelfox HT, Bagshaw SM. Frailty in the critically ill: a novel concept. Critical Care. 2011; 15:301. A815 ICU rehabilitation treatment time: outcomes and barriers C. White1, S. Gregory2, L.G. Forni1 1Royal Surrey County Hospital, ICU and SPACeR research group, Guildford, United Kingdom; 2Royal Surrey County Hospital, Guildford, United Kingdom Correspondence: S. Gregory – Royal Surrey County Hospital, Guildford, United Kingdom Introduction: The Intensive Care Society in the UK has published core standards for the provision of physiotherapy and rehabilitation in ICU patients1. These state:'Patients receiving rehabilitation should be offered a minimum of 45 minutes of each active therapy that is required, for a minimum of 5 days a week of therapy per day, at a level which enables them to meet their goals for as long as they are continuing to benefit from the therapy and are able to tolerate it.'However these standards were adapted from the NICE Quality guidelines for stroke 20102 which targets 45 minutes per day over 5 days or 32 minutes per day over 7 days, rather than the direct needs of patients with Intensive Care Unit Acquired Weakness (ICUAW).

ing to benefit from the therapy and are able to tolerate it.'However these standards were adapted from the NICE Quality guidelines for stroke 20102 which targets 45 minutes per day over 5 days or 32 minutes per day over 7 days, rather than the direct needs of patients with Intensive Care Unit Acquired Weakness (ICUAW). Objectives: We have examined our current rehabilitation service against the recommended 32 minute target for active therapy in the critical care unit to see if we achieve these core standards. Methods: We recorded the therapy time provided per day to our emergency critical care patients on a documented Critical Illness Rehabilitation Pathway. Respiratory physiotherapy was excluded in the treatment times. We excluded patients on post-surgical Enhanced Recovery Pathways (ERP's). The study time run over 4 weeks with prospective data collection. Results: A total of 53 patients were included with 417 physiotherapy contacts over the study period. Patients are offered two treatment sessions per day Monday to Friday and once at a weekend, standard practice in our critical care unit. Patients were described as one of 3 categories: self-ventilating, ventilated via tracheostomy and ventilated via Endotracheal Tube (ETT).

physiotherapy contacts over the study period. Patients are offered two treatment sessions per day Monday to Friday and once at a weekend, standard practice in our critical care unit. Patients were described as one of 3 categories: self-ventilating, ventilated via tracheostomy and ventilated via Endotracheal Tube (ETT). Conclusions: These results demonstrate that the major limiting factors in achieving the core standards are principally driven by factors unrelated to physiotherapy provision. In particular the most significant limiting factor of the self-ventilating and tracheostomy groups was fatigue/exercise tolerance, followed by medical limitation although self-ventilating patients are more likely to refuse treatment. The ventilated via ETT group, although receiving significant respiratory physiotherapy input receive less rehabilitation due to sedation. These data could be used to help inform any potential ESICM guidelines for ICUAW therapy. References 1. The Intensive Care Society. Core standards for Intensive care units 2013. Available at: https://www.ficm.ac.uk/sites/default/files/Core%20Standards%20for%20ICUs%20Ed.1%20(2013).pdf 2. NICE quality standard (QS2) Stroke in Adults 2010. Available at: https://www.nice.org.uk/guidance/qs2 3. The Sentinel Stroke National Audit Programme (SSNAP). Available at: https://www.rcplondon.ac.uk/projects/outputs/sentinel-stroke-national-audit-programme-ssnapTable 4 (abstract A815). Treatment received (%)

1. The Intensive Care Society. Core standards for Intensive care units 2013. Available at: https://www.ficm.ac.uk/sites/default/files/Core%20Standards%20for%20ICUs%20Ed.1%20(2013).pdf 2. NICE quality standard (QS2) Stroke in Adults 2010. Available at: https://www.nice.org.uk/guidance/qs2 3. The Sentinel Stroke National Audit Programme (SSNAP). Available at: https://www.rcplondon.ac.uk/projects/outputs/sentinel-stroke-national-audit-programme-ssnapTable 4 (abstract A815). Treatment received (%) Treatment Offered Self-Ventilating Ventilated via Tracheostomy Ventilated via ETT Bed Exercises 7.2 19.1 0 Chair Exercises 4 1.1 0 Stair practice 0.8 0 0 Balance and standing exercises 10.4 12.4 0 ROM/stretches 4.8 23.5 79 Mobility and gait re-education 45 41.5 0 Sit on edge of balance/sitting balance 7.2 10.1 0 Transfer practice 26.4 17.9 0 Positioning/UL exercises 8 11.1 31 Table 5 (abstract A815). Limitations to Rehab (%) Limitations to treatment Self-ventilating Ventilated via Tracheostomy Ventilated via ETT Fatigue and Exercise tolerance 41.6 39.3 3.57 Refusal/Confusion/Anxiety 24.8 5.6 0 Pain 14.4 1.1 7.14 Procedure/Unavailable 4 8.9 10.71 Acutely unwell/CVS/medical limitations/CVVHD 29.6 20.2 71.4 Sedated/Decreased GCS 1.6 12.4 7.14 Respiratory wean 0 5.6 0 Lack of equipment/Time 1.6 3.3 3.57 Other 16.8 6.7 0 Fig. 5 (abstract A815). Results

tolerance 41.6 39.3 3.57 Refusal/Confusion/Anxiety 24.8 5.6 0 Pain 14.4 1.1 7.14 Procedure/Unavailable 4 8.9 10.71 Acutely unwell/CVS/medical limitations/CVVHD 29.6 20.2 71.4 Sedated/Decreased GCS 1.6 12.4 7.14 Respiratory wean 0 5.6 0 Lack of equipment/Time 1.6 3.3 3.57 Other 16.8 6.7 0 Fig. 5 (abstract A815). Results A816 Exploring current rehabilitation on ITU: can we measure 'tolerance' and 'level' of rehabilitative physiotherapy? E. Flowers, A. Curtis, C.-A. Wood Guys and St Thomas NHS Foundation Trust, London, United Kingdom Correspondence: E. Flowers – Guys and St Thomas NHS Foundation Trust, London, United Kingdom Introduction: Rehabilitation in the Intensive Care Unit (ICU) aims to enhance health, wellbeing and recovery beyond survival of critical illness. Current rehabilitation practice requires description and measurement of effect to enhance exercise prescription1. Objectives: Physiotherapy rehabilitation is a recognised component of ICU care. The Intensive Care Society - Core Standards2 recommends that rehabilitation is 'at a level that enables the patient to meet their rehabilitation goals for as long as they…are able to tolerate it'. In order to investigate and measure the terms 'tolerate' and 'level', physiological measurements and their relationship with self-perceived exertion and tolerance were analysed.

s that rehabilitation is 'at a level that enables the patient to meet their rehabilitation goals for as long as they…are able to tolerate it'. In order to investigate and measure the terms 'tolerate' and 'level', physiological measurements and their relationship with self-perceived exertion and tolerance were analysed. Methods: The project was registered with Guy's & St.Thomas' NHS Foundation Trust, Clinical Audit Group, (Project No. 4701). A convenience sample, of ICU patients undergoing active physiotherapy led rehabilitation, were observed between July and September 2014. A Modified Exertion Scale was used to measure patients' perceived effort. Patients also rated tolerance of the session using a Tolerability Scale, created based on the exertion scale. Sessions were timed, heart rate, blood pressure and oxygen saturation were monitored and the cardiovascular impact of the session measured using Heart Rate Reserve (HRR). Results: Nine rehabilitation sessions were observed; mean length of 17 minutes (range 9–28). Minimum target HRR (>30 %) was achieved, but not sustained, by 3 patients, while 1 peaked within a normal target HRR (40-85 %). Of the 9 patients, 6 were able to use the tolerability scale and 7 the exertion scale. There did not appear to be a relationship between HRR and either perceived scale measurements. There did appear to be a link between perceived exertion and perceived tolerability with 5 of the 6 patients scoring within 3 points.

AH1 N1 influenza virus in critical care patients has not been well estimated, México is considered as the origin of this pandemic disease, and some reports of the CDC (Centers for Disease Control and Prevention) suggest the presentation in this country is usually more severe than in others, with a higher risk of death. Objectives: The objective of this study was to determine the clinical and ventilatory differences betwen patients survivors and non survivors with diagnostic of moderate and severe ARDS due to confirmed 2009 AH1 N1 influenza virus, hospitalized in ICU at a mexican tertiary center. Methods: Observational, retrospective study including adult patients with moderate and severe ARDS, due to RT-PCR confirmed 2009 AH1 N1 influenza virus, hospitalized in the ICU during march 2009 to march 2016. Patients were analyzed studied in two groups:survivors and non survivors.

%). Of the 9 patients, 6 were able to use the tolerability scale and 7 the exertion scale. There did not appear to be a relationship between HRR and either perceived scale measurements. There did appear to be a link between perceived exertion and perceived tolerability with 5 of the 6 patients scoring within 3 points. Conclusions: Reported perception of exertion and physiological markers could both indicate the 'level' patients are working at. We were able to measure effects of rehabilitation on heart rate. The majority of patients were able to use exertion and tolerance scales. However, the change in heart rate was not great enough to suggest a training effect, despite their exertion scores implying high effort levels. To fulfil the ICU society recommendations, a good understanding is needed of how hard patients are working during rehabilitation. Further research is needed to determine why there may be disparity between heart rate and patient-reported measures of exertion; and if either is a useful guide for exercise prescription with ICU patients. References 1. Calvo-Ayala, E., Khan, B., O. Farber, M,. Ely, E., Boustani, M., (2013), 'Interventions to Improve the Physical Function of ICU Survivors', Chest, issue 144, no. 5, pp 1469–1480. 2. Core Standards for Intensive Care Units. Available online at [http://www.ficm.ac.uk/sites/default/files/CoreStandardsforICUsEd.1(2013).pdf] accessed on 23/12/2015

1. Calvo-Ayala, E., Khan, B., O. Farber, M,. Ely, E., Boustani, M., (2013), 'Interventions to Improve the Physical Function of ICU Survivors', Chest, issue 144, no. 5, pp 1469–1480. 2. Core Standards for Intensive Care Units. Available online at [http://www.ficm.ac.uk/sites/default/files/CoreStandardsforICUsEd.1(2013).pdf] accessed on 23/12/2015 A817 Optimising mobilisation in the critically ill - translating knowledge into clinical practice K. Siu1, K. Venkatesan2, J.B.H. Muhammad1, L. Ng1, E. Seet2 1Khoo Teck Puat Hospital, Rehabilitation Services, Singapore, Singapore; 2Khoo Teck Puat Hospital, Anaesthesia & Surgical Intensive Care, Singapore, Singapore Correspondence: K. Siu – Khoo Teck Puat Hospital, Rehabilitation Services, Singapore, Singapore Introduction: Critically ill patients are at risk of developing deconditioning, muscle atrophy and functional impairments long after hospital discharge. There is evidence demonstrating benefits of mobilization in critically ill patients - improved functional outcome and reduced ICU and hospital length of stay. However, there is limited information about how these advances are translated to clinical practice. Objectives: To obtain a baseline data on patients who are eligible for mobilisation in ICU and how many of these patients are optimally mobilised in ICU. This would enable us to undertake a clinical practice improvement project (CPIP) using the Plan-Do-Study-Act (PDSA) implementation strategy to optimise mobilisation in at least 85 % of all eligible ICU patients.

ents who are eligible for mobilisation in ICU and how many of these patients are optimally mobilised in ICU. This would enable us to undertake a clinical practice improvement project (CPIP) using the Plan-Do-Study-Act (PDSA) implementation strategy to optimise mobilisation in at least 85 % of all eligible ICU patients. Methods: Setting. 14-bedded intensivist led closed surgical ICU. The mobilisation team composed of physiotherapists, bedside nurses and respiratory therapists who worked along with an intensivist. Prospective audit conducted to collect data on the patients who met the eligibility criteria of mobilisation over a 3-month period. CPIP team Results: Our audit revealed that at baseline, only 44 % of all eligible patients were optimally mobilised. RCA revealed a total of 21 barriers and through multi-voting and pareto-charting, we identified the top 3 barriers to change. Key barriers identified were: 1. Mobility not being a part of the daily review routine 2. Staff were unsure of the eligibility criteria 3. Lack of knowledge the benefits of optimal mobilisation in the critically ill. The team proposed following strategies to overcome the barriers: 1. Combined ICU multi-disciplinary handover rounds with the lead consultant asking the question “Can this patient be mobilised?” for every patient reviewed. 2. Providing a bedside decision-making algorithm on eligibility criteria, displayed within visibility of staff's work area. 3. Undertake sharing session with ground staff on the importance and benefits of optimising mobility of the critically ill.

he question “Can this patient be mobilised?” for every patient reviewed. 2. Providing a bedside decision-making algorithm on eligibility criteria, displayed within visibility of staff's work area. 3. Undertake sharing session with ground staff on the importance and benefits of optimising mobility of the critically ill. Conclusions: Our audit revealed that less than half of eligible patients received early mobilisation. Our CPIP - a quality improvement initiative identified barriers in translating knowledge into clinical practice. Through various tools of CPIP, we identified the key barriers and strategies to overcome these barriers and thereby achieving the goal of optimising mobilisation in ICU patients. References 1. Engel, Heidi J et al. (2013) ICU Early Mobilization: From Recommendation to Implementation at Three Medical Centres. Crit Care Med 41(9) S69-S80Fig. 6 (abstract A817). Comprised of multi-disciplinary stakeholders - frontline staff and patient representative. A brainstorming storming process identified various barriers to optimal mobilisation. Quality improvement tools such as Affinity diagram Fig. 7 (abstract A817). Root cause analysis (RCA) Fig. 8 (abstract A817). Run charts and Pareto charts were then utilized to form the prioritization matrices

1. Engel, Heidi J et al. (2013) ICU Early Mobilization: From Recommendation to Implementation at Three Medical Centres. Crit Care Med 41(9) S69-S80Fig. 6 (abstract A817). Comprised of multi-disciplinary stakeholders - frontline staff and patient representative. A brainstorming storming process identified various barriers to optimal mobilisation. Quality improvement tools such as Affinity diagram Fig. 7 (abstract A817). Root cause analysis (RCA) Fig. 8 (abstract A817). Run charts and Pareto charts were then utilized to form the prioritization matrices A818 Impact of physiotherapy on the quality of assistance offered by an Angolan ICU N. Baptista1, A. Escoval2, E. Tomas1 1Clinica Sagrada Esperança, Luanda, Angola; 2Escola Nacional de Saude Publica/UNL, Lisboa, Portugal Correspondence: E. Tomas – Clinica Sagrada Esperança, Luanda, Angola Introduction: Bed rest and immobility during critical illness may result in profound physical deconditioning. The multidisciplinary team in intensive care includes physiotherapists, who are responsible for performing diagnoses and procedures for critically ill patients, such as ventilation, respiratory monitoring and assessments of musculoskeletal, neurological, metabolic and cardiovascular diseases, and for the prevention and treatment of the effects of prolonged immobility. Objectives: To evaluate the influence of physiotherapy on quality indicators in the intensive care unit of the Sagrada Esperança Clinic in Luanda, Angola.

A818 Impact of physiotherapy on the quality of assistance offered by an Angolan ICU N. Baptista1, A. Escoval2, E. Tomas1 1Clinica Sagrada Esperança, Luanda, Angola; 2Escola Nacional de Saude Publica/UNL, Lisboa, Portugal Correspondence: E. Tomas – Clinica Sagrada Esperança, Luanda, Angola Introduction: Bed rest and immobility during critical illness may result in profound physical deconditioning. The multidisciplinary team in intensive care includes physiotherapists, who are responsible for performing diagnoses and procedures for critically ill patients, such as ventilation, respiratory monitoring and assessments of musculoskeletal, neurological, metabolic and cardiovascular diseases, and for the prevention and treatment of the effects of prolonged immobility. Objectives: To evaluate the influence of physiotherapy on quality indicators in the intensive care unit of the Sagrada Esperança Clinic in Luanda, Angola. Methods: A retrospective before-after study was designed to assess some quality indicators within the intensive care unit between July and September 2013, where there were no physiotherapists specially trained for respiratory care, and from January to March 2014, where the physiotherapists integrated a multidisciplinary team. The quality indicators analyzed were: the average duration of mechanical ventilation, prevalence of ventilator associated pneumonia and the rate of ventilated patients with non-invasive ventilation. The study population comprised 62 patients for 2013 and 71 for 2014. In this study the patients´ categorization was made by age, sex, pathology and also according with the patient classification systems SAPS 3 and SOFA. The statistical analysis used the systems SPSS version 22 for a 5 % significance level.

ventilation. The study population comprised 62 patients for 2013 and 71 for 2014. In this study the patients´ categorization was made by age, sex, pathology and also according with the patient classification systems SAPS 3 and SOFA. The statistical analysis used the systems SPSS version 22 for a 5 % significance level. Results: The results obtained after analyzing the two homogeneous groups according to age, gender, type of admission and severity influencing the physiotherapy care in ICU quality indicators, in the Sagrada Esperança clinic, highlights the decrease of the average number of days with mechanical ventilation but it is not observed a significant relation between physical therapy and this indicator (p = 0:06). Furthermore, it is also observed a decrease ventilator associated pneumonia, and a significant relation between this indicator and the respiratory physiotherapy. Last, there is a strong relation between the increase on the number of patients without invasive ventilation and physiotherapy (p = 0.017). Conclusions: In this study it is demonstrated the respiratory therapy influences in some quality indicators, namely regarding the reduction of ventilation associated pneumonia and the promotion of non-invasive ventilation in the ICU of the CSE. References 1) Pinto, W. A. M., Rossetti, H. B., Araújo, A., Spósito Júnior, J. J., Salomão, H., Mattos, S. S., Machado, F. R. (2014). Revista Brasileira de Terapia Intensiva, 26(1), 7–13.Fig. 9 (abstract A818). Baseline characteristics Fig. 10 (abstract A818). Outcome

Conclusions: In this study it is demonstrated the respiratory therapy influences in some quality indicators, namely regarding the reduction of ventilation associated pneumonia and the promotion of non-invasive ventilation in the ICU of the CSE. References 1) Pinto, W. A. M., Rossetti, H. B., Araújo, A., Spósito Júnior, J. J., Salomão, H., Mattos, S. S., Machado, F. R. (2014). Revista Brasileira de Terapia Intensiva, 26(1), 7–13.Fig. 9 (abstract A818). Baseline characteristics Fig. 10 (abstract A818). Outcome FUNGAL INFECTION AND INFECTION PREVENTION A820 Epidemiology, treatment pattern and outcomes of candida infection in non neutropenic patients in a medical icu of a developing country R. Agrawal, R. Mathew, A. Varma Fortis Escorts Heart Institute, Critical Care Medicine, New Delhi, India Correspondence: R. Agrawal – Fortis Escorts Heart Institute, Critical Care Medicine, New Delhi, India Introduction: The incidence of fungal infections has increased in Indian ICU`s over last few years1. The mortality due to unrecognized and untreated fungal infections is significantly higher. Objectives: To study the incidence and epidemiology of fungal infections in a medical ICU in non neutropenic patients. We also studied the treatment pattern and patient outcomes in these critically ill patients.

FUNGAL INFECTION AND INFECTION PREVENTION A820 Epidemiology, treatment pattern and outcomes of candida infection in non neutropenic patients in a medical icu of a developing country R. Agrawal, R. Mathew, A. Varma Fortis Escorts Heart Institute, Critical Care Medicine, New Delhi, India Correspondence: R. Agrawal – Fortis Escorts Heart Institute, Critical Care Medicine, New Delhi, India Introduction: The incidence of fungal infections has increased in Indian ICU`s over last few years1. The mortality due to unrecognized and untreated fungal infections is significantly higher. Objectives: To study the incidence and epidemiology of fungal infections in a medical ICU in non neutropenic patients. We also studied the treatment pattern and patient outcomes in these critically ill patients. Methods: The study was conducted from January 2015 to March 2016 over a period of fifteen months.Total patients admitted to the medical ICU were screened and patients with suspected candida infection were noted. The APACHE II score was calculated for assessing the severity of illness.Patients receiving emperical antifungal therapy and type of antifungal drugs used were also noted. The mortality rate in patients with cultures positive for candida was noted and compared to ICU mortality rate.

pected candida infection were noted. The APACHE II score was calculated for assessing the severity of illness.Patients receiving emperical antifungal therapy and type of antifungal drugs used were also noted. The mortality rate in patients with cultures positive for candida was noted and compared to ICU mortality rate. Results: There were total of 3700 patients who were admitted in the 50 bed medical ICU.Out of these 125 (3.3 %) patients were culture positive for candida. Candida tropicalis was the most common species followed by C. albicans (Table 6). The mean APACHE II score in patients with culture positive with candida was 19.2 whereas it was 17.5 in other patients (9.7 % higher in Candida group). Azoles were the most common antifungal used (58 %) followed by echinocandins (21.4 %). The crude mortality rate in patients with positive cultures for candida was 26.4 % (n = 33) which was significantly higher than ICU mortality rate of 5.5 %. (n = 197), (pvalue < 0.001). Conclusions: Candida infections are a significant cause of mortality in ICU`s .C. tropicalis is more commonly isolated than C albicans. Azoles are the most common antifungals used though the usage of ecinocandins is rising. Early recognition and treatment of fungal infections is of critical importance. Avoiding overuse and promoting deescalation protocol of antibiotics can help in controlling fungal infections in the ICU. References 1. Chakrabarti A. J Postgrad Med. 2005; 51(1): 16–20. 5. Dellinger et al. CCM. 2013;41:580–637

Conclusions: Candida infections are a significant cause of mortality in ICU`s .C. tropicalis is more commonly isolated than C albicans. Azoles are the most common antifungals used though the usage of ecinocandins is rising. Early recognition and treatment of fungal infections is of critical importance. Avoiding overuse and promoting deescalation protocol of antibiotics can help in controlling fungal infections in the ICU. References 1. Chakrabarti A. J Postgrad Med. 2005; 51(1): 16–20. 5. Dellinger et al. CCM. 2013;41:580–637 2. Pappas P, Et al. A Prospective Observational Study of Candidemia: Epidemiology, Therapy, and Influences on Mortality in Hospitalized Patients. Clin Infect Dis (2003) 37(5): 634–643. Grant acknowledgement None.Table 6 (abstract A820). Candida species isolated FAMATA 2 1.9 % GLABRATA 14 13.4 % GUILLIERMONDII 1 1.0 % HAEMULONII 2 1.9 % KRUSEI 1 1.0 % PARAPSILOSIS 3 2.8 % RUGOSA 2 1.9 % TROPICALIS 45 42.3 % ALBICANS 37 35.2 %

2. Pappas P, Et al. A Prospective Observational Study of Candidemia: Epidemiology, Therapy, and Influences on Mortality in Hospitalized Patients. Clin Infect Dis (2003) 37(5): 634–643. Grant acknowledgement None.Table 6 (abstract A820). Candida species isolated FAMATA 2 1.9 % GLABRATA 14 13.4 % GUILLIERMONDII 1 1.0 % HAEMULONII 2 1.9 % KRUSEI 1 1.0 % PARAPSILOSIS 3 2.8 % RUGOSA 2 1.9 % TROPICALIS 45 42.3 % ALBICANS 37 35.2 % A821 Candidemia in intensive care unit (ICU) patients: risk factors for non-albicans Candida species and for fluconazole resistance E. Dima1, E. Charitidou2, E. Perivolioti3, M. Pratikaki3, C. Vrettou3, A. Giannopoulos3, S. Zakynthinos3, C. Routsi3 1University of Athens, Medical School, Evangelismos Hospital, Athens, Greece; 2National Technical University of Athens, Athens, Greece; 3Evangelismos Hospital, Athens, Greece Correspondence: E. Dima – University of Athens, Medical School, Evangelismos Hospital, Athens, Greece Introduction: The incidence of candidemia has increased in ICU patients (1). In addition, there are differences in epidemiology among different countries. We have previously shown an increased proportion of non-albicans Candida species in our ICU (2). Objectives: To identify the variables associated with candidemia due to non- albicans Candida species, as well as with fluconazole-resistant strains in a multidisciplinary ICU.

A821 Candidemia in intensive care unit (ICU) patients: risk factors for non-albicans Candida species and for fluconazole resistance E. Dima1, E. Charitidou2, E. Perivolioti3, M. Pratikaki3, C. Vrettou3, A. Giannopoulos3, S. Zakynthinos3, C. Routsi3 1University of Athens, Medical School, Evangelismos Hospital, Athens, Greece; 2National Technical University of Athens, Athens, Greece; 3Evangelismos Hospital, Athens, Greece Correspondence: E. Dima – University of Athens, Medical School, Evangelismos Hospital, Athens, Greece Introduction: The incidence of candidemia has increased in ICU patients (1). In addition, there are differences in epidemiology among different countries. We have previously shown an increased proportion of non-albicans Candida species in our ICU (2). Objectives: To identify the variables associated with candidemia due to non- albicans Candida species, as well as with fluconazole-resistant strains in a multidisciplinary ICU. Methods: All ICU patients with candidemia were prospectively studied over two time periods (2005–2008 and 2012–2015). Demographics, illness severity, clinical and laboratory variables were recorded. SOFA score value on ICU admission subtracted from the value on the day of candidemia occurrence was defined as Delta SOFA. Patients with C. albicans candidemia were compared to those with non-albicans candidemia. Also, patients with fluconazole-resistant candidemia were compared to those without fluconazole resistance.

SOFA score value on ICU admission subtracted from the value on the day of candidemia occurrence was defined as Delta SOFA. Patients with C. albicans candidemia were compared to those with non-albicans candidemia. Also, patients with fluconazole-resistant candidemia were compared to those without fluconazole resistance. Results: Among 143 patients with ICU-acquired candidemia, in 55 patients candidemia was due to C. albicans and in 88 patients to non-albicans species. C. parapsilosis was the most common (46 %) followed by C. albicans (38 %). The median time from ICU admission to candidemia onset was 12 and 23 days for C. albicans and non-albicans respectively, p = 0.02. Similarly, the median time for candidemia due to fluconazole sensitive isolate was 14 days and 32 days for fluconazole resistant, p < 0.001. Resistance to fluconazole was 9 % and 51 % in C. albicans and in non-albicans species respectively, p < 0.001).Presence of shock on candidemia day (OR 6.85; CI: 2.2-25, p = 0.001) and the Delta SOFA score (OR 0.74; CI: 0.60-0.89, p = 0.002) were independently associated with candidemia due to C. albicans. Independent risk factors for fluconazole resistant isolates were the length of ICU stay before the development of candidemia (OR 1.03; CI: 1.01-1.05, p = 0.003) and the presence of shock on candidemia day (OR 0.23; CI: 0.07-0.64, p = 0.006). Previous fluconazole exposure (10 patients) was not associated with fluconazole resistance.

isk factors for fluconazole resistant isolates were the length of ICU stay before the development of candidemia (OR 1.03; CI: 1.01-1.05, p = 0.003) and the presence of shock on candidemia day (OR 0.23; CI: 0.07-0.64, p = 0.006). Previous fluconazole exposure (10 patients) was not associated with fluconazole resistance. Conclusions: This study confirms the predominance of non-albicans Candida species, in our ICU patients with candidemia, with high prevalence of fluconazole resistance. Early onset of candidemia and the presence of shock were most likely due to C. albicans whereas late onset was associated with fluconazole-resistant non- albicans species. These findings may be of value for empiric antifungal treatment selection. References (1) Intensive Care Med 2014;40:1303 (2) Mycoses 2011;54:154 A822 Early fungal infections after lung transplantation E. Atchade, S. Houzé, S. Jean-Baptiste, G. Thabut, C. Genève, S. Tanaka, B. Lortat-Jacob, P. Augustin, M. Desmard, P. Montravers Hopital Bichat Claude Bernard, Paris, France Correspondence: E. Atchade – Hopital Bichat Claude Bernard, Paris, France Introduction: Fungal infections (FI) after lung transplantation (LT) are common and associated with high mortality and morbidity rates (1). Published studies report late invasive infections caused by Aspergillus sp, but the early post-transplant period in intensive care unit (ICU) patients is rarely assessed.

s, France Introduction: Fungal infections (FI) after lung transplantation (LT) are common and associated with high mortality and morbidity rates (1). Published studies report late invasive infections caused by Aspergillus sp, but the early post-transplant period in intensive care unit (ICU) patients is rarely assessed. Objectives: The primary goal of the study was to describe the epidemiology of early FI in ICU after LT. Secondary aims were to evaluate its impact on the outcome on ICU stay and to determine the risk factors for fungal colonisation. Methods: This observational, retrospective, monocentric study analysed microbiologic results, clinical evolution and outcome of 176 LT, in ICU, during a 6-year period. Fungal positive respiratory sample was considered as colonisation when no clinical, radiological or histological criteria for invasive infection were present. Results are expressed as median and interquartile range. Statistical analyses were performed using Chi square, Mann-Whitney and Kruskal Wallis tests. The level of statistical significance was set at 5 %. The local Ethic Committee approval was obtained for the study.

logical criteria for invasive infection were present. Results are expressed as median and interquartile range. Statistical analyses were performed using Chi square, Mann-Whitney and Kruskal Wallis tests. The level of statistical significance was set at 5 %. The local Ethic Committee approval was obtained for the study. Results: During the pre-transplantation period, Candida sp colonisation was reported in 17 % of the patients (87 % C. albicans), while Aspergillus sp colonisation was observed in 4 % of them. In the post-transplantation period, 69 % of patients were colonised with fungi, mainly C.albicans (33 % of cases), rarely Aspergillus spp (7 %). Median time to onset of Candida colonisation was 4 days [1–8] and 5 days [3–14] for Aspergillus colonisation. Fungi were significantly associated with the presence of Enterobacteriaceae (OR = 2.61, 95%CI [1.36-5]; p = 0.003) and enterococci (OR = 5.09, 95%CI [1.14, 22.75]; p = 0.02). Risk factors for fungal post-operative colonisation were bi-pulmonary transplantation (OR = 2.49, 95%CI [1.15-5.45], p = 0.02) and COPD (OR = 2.71, 95%CI [1.39-5.32], p = 0.003). Among the patients developing candida colonisation in the post-transplantation period, 57 % received an antifungal treatment during their ICU stay (10 % echinocandins, 97 % azoles). A significant association between pre-transplant Candida sp colonisation and mortality rate in ICU was observed (OR = 2.35, 95%CI [1.14-4.9], p = 0.03). Post-operative fungal colonisation was not associated with increased death rate in ICU (OR = 1.32, 95%CI [0.62, 2.82], p = 0.57) or duration of mechanical ventilation ≥3 days (OR = 1.65, 95%CI [0.87-3.12]) while post-operative Candida sp colonisation was associated with prolonged ICU stay (p < 0.001) and increased duration of mechanical ventilation (p = 0.002).

associated with increased death rate in ICU (OR = 1.32, 95%CI [0.62, 2.82], p = 0.57) or duration of mechanical ventilation ≥3 days (OR = 1.65, 95%CI [0.87-3.12]) while post-operative Candida sp colonisation was associated with prolonged ICU stay (p < 0.001) and increased duration of mechanical ventilation (p = 0.002). Conclusions: Prevalence of fungal colonisation is high after LT, most commonly caused by Candida sp. The lack of association between post-transplant fungal colonisation and mortality and morbidity suggests to avoid antifungal therapy when no clinical signs of fungal infection are observed. References (1) Solé A, Transplant Rev Orlando Fla, 2008; 22(2): 89–104 Grant acknowledgement: None.

Conclusions: Prevalence of fungal colonisation is high after LT, most commonly caused by Candida sp. The lack of association between post-transplant fungal colonisation and mortality and morbidity suggests to avoid antifungal therapy when no clinical signs of fungal infection are observed. References (1) Solé A, Transplant Rev Orlando Fla, 2008; 22(2): 89–104 Grant acknowledgement: None. A823 Clinical significance of Aspergillus isolation in critically ill H1 N1 patients F.J. González de Molina1,2, S. Barbadillo3,4, R. Alejandro5, F. Álvarez-Lerma6, J. Vallés7, R.M. Catalán8, E. Palencia9, A. Jareño10, R.M. Granada11, M.-L. Ignacio12, GETGAG Working Group 1Hospital Universitari Mútua de Terrassa, Intensive Care Department, Terrassa, Spain; 2AGAUR, Grup Recerca Emergent, Terrassa, Spain; 3Hospital General de Catalunya, Intensive Care Department, Sant Cugat del Vallés, Spain; 4Universidad Autónoma de Barcelona, Departament de Medicina, Barcelona, Spain; 5Hospital Universitari de Tarragona Joan XXIII, Intensive Care Department, Tarragona, Spain; 6Hospital del Mar, Intensive Care Department, Barcelona, Spain; 7Hospital Parc Taulí, Intensive Care Department, Sabadell, Spain; 8Hospital General de Vic, Intensive Care Department, Vic, Spain; 9Hospital Infanta Leonor, Intensive Care Department, Madrid, Spain; 10Hospital del SAS de Jerez, Intensive Care Department, Jerez de la Frontera, Spain; 11Hospital de Bellvitge, Intensive Care Department, Barcelona, Spain; 12St James's University Hospital, Dublin, Ireland Correspondence: F.J. González de Molina – Hospital Universitari Mútua de Terrassa, Intensive Care Department, Terrassa, Spain Introduction: Invasive Aspergillus infections are well-known complications of immunocompromised states, chronic obstructive pulmonary disease and haematopoietic stem cell transplant. Bacterial coinfection is well described in influenza literature but there is scarce data on invasive aspergillosis complicated severe influenza infection.

asive Aspergillus infections are well-known complications of immunocompromised states, chronic obstructive pulmonary disease and haematopoietic stem cell transplant. Bacterial coinfection is well described in influenza literature but there is scarce data on invasive aspergillosis complicated severe influenza infection. Objectives: The aim of this study is to describe the clinical and demographic characteristics of patients with Aspergillus isolation in severe influenza A(H1 N1) pneumonia. Methods: Prospective, observational, multicenter study conducted in 148 Spanish ICUs from 2009 to 2015. All individuals with severe primary influenza A(H1 N1) pneumonia requiring invasive mechanical ventilation were included in the study. Influenza A(H1 N1) patients without coinfection were compared with those with Aspergillus isolation in respiratory samples. 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 counts (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.

outcomes were recorded. Discrete variables are expressed as counts (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 1594 intubated patients with confirmed influenza A (H1 N1) pneumonia at ICU admission, 385 were excluded due to other microorganism coinfection. At all, 1185 patients with H1 N1 pneumonia were compared to 24 patients with H1 N1 pneumonia and Aspegillus isolation (AI) in respiratory samples. Patients with AI were older (64 [54–71] vs 49 [38–60], P < 0.001), presented a higher proportion of COPD (39.1 % vs 17.8 %, P =0.024), chronic renal failure (21.7 % vs 7.1 %, P = < 0.023), and immunodeficiency (34.8 % vs 10.8 %, P =0.002). Patients with AI developed more acute kidney injury (47.6 % vs 28.0 %, P =0.048) and were treated more frecuently with corticosteroids (71.4 % vs 47.1 %, P = < 0.044). Overall mortality was much higher in those patients with AI (65.2 % vs 29.6 %, P < 0.001). Conclusions: The mortality rate was significantly higher in H1 N1 patients with Aspergillus isolation in respiratory samples. Diagnosis of invasive aspergillosis in critically ill patients in the post-influenza era must be re-evaluated. Clinical studies should be conducted in order to know the clinical significance of Aspergillus isolation in respiratory samples in intubated patients with primary influenza A(H1 N1) pneumonia. References

Conclusions: The mortality rate was significantly higher in H1 N1 patients with Aspergillus isolation in respiratory samples. Diagnosis of invasive aspergillosis in critically ill patients in the post-influenza era must be re-evaluated. Clinical studies should be conducted in order to know the clinical significance of Aspergillus isolation in respiratory samples in intubated patients with primary influenza A(H1 N1) pneumonia. References 1.- Wauters J, et al. Invasive pulmonary aspergillosis is a frequent complication of critically ill H1 N1 patients: a retrospective study. Intensive Care Med. 2012;38(11):1761–8. 2.- Adalja AA, et al. Isolation of Aspergillus in three 2009 H1 N1 influenza patients. Influenza Other Respir Viruses. 2011;5(4):225–9. A824 Initial Therapeutic Strategy Of Invasive Candidiasis For Intensive Care Unit Patients: An Analysis From The China-Scan Study N. Cui1, D. Liu1, H. Wang1, L. Su1, H. Qiu2, R. Li3 1Peking Union Medical College Hospital, Critical Care Medicine, Beijing, China; 2Nanjing Zhongda Hospital, Southeast University School of Medicine, Critical Care Medicine, Nanjing, China; 3Peking University First Hospital, Peking University, Research Center for Medical Mycology, Beijing, China Correspondence: N. Cui – Peking Union Medical College Hospital, Critical Care Medicine, Beijing, China Introduction: The empiric or pre-emptive approach can be used as a better target therapy in antifungal treatment and affect mortality.

Hospital, Peking University, Research Center for Medical Mycology, Beijing, China Correspondence: N. Cui – Peking Union Medical College Hospital, Critical Care Medicine, Beijing, China Introduction: The empiric or pre-emptive approach can be used as a better target therapy in antifungal treatment and affect mortality. Objectives: To investigate the impact of initial antifungal therapeutic strategies on the prognosis of invasive Candida infections (ICIs) in intensive care units (ICUs) in China. Methods: A total of 306 patients with proven ICIs in the China Survey of Candidiasis study were analyzed. Empiric, pre-emptive, and targeted therapy were adopted based on starting criteria including clinical, microbiological, and other conventional prediction rules. The primary outcome was ICU/hospital mortality.

Objectives: To investigate the impact of initial antifungal therapeutic strategies on the prognosis of invasive Candida infections (ICIs) in intensive care units (ICUs) in China. Methods: A total of 306 patients with proven ICIs in the China Survey of Candidiasis study were analyzed. Empiric, pre-emptive, and targeted therapy were adopted based on starting criteria including clinical, microbiological, and other conventional prediction rules. The primary outcome was ICU/hospital mortality. Results: Compared with the empirical initial antifungal therapy and targeted initial antifungal therapy, patients with pre-emptive initial antifungal therapy had significantly less clinical remission [11/53 (21.2 %) vs. 61/142 (43.3 %) vs. 22/73 (30.1 %), p = 0.009], higher ICU [26/53 (57.8 %) vs. 42/142 (32.2 %) vs. 27/73 (43.5 %), p = 0.008] and hospital mortality [27/53 (60.0 %) vs. 43/142 (32.8 %) vs. 29/73 (46.8 %), p = 0.004] and more microbiological persistence [9/53 (17.0 %) vs. 6/142 (4.2 %) vs. 9/73 (12.3 %), p = 0.011]. Kaplan-Meier survival analysis revealed that ICI patients with pre-emptive initial antifungal therapy and targeted initial antifungal therapy were associated with reduced hospital duration compared with patients with empirical initial antifungal therapy after confirmation of fungal infection (log-rank test: p = 0.021).Multivariate regression analysis provided evidence that initial empirical antifungal therapy was an independent predictor for hospital mortality in ICI patients[odds ratio 0.349(95 % confidence interval 0.168-0.724); p = 0.005).

nitial antifungal therapy after confirmation of fungal infection (log-rank test: p = 0.021).Multivariate regression analysis provided evidence that initial empirical antifungal therapy was an independent predictor for hospital mortality in ICI patients[odds ratio 0.349(95 % confidence interval 0.168-0.724); p = 0.005). Conclusions: The initial therapeutic strategy for invasive candidiasis was independently associated with hospital mortality. Prompt empirical antifungal therapy could be critical to decrease early hospital mortality. Acknowledgement This study was supported by Merck Sharp & Dohme China. We would like to thank the patients and investigators who participated in this study. We also acknowledge the investigators at each study site, without whom this study would not have been possible. A825 De-escalation of antifungal treatment in critically ill patients: incidence, associated factors and safety K. Jaffal1, A. Rouzé1, J. Poissy1, B. Sendid2, S. Nseir1 1Lille University Hospital, ICU, Lille, France; 2Lille University Hospital, Mycology and Parasitology Lab, Lille, France Correspondence: S. Nseir – Lille University Hospital, ICU, Lille, France Introduction: Although antifungal treatment is common in critically ill patients, only a small proportion of patients receiving antifungal treatment have confirmed fungal infection. Side effects of this treatment include toxicity, resistance and unnecessary high costs. In addition, recent studies suggested no benefit of empirical antifungal treatment in these patients.

on in critically ill patients, only a small proportion of patients receiving antifungal treatment have confirmed fungal infection. Side effects of this treatment include toxicity, resistance and unnecessary high costs. In addition, recent studies suggested no benefit of empirical antifungal treatment in these patients. Objectives: The aim of this study was to identify the incidence, associated factors, and safety of de-escalation of antifungal treatment. Methods: Retrospective study, conducted in a 30-bed mixed ICU, during a 1-year period. All patients hospitalized for >5d and treated with antifungals for first suspected or proven fungal infection were included. Patients receiving prophylactic antifungals were excluded. De-escalation was defined as switch from initial systemic antifungal therapy drugs (except fluconazole) to triazoles, or stopping initial drugs within 6 days following their initiation. Patients with de-escalation were compared with those with no de-escalation using univariate and multivariate analysis. Results: Among the 234 patients who received systemic antifungals, 44 % received empirical, 23 % preemptive, and 33 % targeted treatment. Caspofungin (51 %), fluconazole (34 %), voriconazole (11 %), and liposomal amphotericin B (2 %) were the most frequently used antifungals. Antifungal treatment was de-escalated in 48 (20.5 %) patients.

234 patients who received systemic antifungals, 44 % received empirical, 23 % preemptive, and 33 % targeted treatment. Caspofungin (51 %), fluconazole (34 %), voriconazole (11 %), and liposomal amphotericin B (2 %) were the most frequently used antifungals. Antifungal treatment was de-escalated in 48 (20.5 %) patients. Factors associated with higher rate of de-escalation in univariate analysis were: sterile repeated mycological samples, empirical treatment, preemptive treatment, positive bacterial samples, apyrexia, and catecholamine withdrawal at 72 hours after initiation of antifungal treatment. Multifocal Candida colonization, and mechanical ventilation were associated with significantly lower rate of de-escalation by univariate analysis. In multivariate analysis, apyrexia at 72 h (OR 15 [2–120], p = 0.009), empirical treatment (3.5 [1.7-7.5], p = 0.001), and mechanical ventilation (0.22 [0.08-0.62], p = 0.004) were independently associated with de-escalation. No significant difference was found in duration of mechanical ventilation (med (IQR) 17 (5, 30) vs 20d (10, 35), p = 0.06), or length of ICU stay (24 (14, 40) vs 25d (14, 39) p = 0.79) between patients with de-escalation, and those with no de-escalation, respectively. Patients with de-escalation had lower ICU mortality rate, compared with those with no de-escalation (42 % vs 60 %, p = 0.017). However, 1-year mortality rate was similar in the two groups (60 % vs 66 %, p = 0.46).

25d (14, 39) p = 0.79) between patients with de-escalation, and those with no de-escalation, respectively. Patients with de-escalation had lower ICU mortality rate, compared with those with no de-escalation (42 % vs 60 %, p = 0.017). However, 1-year mortality rate was similar in the two groups (60 % vs 66 %, p = 0.46). Conclusions: De-escalation was performed in 20.5 % of patients receiving systemic antifungals. Factors independently associated with de-escalation were apyrexia at 72 hours after initiation of antifungals, empirical treatment strategy, and mechanical ventilation. De-escalation of antifungal treatment seems to be safe in critically ill patients. A826 Epidemiology and predictors of candidemia mortality in a critical care setting E. Paramythiotou, M. Rizos, F. Frantzeskaki, A. Antoniadou, S. Vourli, L. Zerva, A. Armaganidis Attikon University Hospital, Athens, Greece Correspondence: E. Paramythiotou – Attikon University Hospital, Athens, Greece Introduction: Invasive candidiasis is an important cause of morbidity and mortality in the nosocomial setting and particularly in the Intensive Care Unit (ICU). Candidemia is ranked fourth as the cause of bloodstream infection in USA and it is the cause of 7 % of positive blood cultures in Europe. The increasing incidence of non-albicans Candida species is a matter of concern. Objectives: The aim of the present study was to record the epidemiology, risk factors, mortality, strain susceptibility to antifungal drugs, and to evaluate Ostrosky rule's capability to predict invasive candidiasis.

A826 Epidemiology and predictors of candidemia mortality in a critical care setting E. Paramythiotou, M. Rizos, F. Frantzeskaki, A. Antoniadou, S. Vourli, L. Zerva, A. Armaganidis Attikon University Hospital, Athens, Greece Correspondence: E. Paramythiotou – Attikon University Hospital, Athens, Greece Introduction: Invasive candidiasis is an important cause of morbidity and mortality in the nosocomial setting and particularly in the Intensive Care Unit (ICU). Candidemia is ranked fourth as the cause of bloodstream infection in USA and it is the cause of 7 % of positive blood cultures in Europe. The increasing incidence of non-albicans Candida species is a matter of concern. Objectives: The aim of the present study was to record the epidemiology, risk factors, mortality, strain susceptibility to antifungal drugs, and to evaluate Ostrosky rule's capability to predict invasive candidiasis. Methods: This is a clinical and microbiological retrospective study of all candidemia episodes which were registered during a ten - year period (between 1/1/2004 and 31/12/2014). Patient identification was performed through the records of the laboratory of Attikon university hospital. Medical records were then retrieved. Only the first candidemia episode was evaluated. Special forms were completed for each patient including demographic information, concomitant conditions, Apache II and Sofa severity scores the day of ICU admission, the risk factors within the preceding 10 days, data of colonization and candidemia related information.

. Only the first candidemia episode was evaluated. Special forms were completed for each patient including demographic information, concomitant conditions, Apache II and Sofa severity scores the day of ICU admission, the risk factors within the preceding 10 days, data of colonization and candidemia related information. Results: Attikon hospital is a 640 - bed teaching tertiary care hospital with a 25 - bed medical and surgical ICU. Among them 70 patients developed candidemia. For 7 patients the medical records were incomplete so they were excluded. Mean patients' age was 67 years ± 15.22. Median ICU length of stay was 40.74 + 36 days. Former duration of hospitalization was 16 + 16 days. Medical cause of admission was present in 29 cases and surgical in 34 cases. Species isolated included C. albicans 16,C.parapsilosis 15 , C. tropicalis 5 , Candida spp, 18, C.glabrata 4, C krusei 2, and C. non - albicans 3. Median time elapsed between ICU admission and candidemia was 19 days (3–85). Mean Apache II score was 19.5 ± 6.4 (range 12–37) on the day of admission, SOFA score 8 ± 3.6 and overall mortality was 74.6 %. Candidemia was considered the cause of death in 13 cases (20.6 %). Ostrosky's prediction rule was positive in 30 patients. Thirteen pts were submitted to an intra-abdominal operation. Blood cultures were sterilized in thirty patients. Twenty pts received TPN prior to candidemia episode. Fourteen pts were receiving steroid therapy and three were receiving immunocompromising therapy. Caspofungin was the most commonly introduced treatment.

s. Thirteen pts were submitted to an intra-abdominal operation. Blood cultures were sterilized in thirty patients. Twenty pts received TPN prior to candidemia episode. Fourteen pts were receiving steroid therapy and three were receiving immunocompromising therapy. Caspofungin was the most commonly introduced treatment. Conclusions: The burden of fungemia episodes is not very high when compared to other blood infections but they are associated with a high mortality perhaps due to the severity of the underlying disease. A high Apache II score at admission, multiple site colonization in combination with abdominal surgery should raise a high suspicion index and a prophylactic therapy should start. Non - albicans species are increasing.

s but they are associated with a high mortality perhaps due to the severity of the underlying disease. A high Apache II score at admission, multiple site colonization in combination with abdominal surgery should raise a high suspicion index and a prophylactic therapy should start. Non - albicans species are increasing. A827 Host or worst? Role of enterococcus and candida in respiratory cultures in lung transplant recipients J. Riera1,2,3, J. Gottlieb4,5, M. Greer4,5, O. Wiesner4, M. Martínez1, M. Acuña1, J. Rello1,3, T. Welte4,5 1Vall d'Hebron University Hospital, Critical Care, Barcelona, Spain; 2Universitat Autònoma de Barcelona, Barcelona, Spain; 3CIBERES, Madrid, Spain; 4Hannover Medical School, Department of Respiratory Medicine, Hannover, Germany; 5German Centre of Lung Research (DZL/BREATH), Hannover, Germany Correspondence: J. Riera – Vall d'Hebron University Hospital, Critical Care, Barcelona, Spain Introduction: Lung transplant (LT) recipients often receive antibiotic treatments and prophylaxis that may modify the normal respiratory microbiome [1]. Recent studies suggest that a priori non-pathogenic organisms, such as enterococci and candida, may cause infection, especially in immunosuppressed patients [2,3]. In the LT recipient it can be difficult to distinguish between organisms causing pneumonia, tracheobronchitis or if they are merely colonizers [4].

ome [1]. Recent studies suggest that a priori non-pathogenic organisms, such as enterococci and candida, may cause infection, especially in immunosuppressed patients [2,3]. In the LT recipient it can be difficult to distinguish between organisms causing pneumonia, tracheobronchitis or if they are merely colonizers [4]. Objectives: Assess the association between candida spp and enterococcus spp isolates in bronchoalveolar lavage (BAL) with signs and symptoms of respiratory infection, and compare this association with 2 control groups: the first formed by BAL with positive cultures for pseudomonas aeruginosa and the second one by sterile BAL. Methods: Retrospective analysis of bronchoscopies performed in LT recipients at Hannover Medical School between January 2008 and August 2015. BAL cultures positive for candida spp, enterococcus spp and pseudomonas aeruginosa were identified, as well as sterile cultures (Fig. 11) Reported symptoms and signs suggestive of respiratory infection at the time of the bronchoscopy were then evaluated. Symptoms were recorded since August 2012 using a questionnaire fulfilled by the patients.

r candida spp, enterococcus spp and pseudomonas aeruginosa were identified, as well as sterile cultures (Fig. 11) Reported symptoms and signs suggestive of respiratory infection at the time of the bronchoscopy were then evaluated. Symptoms were recorded since August 2012 using a questionnaire fulfilled by the patients. Results: BAL cultures from 1995 bronchoscopies were analyzed. In total 110 BAL cultures positive for candida spp were identified along with 125 for enterococcus spp, 1103 for pseudomonas aeruginosa and 359 sterile samples (Figures are displayed in pic_01). Levels of alveolar neutrophils in the BAL of patients with positive cultures for candida spp were not different than those of patients with positive cultures for pseudomonas aeruginosa. The same was true for serum CRP (P = 0.45) and blood leucocytes counts (P = 0.09). Patients with positive cultures for enterococcus spp had lower number of alveolar neutrophils compared with patients with pseudomonas aeruginosa (P = 0.02) and lower blood leucocyte count (P = 0.01). No differences were found between the serum level of CRP (P = 0.15). Patients with candida spp positive cultures had significantly more alveolar neutrophils (P < 0.01), higher blood leucocytes (P < 0.01) and higher serum CRP (P = 0.04) than patients with sterile BAL. Patients with enterococcus spp positive cultures had more alveolar neutrophils (P < 0.001) and higher serum CRP (P = 0.04) than patients with sterile BAL. No differences regarding blood leucocytes (P = 0.25) were evident.

higher blood leucocytes (P < 0.01) and higher serum CRP (P = 0.04) than patients with sterile BAL. Patients with enterococcus spp positive cultures had more alveolar neutrophils (P < 0.001) and higher serum CRP (P = 0.04) than patients with sterile BAL. No differences regarding blood leucocytes (P = 0.25) were evident. Conclusions: Candida spp and enterococcus spp may cause some degree of inflammation in the lung allograft. It would be reasonable to treat a LT recipient with signs of infection and positive BAL cultures for candida spp. Enterococcus spp may cause a milder degree of inflammation. References 1. Charlson et al. AJRCCM 2012;186:536–45. 2. Albert at el. ICM 2014;40:1313–22. 3. Bucheli et al. Trans Inf Dis 2014;16:26–36. 4. Riera et al. ERJ 2015;45:726–37. Grant acknowledgement Supported by FUCAP and CIBERES.Fig. 11 (abstract A827). ᅟ A828 Use of biomarkers in Candida sp infections after lung transplantation E. Atchade, T. Mignot, S. Houzé, S. Jean-Baptiste, G. Thabut, B. Lortat-Jacob, S. Tanaka, P. Augustin, M. Desmard, P. Montravers Hopital Bichat Claude Bernard, Paris, France Correspondence: E. Atchade – Hopital Bichat Claude Bernard, Paris, France Introduction:Candida sp. colonisation in respiratory tract is frequent after lung transplantation (LT), but only a small proportion of lung recipients develop an invasive candida infection (1). The distinction between colonisation and invasive infection remains difficult leading clinicians to use biomarkers in their decision making process.

p. colonisation in respiratory tract is frequent after lung transplantation (LT), but only a small proportion of lung recipients develop an invasive candida infection (1). The distinction between colonisation and invasive infection remains difficult leading clinicians to use biomarkers in their decision making process. Objectives: To assess the value of fungal biomarkers in the diagnosis of Candida sp infections after LT. Methods: This prospective, monocentric, study performed over 12-months assessed the value of a twice weekly dosage of fungal biomakers (Candida serology IgA, IgM, IgG, ß-D-Glucan (BDG) and mannan antigens) in ICU patients after LT. Proven/probable/possible infection was defined according to the EORTC/MSG criteria. Colonisation was defined by presence of Candida sp in respiratory samples without any sign of invasive infection. The study was approved by the Ethical Committee. Results are presented as means.

annan antigens) in ICU patients after LT. Proven/probable/possible infection was defined according to the EORTC/MSG criteria. Colonisation was defined by presence of Candida sp in respiratory samples without any sign of invasive infection. The study was approved by the Ethical Committee. Results are presented as means. Results: We analysed 39 ICU patients after LT. 77 % had a Candida sp colonisation while an invasive infection was proven in 4 (10 %) patients. Candida albicans was cultured from 72 % of the pulmonary samples. 50 % of the invasive infections were related to C. glabrata. Results of biomarkers dosages are presented in the table. Positive Candida IgG serology was observed in 45 % of the cases. Mortality rate at 6-months after LT was 71 % in the immunised patients versus 26 % in non-immunised patients. An invasive candidiasis (IC) was present in 14 % of the immunised patients versus 10 % in non-immunised patients. At least one BDG dosage was positive in 60 % of the cases. BDG dosage value decreased after surgery, reaching a non-significant value after the 4th day. In proven IC, BDG measurements reached concentrations >500 pg/ml, 16 days before initiation of antifungal treatment. No patient had positive mannan antigen measurement. Conclusions: A twice weekly dosage of BDG seems to be useful in the decision making process for early initiation of antifungal therapy in LT patients. The cutoff for a significant value of BDG needs to be defined. Pre-transplantation assessment of Candida IgG serology could help to identify patients at risk of post-operative fungal infection.

kly dosage of BDG seems to be useful in the decision making process for early initiation of antifungal therapy in LT patients. The cutoff for a significant value of BDG needs to be defined. Pre-transplantation assessment of Candida IgG serology could help to identify patients at risk of post-operative fungal infection. References (1) Flume, Am J Respir Crit Care Med,1994; 149(6):1601–7. Grant acknowledgement None.Table 7 (abstract A828). Dosage of biomarkers after LT Colonised pts Non colonised pts Infected pts Ischemic bronchitis - + - + BDG+,%;mean (pg/mL) 14;32 50;125 11;38 50;63 100;239 IgM + , % 10 0 0 0 0 Ig G +, % 25 56 0 0 33 Ig G + < 7 days, % 20 38 0 33 IgG seroconversions, % 0 21 - 0 100 Ig A +, % 0 14 0 0 33 Mannan antigen +, % 0 0 0 0 0

None.Table 7 (abstract A828). Dosage of biomarkers after LT Colonised pts Non colonised pts Infected pts Ischemic bronchitis - + - + BDG+,%;mean (pg/mL) 14;32 50;125 11;38 50;63 100;239 IgM + , % 10 0 0 0 0 Ig G +, % 25 56 0 0 33 Ig G + < 7 days, % 20 38 0 33 IgG seroconversions, % 0 21 - 0 100 Ig A +, % 0 14 0 0 33 Mannan antigen +, % 0 0 0 0 0 A829 Curative or pre-emptive antifungal therapy in invasive candidiasis in severely burned patients? S. Soussi1, E. Dudoignon1, A. Ferry1, M. Chaussard1, M. Benyamina1, A. Alanio2, S. Touratier3, M. Chaouat4, M. Lafaurie5, M. Mimoun4, A. Mebazaa1, M. Legrand1 1AP-HP, GH St-Louis-Lariboisière, Department of Anesthesiology and Critical Care and SMUR and Burn Unit, Paris, France; 2AP-HP, GH St-Louis-Lariboisière, Mycology Unit, Paris, France; 3AP-HP, GH St-Louis-Lariboisière, Pharmacy Unit, Paris, France; 4AP-HP, GH St-Louis-Lariboisière, Plastic Surgery and Burn Unit, Paris, France; 5AP-HP, GH St-Louis-Lariboisière, Department of Infectious Diseases, Paris, France Correspondence: S. Soussi – AP-HP, GH St-Louis-Lariboisière, Department of Anesthesiology and Critical Care and SMUR and Burn Unit, Paris, France Introduction: The antifungal (AF) therapy strategy (pre-emptive vs culture based treatment) in intensive care unit is a matter of debate [1]. The necessity to not delay the initiation of the AF in invasive candidiasis (IC) must be balanced with the cost and risk of selecting resistant pathogens when AF are prescribed too widely. Burn patients are at risk of IC because of the frequent use of antibiotics and immunodeficiency.

re unit is a matter of debate [1]. The necessity to not delay the initiation of the AF in invasive candidiasis (IC) must be balanced with the cost and risk of selecting resistant pathogens when AF are prescribed too widely. Burn patients are at risk of IC because of the frequent use of antibiotics and immunodeficiency. Objectives: To evaluate our antifungal (AF) therapy strategy in suspected or proven IC in terms of prognosis and risk factors of IC. Methods: Observational, descriptive, retrospective study conducted from June 2012 to September 2015 in the Saint Louis Hospital Burn Unit. Inclusion criteria: patients treated with pre-emptive (severe sepsis or septic shock with Candida Sp colonization) or curative (proven, PIC) AF. The outcome was the PIC (candidemia and/or positive peritoneal sample). Clinical characteristics, organ supports, AF treatments and outcome were collected and compared between PIC and suspected IC (SIC). The results are presented in median (IQR) or n (%).

hock with Candida Sp colonization) or curative (proven, PIC) AF. The outcome was the PIC (candidemia and/or positive peritoneal sample). Clinical characteristics, organ supports, AF treatments and outcome were collected and compared between PIC and suspected IC (SIC). The results are presented in median (IQR) or n (%). Results: 715 patients were admitted during the study period including 184 with a total body surface area (TBSA) >20 %. 44 treated with AF including 14 PIC (32 %). In those 44 patients: age 55 (43–67), TBSA (43–67), SAPSII 41 (29–53), ABSI 10 (8–12) and SOFA 10 (5–15). Renal replacement therapy 22 (50 %), mechanical ventilation 36 (82 %), parenteral nutrition 7 (16 %). Inhospital mortality = 60 % (68 % SIC vs 32 % PIC, p = 0.33). 3 patients with PIC (21 %) were treated before the IC diagnosis (2 because of filamentous infection before the PIC). The delay between admission and AF treatment initiation was 16 days. Patients characteristics, organs supports were not significantly different between PIC and SIC at the treatment initiation except for the SAPSII (PIC 43 (29–57) VS SIC 37 (26–49), p = 0.03). 30 patients (68 %) received an echinocandin as a first-line treatment. 6 (5.5-6.5) sites were monitored for Candida colonization the week before treatment initiation. Patients with PIC had higher colonization index than those with SIC (55 % vs 33 %, p = 0.01) and a candida score significantly higher (4 vs 3 (0.5-3.5) respectively, p = 0.03). A semi-quantitative estimation of the fungal inoculum had no predictive value.

d for Candida colonization the week before treatment initiation. Patients with PIC had higher colonization index than those with SIC (55 % vs 33 %, p = 0.01) and a candida score significantly higher (4 vs 3 (0.5-3.5) respectively, p = 0.03). A semi-quantitative estimation of the fungal inoculum had no predictive value. Conclusions: In this study, the majority of PIC were treated after diagnosis confirmation. Only 1/30 (3 %) patient treated preemptively did declare a PIC. The outcome was not different when the treatment was initiated after confirmation. The results of this study highlight the difficulty to identify patients at highest risk of IC, and question the strategy of preemptive treatment in this population. References [1] Intensive care Med (2014) 40: 1429–48

Conclusions: In this study, the majority of PIC were treated after diagnosis confirmation. Only 1/30 (3 %) patient treated preemptively did declare a PIC. The outcome was not different when the treatment was initiated after confirmation. The results of this study highlight the difficulty to identify patients at highest risk of IC, and question the strategy of preemptive treatment in this population. References [1] Intensive care Med (2014) 40: 1429–48 A830 An observational study demonstrating a possible link between a procalcitonin driven reduction in antibiotic use and systemic fungal infections M.A. Sheils1, C. Patel1, L. Mohankumar2, N. Akhtar1 1Dudley Group of Hospitals NHS Foundation Trust, Anaesthetics, Dudley, United Kingdom; 2Dudley Group of Hospitals NHS Foundation Trust, Microbiology, Dudley, United Kingdom Correspondence: M.A. Sheils – Dudley Group of Hospitals NHS Foundation Trust, Anaesthetics, Dudley, United Kingdom Introduction: Invasive candida infections are implicated in 10 % of intensive care acquired bloodstream infections with a reported mortality of 42.5 %. Amongst the risk factors for invasive fungal infections is exposure to broad spectrum antibiotics and fungal colonisation of mucosal surfaces. Procalcitonin (PCT) guided prescribing algorithms have been proven to safely reduce patient exposure to antibiotics. Objectives: We wanted to determine whether PCT guided antibiotic rationalisation could reduce fungal colonisation and antifungal usage.

A830 An observational study demonstrating a possible link between a procalcitonin driven reduction in antibiotic use and systemic fungal infections M.A. Sheils1, C. Patel1, L. Mohankumar2, N. Akhtar1 1Dudley Group of Hospitals NHS Foundation Trust, Anaesthetics, Dudley, United Kingdom; 2Dudley Group of Hospitals NHS Foundation Trust, Microbiology, Dudley, United Kingdom Correspondence: M.A. Sheils – Dudley Group of Hospitals NHS Foundation Trust, Anaesthetics, Dudley, United Kingdom Introduction: Invasive candida infections are implicated in 10 % of intensive care acquired bloodstream infections with a reported mortality of 42.5 %. Amongst the risk factors for invasive fungal infections is exposure to broad spectrum antibiotics and fungal colonisation of mucosal surfaces. Procalcitonin (PCT) guided prescribing algorithms have been proven to safely reduce patient exposure to antibiotics. Objectives: We wanted to determine whether PCT guided antibiotic rationalisation could reduce fungal colonisation and antifungal usage. Methods: We undertook a retrospective observational study at a nine bedded ICU department in the United Kingdom. We collected data on all patients admitted to the unit in the year prior and post the introduction of PCT guided rationalisation of antibiotics. We used the pharmacy database to assess the use of antibiotics, correcting for changes in costs over this time. We used the microbiology database to assess the rate of patients colonising fungal species and those requiring treatment.

the year prior and post the introduction of PCT guided rationalisation of antibiotics. We used the pharmacy database to assess the use of antibiotics, correcting for changes in costs over this time. We used the microbiology database to assess the rate of patients colonising fungal species and those requiring treatment. Results: Since the introduction of PCT, the average expenditure on antibiotics per ICU admission fell 35.11 % (p 0.01). The rate of ICU patients colonised with a fungal species fell from 32.3 % to 14.89 % (p < 0.0001). The incidence of patient's prescribed systemic antifungal therapy fell from 14.58 % to 4.26 % (p < 0.0001). Conclusions: We demonstrated a significant reduction in patients colonised with fungal species and those requiring anti-fungal therapy since introducing PCT guided rationalisation of antibiotics. A prospective randomised controlled trial is required to assess whether this equates to improved patient outcome. References 1) Kett DH, Azoulay E, Echeverria PM, Vincent JL. Candida bloodstream infections in intensive care units: Analysis of the extended prevalence of infection in intensive care unit study. Crit Care Med. 2011;39:665–670. 2) Eggimann P, Bille J, Marchetti O. Diagnosis of invasive candidiasis in ICU. Ann Intensive Care. 2011; 1: 37. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3224461/). 3) Eggimann P, Garbino J, Pittet D. Epidemiology of Candida species infections in critically ill non-immunosuppressed patient. Lancet Infect Dis. 2003;3:685–702. doi: 10.1016/S1473-3099(03)00801-6. 4) Http://www.who.int/dg/speeches/2012/amr_20120314/en/index.html

2) Eggimann P, Bille J, Marchetti O. Diagnosis of invasive candidiasis in ICU. Ann Intensive Care. 2011; 1: 37. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3224461/). 3) Eggimann P, Garbino J, Pittet D. Epidemiology of Candida species infections in critically ill non-immunosuppressed patient. Lancet Infect Dis. 2003;3:685–702. doi: 10.1016/S1473-3099(03)00801-6. 4) Http://www.who.int/dg/speeches/2012/amr_20120314/en/index.html 5) Karlsson S, Heikkinen M, Pettilä V et al.: Predictive value of procalcitonin decrease in patients with severe sepsis: a prospective observational study. Crit care 2010, Grant acknowledgement None A831 Invasive fungal infections in critically ill burn patients admitted to a colombian national referral intensive care unit (ICU) S.K. Pacheco Noriega, N. Navarrete Aldana, J.L. Ávila León, J. Durand Baquero, F. Fernández Bernal Hospital Simón Bolívar, Burn ICU, Bogotá, Colombia Correspondence: S.K. Pacheco Noriega – Hospital Simón Bolívar, Burn ICU, Bogotá, Colombia Introduction: Invasive fungal infections remain a challenge in burn intensive care. In our country there is not available data about the epidemiology of the fungal species involved, the correct management, and the patients' outcome. Objectives: To identify 1) the fungal organism/sp most commonly isolated 2) the fungal infection most commonly diagnosed based on microbiological data 3) the most prevalent risk factors involved 5) the most used antifungal agent.

A831 Invasive fungal infections in critically ill burn patients admitted to a colombian national referral intensive care unit (ICU) S.K. Pacheco Noriega, N. Navarrete Aldana, J.L. Ávila León, J. Durand Baquero, F. Fernández Bernal Hospital Simón Bolívar, Burn ICU, Bogotá, Colombia Correspondence: S.K. Pacheco Noriega – Hospital Simón Bolívar, Burn ICU, Bogotá, Colombia Introduction: Invasive fungal infections remain a challenge in burn intensive care. In our country there is not available data about the epidemiology of the fungal species involved, the correct management, and the patients' outcome. Objectives: To identify 1) the fungal organism/sp most commonly isolated 2) the fungal infection most commonly diagnosed based on microbiological data 3) the most prevalent risk factors involved 5) the most used antifungal agent. Methods: An observational, descriptive, and retrospective study was done including consecutive patients from June 2012 to December 2015. Setting: National referral Burn ICU in a developing country, 10 beds. Inclusion criteria: Every patient admitted to ICU with positive fungal samples. Patients were followed until discharge from hospital. Microbiology samples were collected before the beginning of antifungal treatment and during the follow- up.

er 2015. Setting: National referral Burn ICU in a developing country, 10 beds. Inclusion criteria: Every patient admitted to ICU with positive fungal samples. Patients were followed until discharge from hospital. Microbiology samples were collected before the beginning of antifungal treatment and during the follow- up. Results: 52 patients were included, 65.38 % were male. Mean Age 45.2 (SD +/−19.1), median Apache II score was 15, median SAPS II was 41, median ABSI was 7. 42.3 % were from rural areas and 28.85%were working during the injury, followed by home incidents 21.15 %. Fire was the most common cause found in 50 % of patients, 17.3 % had electrical burns and 17.3 % were in an explosion context. 1 patient suffered a chemical burn. Inhalation injury was present 57.7 %. Total burn surface area (TBSA) was 44.2 %: 40-60 %, 30 %:20 %-40 %, and 15.4 %:60-80 %. Mean stay in ICU 43.8 days (SD 52.5), median 28 days. C. albicans was the most commonly isolated species (sp) (34,6 %), Candidemia was the most common diagnosed infection. The sp isolated in blood cultures (BC) were: 23,1 % C. albicans, 21,1 % C. haemulonii, 14.4 % C. parasilopsis. We had 2 + BC for Trichosporon asahii. Catheter -related infection by C. albicans, C. parasilopsis, and candida haemulonii was diagnosed in 6 patients. Positive Urine samples were found mostly for C. albicans 11.5 %. The most frequently factors associated with fungal infection were: > than 7 days in the ICU 90.4 %, urinary catheter 90.4 %, broad- spectrum antibiotic exposure 88.5 %, indwelling central venous catheter (CVC) 82.7 %, feeding tube 82.7 %, total parenteral nutrition 63.5 %, Invasive mechanical ventilation 63.5 %. To highlight an association with Acinetobacter baumannii in 44.2 % of our patients. Doctors chose Fluconazole in 88.9 % as a first line of therapy. An antibiogram was performed and the susceptibility was confirmed. ICU mortality rate, 26.9 %.

%, total parenteral nutrition 63.5 %, Invasive mechanical ventilation 63.5 %. To highlight an association with Acinetobacter baumannii in 44.2 % of our patients. Doctors chose Fluconazole in 88.9 % as a first line of therapy. An antibiogram was performed and the susceptibility was confirmed. ICU mortality rate, 26.9 %. Conclusions: In our environment, C. albicans continues to be the species that causes the largest number of invasive candidiasis. Prolonged stay in the ICU is an important risk factor to develop fungal infections. Even with the particular features of a burn patient, their complexity, and the negative impact of each infection; Fluconazole keeps having an important role in the treatment as a first line.

s the largest number of invasive candidiasis. Prolonged stay in the ICU is an important risk factor to develop fungal infections. Even with the particular features of a burn patient, their complexity, and the negative impact of each infection; Fluconazole keeps having an important role in the treatment as a first line. A832 The effect of introduction of daily chlorhexidine bathing on healthcare-associated infections and acquisition of multi-drug resistant organisms E. Ahmadnia1, J.S. Hadley1,2, M. Millar3, D. Hall1, H. Hewitt1 1Adult Critical Care Unit, Royal London Hospital, Barts Health NHS Trust, London, United Kingdom; 2Queen Mary University of London, London, United Kingdom; 3Royal London Hospital, Barts Health NHS Trust, Microbiology Department, London, United Kingdom Correspondence: E. Ahmadnia – Adult Critical Care Unit, Royal London Hospital, Barts Health NHS Trust, London, United Kingdom Introduction: It has been suggested that daily bathing with chlorhexidine impregnated cloths may significantly reduce the acquisition of multi-drug resistant organisms (MDROs), incidence of central line associated bloodstream infections (CLABSIs), and the development of intensive care unit (ICU) acquired bloodstream infections [1]. However, more recent data have failed to support daily bathing of critically ill patients with chlorhexidine for these purposes [2]. Objectives: To determine if the implementation of a daily chlorhexidine bathing regimen affects acquisition rates of MDROs, the incidence of CLABSIs, and ICU bacteraemias.

A832 The effect of introduction of daily chlorhexidine bathing on healthcare-associated infections and acquisition of multi-drug resistant organisms E. Ahmadnia1, J.S. Hadley1,2, M. Millar3, D. Hall1, H. Hewitt1 1Adult Critical Care Unit, Royal London Hospital, Barts Health NHS Trust, London, United Kingdom; 2Queen Mary University of London, London, United Kingdom; 3Royal London Hospital, Barts Health NHS Trust, Microbiology Department, London, United Kingdom Correspondence: E. Ahmadnia – Adult Critical Care Unit, Royal London Hospital, Barts Health NHS Trust, London, United Kingdom Introduction: It has been suggested that daily bathing with chlorhexidine impregnated cloths may significantly reduce the acquisition of multi-drug resistant organisms (MDROs), incidence of central line associated bloodstream infections (CLABSIs), and the development of intensive care unit (ICU) acquired bloodstream infections [1]. However, more recent data have failed to support daily bathing of critically ill patients with chlorhexidine for these purposes [2]. Objectives: To determine if the implementation of a daily chlorhexidine bathing regimen affects acquisition rates of MDROs, the incidence of CLABSIs, and ICU bacteraemias. Methods: A quality improvement project was conducted at a 44 bedded adult critical care unit within a UK University Hospital (incorporating major trauma, medical, and surgical patients). During the 1 year control period (December 2013 to November 2014), all patients were bathed using soap and water. During the subsequent intervention period (December 2014 to November 2015), all patients were bathed using 2 % chlorhexidine impregnated cloths (Clinell, GAMA Healthcare). The acquisition of MDROs, incidence of CLABSIs and ICU bacteraemias were recorded during these periods (6 months pre- and 6 months post-chlorhexidine for CLABSIs, one year for the other outcomes).

ber 2014 to November 2015), all patients were bathed using 2 % chlorhexidine impregnated cloths (Clinell, GAMA Healthcare). The acquisition of MDROs, incidence of CLABSIs and ICU bacteraemias were recorded during these periods (6 months pre- and 6 months post-chlorhexidine for CLABSIs, one year for the other outcomes). Results: The study covered 34317 patient bed days (16887 pre- and 17430 post-introduction of chlorhexidine bathing). There were an identical number of MDRO acquisitions in each group (290), giving rise to an MDRO acquisition rate per 1000 bed days of 17.17 in the control group compared to 16.63 in the chlorhexidine group (P = 0.70). CLABSI incidence per 1000 bed days was higher in the control group compared to the chlorhexidine group (9.51 vs 7.90; P = 0.26). The incidence of significant bacteraemias per 1000 bed days was similar in the the two groups (5.57 before and 5.45 during chlorhexidine bathing; P = 0.88), but the incidence of bacteraemias due to skin commensals per 1000 bed days was lower in the chlorhexidine group (7.22 vs 5.45; P = 0.04).

(9.51 vs 7.90; P = 0.26). The incidence of significant bacteraemias per 1000 bed days was similar in the the two groups (5.57 before and 5.45 during chlorhexidine bathing; P = 0.88), but the incidence of bacteraemias due to skin commensals per 1000 bed days was lower in the chlorhexidine group (7.22 vs 5.45; P = 0.04). Conclusions: At our large University Hospital ICU with a heterogeneous patient population, the introduction of routine daily chlorhexidine-impregnated cloth bathing appears to significantly reduce the incidence of bacteraemias due to skin commensals and demonstrates a non-significant reduction in CLABSIs. Given the uncertainties surrounding diagnosis in the ICU, the effect seen may be of benefit in reducing the use of antibiotics to cover for these skin commensals - both in terms of antibiotic stewardship and health economics. References [1] Climo MW et al. Effect of daily chlorhexidine bathing on hospital-acquired infection. N Engl J Med. 2013;368(6):533–542. [2] Noto MJ et al. Chlorhexidine bathing and health care-associated infections: a randomized clinical trial. JAMA. 2015;313(4):369–378. Grant acknowledgement Clinell (GAMA Healthcare) provided the first 6 months of chlorhexidine cloths without charge.

[1] Climo MW et al. Effect of daily chlorhexidine bathing on hospital-acquired infection. N Engl J Med. 2013;368(6):533–542. [2] Noto MJ et al. Chlorhexidine bathing and health care-associated infections: a randomized clinical trial. JAMA. 2015;313(4):369–378. Grant acknowledgement Clinell (GAMA Healthcare) provided the first 6 months of chlorhexidine cloths without charge. A833 Comparison of the efficacy of three cutaneous antiseptic solutions for preventing catheter colonization: a multicenter, prospective, open-label, parallel, randomized controlled study H. Yasuda1,2, M. Sanui3, T. Komuro4, S. Kawano4, K. Andoh4, H. Yamamoto4, E. Noda4, J. Hatakeyama4, N. Saitou4, H. Okamoto4, A. Kobayashi4, T. Takei4, S. Matsukubo4, JSEPTIC (Japanese Society of Education for Physicians and Trainees in Intensive Care) Clinical Trial Group 1Japanese Red Cross Musashino Hospital, Emergency And Critical Care Medicine, Tokyo, Japan; 2Kameda Medical Center, Intensive Care Medicine, Chiba, Japan; 3Saitama Medical Center, Jichi Medical University, Department of Anesthesiology and Critical Care Medicine, Saitama, Japan; 4Japanese Society of Education for Physicians and Trainees in Intensive Care, Tokyo, Japan Correspondence: H. Yasuda – Japanese Red Cross Musashino Hospital, Emergency And Critical Care Medicine, Tokyo, Japan Introduction: The current CDC guideline published in 2011 for the prevention of intravascular catheter-related infections recommends skin preparation with a greater than 0.5 % chlorhexidine with alcohol solution before CVCs or ACs placement and with dressing changes, which was changed from 2 % chlorhexidine recommended in the 2002 guideline. However, few studies investigated the superiority of 1 % CHG over either 0.5 % CHG or 10 % PVI for the prevention of catheter colonization as CDC guideline recommends.

ith alcohol solution before CVCs or ACs placement and with dressing changes, which was changed from 2 % chlorhexidine recommended in the 2002 guideline. However, few studies investigated the superiority of 1 % CHG over either 0.5 % CHG or 10 % PVI for the prevention of catheter colonization as CDC guideline recommends. Objectives: Efficacy comparison of three antiseptic solutions [10 % aqueous povidone-iodine (PVI), and 0.5 % and 1.0 % alcoholic chlorhexidine gluconate (CHG)] for preventing intravascular catheter colonization. Methods: This was a open-label, multicenter, prospective, randomized controlled trial conducted at 15 ICUs in Japan. The intravascular catheters included central venous catheters (CVCs) and arterial catheters (ACs). Patients aged >18 years of age undergoing CVC and AC insertion in ICU were randomized to receive one of three antiseptic preparations pre-insertion. Catheters were removed when no longer necessary or if catheter-related infection was suspected. After catheter removal, distal tips were cultured using semi-quantitative/quantitative techniques. Catheter colonization and catheter-related bloodstream infection (CRBSI) incidences were compared.

reparations pre-insertion. Catheters were removed when no longer necessary or if catheter-related infection was suspected. After catheter removal, distal tips were cultured using semi-quantitative/quantitative techniques. Catheter colonization and catheter-related bloodstream infection (CRBSI) incidences were compared. Results: While a total of 1132 catheters were randomized, several catheters were excluded due to withdraw of their informed consent and lack of cultured catheters after randomization, and 796 (70 %) catheters were included in the full analysis (0.5 % CHG n = 261, 1.0 % CHG n = 278, and 10 % PVI n = 257). The median catheterization duration was 3.8 days (95 % CI: 2.0-6.7 days); no significant intergroup differences were observed (p = 0.43). Catheter-tip colonization incidence (per 1000 catheter days) was 10.5, 3.9, and 3.7 events in 10 % PVI, 1 % CHG, and 0.5 % CHG groups, respectively (p = 0.04). Catheter colonization risk was significantly higher in the 10 % PVI group. No significant intergroup differences CRBSI probability were observed (3.2 vs. 2.0 vs. 4.9 per 1000 catheter days, p = 0.42). Conclusions: In this multicenter prospective randomized controlled trial comparing the effectiveness of three cutaneous antiseptic solutions for the prevention of catheter colonization, either 0.5 % or 1.0 % CHG was superior to 10 % PVI.Fig. 12 (abstract A833). Catheter colonization

Results: While a total of 1132 catheters were randomized, several catheters were excluded due to withdraw of their informed consent and lack of cultured catheters after randomization, and 796 (70 %) catheters were included in the full analysis (0.5 % CHG n = 261, 1.0 % CHG n = 278, and 10 % PVI n = 257). The median catheterization duration was 3.8 days (95 % CI: 2.0-6.7 days); no significant intergroup differences were observed (p = 0.43). Catheter-tip colonization incidence (per 1000 catheter days) was 10.5, 3.9, and 3.7 events in 10 % PVI, 1 % CHG, and 0.5 % CHG groups, respectively (p = 0.04). Catheter colonization risk was significantly higher in the 10 % PVI group. No significant intergroup differences CRBSI probability were observed (3.2 vs. 2.0 vs. 4.9 per 1000 catheter days, p = 0.42). Conclusions: In this multicenter prospective randomized controlled trial comparing the effectiveness of three cutaneous antiseptic solutions for the prevention of catheter colonization, either 0.5 % or 1.0 % CHG was superior to 10 % PVI.Fig. 12 (abstract A833). Catheter colonization OUTCOMES IN NEUROINTENSIVE CARE A834 Hematoma volume and location as prognostic factors for clinical outcome in patients with spontaneous intracerebral hemorrhage H.B. Rotzel, A. Serrano Lázaro, D. Aguillón Prada, M. Rodriguez Gimillo, O. Diaz Barinas, M.L. Blasco Cortes, J. Ferreres Franco, J.M. Segura Roca, A. Carratalá Hospital Clinic Universitari Valencia, Valencia, Spain Correspondence: H.B. Rotzel – Hospital Clinic Universitari Valencia, Valencia, Spain Introduction: Spontaneous intracerebral hemorrhage (ICH) is the most fatal stroke subtype worldwide caused by spontaneous vascular rupture due to hypertension or amyloid angiopathy. An accurate prediction of ICH outcome would assist both families and physicians to decide therapies and monitorization at an early stage.

Introduction: Spontaneous intracerebral hemorrhage (ICH) is the most fatal stroke subtype worldwide caused by spontaneous vascular rupture due to hypertension or amyloid angiopathy. An accurate prediction of ICH outcome would assist both families and physicians to decide therapies and monitorization at an early stage. Objectives: To evaluate the relationship between the hematoma volume and location with mortality and functional outcome in patients with spontaneous ICH. Methods: We performed a prospective observational study, included patients admitted in ICU with spontaneous ICH. We determined hematoma volume at admission with Kothari modified formula (AxBxC/2) and divided them in two groups according the location as infratentorial or supratentorial. We collected GCS, SOFA, APACHE II and GRAEB at admission, medical history and complications during the first week in the ICU. We established modified Rankin Scale (mRS: poor outcome >2) and Glasgow Outcome Scale (GOS, poor outcome < 4) at ICU discharge. We used %, mean (SD) and median (minimal/maximum). T-Student and χ2 (p < 0.05) were used for the univariable analysis. We conducted a multivariable analysis for mortality with binary logistic regression (95 % CI, OR) p < 0,05. ROC curve was determined for the volume of hematoma associated with mortality (IC 95 % p < 0.05).

sed %, mean (SD) and median (minimal/maximum). T-Student and χ2 (p < 0.05) were used for the univariable analysis. We conducted a multivariable analysis for mortality with binary logistic regression (95 % CI, OR) p < 0,05. ROC curve was determined for the volume of hematoma associated with mortality (IC 95 % p < 0.05). Results: We enrolled 120 patients. 68 % were men, mean age 62 (±12.6) years. Global mortality was 34.2 %. 83.3 % were supratentorial and 16.7 % infratentorial. Mean APACHE II 14 (±6, 5) and GCS 10.4 (±4,1) and median SOFA 4 (0–14) and hematoma volume 21,35 cc3 (1–252). There were no significant differences between the two groups (infra and supratentorial) except ICH volume (p 0.000) and length of stay (LOS)-ICU (p 0.021). In the univariable analysis worse outcome with mRS was related with the volume of the hematoma (p 0,03) but not with GOS (p 0,29). Variables associated with mortality: GCS (p 0.000), APACHE II (p 0.000), GRAEB (p 0.001), SOFA (p 0.000), LOS-ICU (p 0.004) and ICH volume (p 0.000). After the multivariable analysis we determined hematoma volume was an independent risk factor for mortality (OR 1,032; 95 % CI 1,019-1,046; p 0.000). According the location we obtained a significantly association with mortality in the supratentorial group (p 0,000). We performed a ROC curve of this group and obtained an AUC 0,718 (95 % CI 0,603-0,834; p 0.000) with cutoff point of 20,2 cc3.

ependent risk factor for mortality (OR 1,032; 95 % CI 1,019-1,046; p 0.000). According the location we obtained a significantly association with mortality in the supratentorial group (p 0,000). We performed a ROC curve of this group and obtained an AUC 0,718 (95 % CI 0,603-0,834; p 0.000) with cutoff point of 20,2 cc3. Conclusions: Hematoma volume and LOS-ICU are greater in supratentorial ICH. The hematoma volume is associated with a worse outcome at ICU discharge and a supratentorial ICH volume above 20.2 cc3 is related to higher risk of mortality. References: Guidelines for the Management of Spontaneous Intracerebral Hemorrhage. AHA/ASA Guidelines. June 2015.Fig. 13 (abstract A834). ROC curve for supratentorial ICH

Conclusions: Hematoma volume and LOS-ICU are greater in supratentorial ICH. The hematoma volume is associated with a worse outcome at ICU discharge and a supratentorial ICH volume above 20.2 cc3 is related to higher risk of mortality. References: Guidelines for the Management of Spontaneous Intracerebral Hemorrhage. AHA/ASA Guidelines. June 2015.Fig. 13 (abstract A834). ROC curve for supratentorial ICH A835 Subarachnoid hemorrhage in a high volume center in Rio de Janeiro - neurological and clinical complications associated with unfavorable outcomes B. Gonçalves, R. Turon, A. Mendes, F. Miranda, P.J. Mata, D. Cavalcanti, N. Melo, P. Lacerda, P. Kurtz, C. Righy Paulo Niemeyer State Brain Institute, Rio de Janeiro, Brazil Correspondence: B. Gonçalves – Paulo Niemeyer State Brain Institute, Rio de Janeiro, Brazil Introduction: Aneurysmal subarachnoid hemorrhage (SAH) is an acute cerebrovascular event, which leads to devastating consequences, high mortality and is an important cause of neurologic disability among survivors. Incidence is reported between 2 to 16/100000 and mortality rates vary widely, ranging from 8 to 67 % among different authors. Many complications associated with SAH, such as delayed cerebral ischemia or hydrocephalus, also play a role in the poor functional outcome in survivors. Paulo Niemeyer State Brain Institute is a reference and high-volume center for SAH, located in Rio de Janeiro, Brazil, receiving patients from all over the state.

rs. Many complications associated with SAH, such as delayed cerebral ischemia or hydrocephalus, also play a role in the poor functional outcome in survivors. Paulo Niemeyer State Brain Institute is a reference and high-volume center for SAH, located in Rio de Janeiro, Brazil, receiving patients from all over the state. Objectives: The aim of the study was to describe the characteristics of patients with SAH admitted to the ICU, as part of a large prospective ongoing study, and to evaluate the factors associated with outcome. Methods: From July 2015 to March 2016, every patient admitted to the ICU with aneurysmal SAH, 18 years and older was enrolled in the study. Data were collected prospectively during hospital stay. The primary endpoint was mortality and dichotomized functional outcome, (poor outcome defined as Modified Rankin Scale 4–6) at hospital discharge.

5 to March 2016, every patient admitted to the ICU with aneurysmal SAH, 18 years and older was enrolled in the study. Data were collected prospectively during hospital stay. The primary endpoint was mortality and dichotomized functional outcome, (poor outcome defined as Modified Rankin Scale 4–6) at hospital discharge. Results: A total of 53 patients were included. The median age was 56 (47–66), 43 patients (81 %) were female. Demographic characteristics are presented in Tables 8 and 9. Twenty-nine patients (55 %) were treated by clipping, and 13 patients (25 %) were hydrocephalic and needed an EVD. An intracranial pressure monitor was inserted in 11 patients (21 %). Nine patients (17 %) developed sepsis or septic shock during ICU stay and pneumonia was present in 14 (26 %) patients. Rebleeding was diagnosed in 8 patients (15 %), vasospasm was present in 17 (32 %) patients, post-surgical deterioration was diagnosed in 16 (30 %) patients and 13 (25 %) patients developed DCI. Twenty-two (42 %) patients were mechanically ventilated. Hospital mortality was 11 % (6 patients); and 26 patients had unfavorable (49 %). In univariate analysis, factors most frequently seen in patients with unfavorable outcome were rebleeding (69 % vs 0 %, p = 0.002), vasospasm (46 % vs 19 %, p = 0.031), post-surgical neurological deterioration (46 % vs 15 %, p = 0.013), DCI (42 % vs 7 %, p = 0.003) and pneumonia (42 % vs 11 %, p = 0.01). Although not statistically significant, there was a trend towards the association between sepsis/septic shock (27 % vs 7 %, p = 0.076) and unfavorable outcome.

% vs 19 %, p = 0.031), post-surgical neurological deterioration (46 % vs 15 %, p = 0.013), DCI (42 % vs 7 %, p = 0.003) and pneumonia (42 % vs 11 %, p = 0.01). Although not statistically significant, there was a trend towards the association between sepsis/septic shock (27 % vs 7 %, p = 0.076) and unfavorable outcome. Conclusions: SAH is associated with high morbidity. Neurological complications such as rebleeding, vasospasm, post-surgical neurological deterioration and DCI, as well as clinical complications (eg. pneumonia) were associated with unfavorable outcomes. Therapeutic interventions to prevent neurological and systemic complications may have an impact on clinical outcomes.Table 8 (abstract A835). Demographic characteristics Variable All patients (n = 53) Female Gender 43 (81 %) Age 56 (47–66) Treatment Clipping 29 (55 %) Coil 22 (42 %) Mechanical Ventilation 22 (42 %) Sepsis/Septic Shock 9 (17 %) Hospital Mortality 6 (11 %) Unfavorable Outcome 26 (49 %) Table 9 (abstract A835). SAH Scores on Admisson Admission WFNS - 1 24 (45 %) 2 12 (23 %) 3 6 (11 %) 4 7 (13 %) 5 3 (6 %) Admission Modified Fisher Scale - 0 6 (11 %) 1 11 (21 %) 2 7 (13 %) 3 16 (30 %) 4 12 (23 %)

Variable All patients (n = 53) Female Gender 43 (81 %) Age 56 (47–66) Treatment Clipping 29 (55 %) Coil 22 (42 %) Mechanical Ventilation 22 (42 %) Sepsis/Septic Shock 9 (17 %) Hospital Mortality 6 (11 %) Unfavorable Outcome 26 (49 %) Table 9 (abstract A835). SAH Scores on Admisson Admission WFNS - 1 24 (45 %) 2 12 (23 %) 3 6 (11 %) 4 7 (13 %) 5 3 (6 %) Admission Modified Fisher Scale - 0 6 (11 %) 1 11 (21 %) 2 7 (13 %) 3 16 (30 %) 4 12 (23 %) A836 Patient’s course into the icu after been submitted to a CNS tumor resection L.E. de la Cruz Rosario1, S.P. Gómez Lesmes1, J.C. García Romero2, A.N. García Herrera1, E.D. Díaz Pertuz3, M.J. Gómez Sánchez1, E. Regidor Sanz1, J. Barado Hualde1, A. Ansotegui Hernández1, J.M. Guergué Irazabal1 1Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Navarra, Spain; 2Complejo Hospitalario de Navarra, Neurosurgery, Pamplona, Navarra, Spain; 3Complejo Hospitalario de Navarra, Neurology, Pamplona, Navarra, Spain Correspondence: L.E. de la Cruz Rosario – Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Navarra, Spain Introduction: The management of patient into the ICU after been submitted to a CNS resection is an important challenge. Surgery is indicated for diagnosis, to reduce tumor bulk and to manage raised intracranial pressure. Primary brain tumors are classified based on their cellular origin and histologic appearance. The most common malignant brain tumor is glioblastoma multiforme, this group have a poor prognosis.

s an important challenge. Surgery is indicated for diagnosis, to reduce tumor bulk and to manage raised intracranial pressure. Primary brain tumors are classified based on their cellular origin and histologic appearance. The most common malignant brain tumor is glioblastoma multiforme, this group have a poor prognosis. Objectives: The goal was to make a descriptive analysis about the evolution of patient submitted in the ICU for postoperative control following a surgical resection of intracranial tumors. Methods: A retrospective and observational study was conducted on all elective consecutive surgical procedures for tumor resection admitted into the ICU. We analyzed variables related with the tumor, predisposing pathology, surgical data and evolution in the ICU. We considered as an unfavorable evolution the death into the first month after the intervention or the decrease in two points or more of the Canadian´s scale score (CSS). Is a comparative study analyzed by Student´s t-test, ANOVA of one factor and Pearson´s chi-square test. Comparative study expressed by: mean difference, relative risk and confidence intervals at 95 %.

into the first month after the intervention or the decrease in two points or more of the Canadian´s scale score (CSS). Is a comparative study analyzed by Student´s t-test, ANOVA of one factor and Pearson´s chi-square test. Comparative study expressed by: mean difference, relative risk and confidence intervals at 95 %. Results: We analyzed 271 patients over of 5 years (2011–2015). Of the total, 31.3 % are high-grade gliomas, 9.6 % low grade gliomas, 32,7 % meningiomas , 8.5 % metastasis and 14.2 % other type of tumors. Average age is 54.6 years (SD 14.8), it is significantly lower in the low-grade gliomas, and in the group of other tumor types compared to other groups. 50.6 % are men , the most common in men (63.5 %) and meningiomas and other tumors in women (60.9 % and 62.5 % respectively) gliomas. 88.7 % are supratentorial location. Average size is 29.9 mL (SD 28.2) . The average score in the preoperative Karnofsky scale is 73.2 (SD 12.9) . The average income APACHE is 6.9 points (SD 4.5). An unfavorable evolution is observed in 14.0 % of patients (6.6 % per patient died and 7.4 % decline in the CSS) after one month , with no differences between different types of tumors. The percentage of deaths in the first month is higher in those undergoing surgery for metastasis (20.8 %, RR 4.0, CI 1.5 to 10.2). Mortality at two years of intervention is 31.5 %, being higher in sifnificativamente undergoing metastasis (75 %; relative risk 5.7, CI 3.6 to 9.0) and high-grade gliomas (63.6 %; relative risk 4.6, CI 3.1 to 7.4).

in the first month is higher in those undergoing surgery for metastasis (20.8 %, RR 4.0, CI 1.5 to 10.2). Mortality at two years of intervention is 31.5 %, being higher in sifnificativamente undergoing metastasis (75 %; relative risk 5.7, CI 3.6 to 9.0) and high-grade gliomas (63.6 %; relative risk 4.6, CI 3.1 to 7.4). Conclusions: Patients undergoing brain tumors have a significant risk of poor outcome , which is significantly higher in metastatic patients from the first month of intervention and in patients undergoing high-grade gliomas at two years. References 1. Outcome of elderly patients undergoing intracranial meningioma resection - a systematic review and meta-analysis.M. Tin-Chung Poon1 ,2, L. Hing-Kai Fung1 , J. K.-Suen Pu1 & G. Ka-Kit Leung1BJN, 2013; Early Online: 1–7 A837 Preventive multimodal nosocomial infection protocol in neurocritical care V. Spatenkova1, O. Bradac2, P. Suchomel3 1Regional Hospital, Neurocenter, Neurointensive Care Unit, Liberec, Czech Republic; 2Military University Hospital and First Medical School, Charles University, Department of Neurosurgery, Prague, Czech Republic; 3Regional Hospital, Neurocenter, Department of Neurosurgery, Liberec, Czech Republic Correspondence: V. Spatenkova – Regional Hospital, Neurocenter, Neurointensive Care Unit, Liberec, Czech Republic Introduction: Nosocomial infection (NI) is still an issue in neuroritical care. Objectives: We analysed NI in a preventive multimodal protocol in patients with acute brain disease.

A837 Preventive multimodal nosocomial infection protocol in neurocritical care V. Spatenkova1, O. Bradac2, P. Suchomel3 1Regional Hospital, Neurocenter, Neurointensive Care Unit, Liberec, Czech Republic; 2Military University Hospital and First Medical School, Charles University, Department of Neurosurgery, Prague, Czech Republic; 3Regional Hospital, Neurocenter, Department of Neurosurgery, Liberec, Czech Republic Correspondence: V. Spatenkova – Regional Hospital, Neurocenter, Neurointensive Care Unit, Liberec, Czech Republic Introduction: Nosocomial infection (NI) is still an issue in neuroritical care. Objectives: We analysed NI in a preventive multimodal protocol in patients with acute brain disease. Method: We performed a 10-year prospective observational cohort study in 3464 patients (pts) with acute brain disease admitted to an eight-bed adult neuro-intensive care unit (NICU). We defined our preventive multimodal protocol as: 1) keeping a hygienic and epidemiological regime including isolation of pts with multi-drug resistant bacteria 2) correct antibiotic policy, and 3) regular microbiological screening. There were 198 (5.7 %; wound 2.1 %, respiratory 1.8 %, urinary 1.0 %, bloodstream 0.6 % and other 0.2 %) pts with NI. We compared NI group pts with the control group of 3266 pts and searching predictors of NI in univariete analysis.

istant bacteria 2) correct antibiotic policy, and 3) regular microbiological screening. There were 198 (5.7 %; wound 2.1 %, respiratory 1.8 %, urinary 1.0 %, bloodstream 0.6 % and other 0.2 %) pts with NI. We compared NI group pts with the control group of 3266 pts and searching predictors of NI in univariete analysis. Results: We did not find differences in age (p = 0.416), male (p = 0.716), weight (p = 0.423) or body mass index (p = 0.966), but there were more stroke pts and fewer tumour pts (p < 0.001). NI pts stayed in NICU longer (mean 15.3 vs 4.8, p < 0.001), on admission had lower Glasgow Coma Scale (mean 11.5 vs 13.1, p < 0.001), higher Therapeutic Intervention Scoring System (TISS, p < 0.001), Acute Physiology and Chronic Health Evaluation II (p < 0.001), and CRP (p < 0.001); in the NICU they had higher CRP (p < 0.001) and NICU mortality (p < 0.001); on discharge they had worse Glasgow Outcome Scale (p < 0.001) and higher TISS sums (p < 0.001). NI pts had more accesses, which were strong predictors of NI: artery (odds ratio [OR] 3.68, 95 % CI 2.65-5.11, p < 0.001), central venous (OR 4.97, 95 % CI 3.49-7.07, p < 0.001), airways (OR 7.40, 95 % CI 5.27-10.39, p < 0.001), artificial ventilation (OR 6.74, 95 % CI 4.84-9.40, p < 0.001), urine (OR 4.23, 95 % CI 1.56-11.50, p = 0.005), operations (OR 1.65, 95 % CI 1.14-2.39, p = 0.008), drainage (OR 2.42, 95 % CI 1.71-3.42, p < 0.001). Other predictors were transfusions (OR 6.97, 95 % CI 4.62-10.50, p < 0.001), ulcer profylaxis (OR 2.12, 95 % CI 1.49-3.02, p < 0.001), wound complications (OR 7.21, 95 % CI 4.60-11.30, p < 0.001) and Methicillin-resistant Staphylococcus aureus (OR 2.90, 95 % CI 1.22-6.89, p = 0.016) to contrast Extended spectrum beta-lactamase (p = 0.227).

transfusions (OR 6.97, 95 % CI 4.62-10.50, p < 0.001), ulcer profylaxis (OR 2.12, 95 % CI 1.49-3.02, p < 0.001), wound complications (OR 7.21, 95 % CI 4.60-11.30, p < 0.001) and Methicillin-resistant Staphylococcus aureus (OR 2.90, 95 % CI 1.22-6.89, p = 0.016) to contrast Extended spectrum beta-lactamase (p = 0.227). Conclusions: Our study confirmed that nosocomial infection is associated with worse outcome and higher cost, and that accesses are still risk factors in a preventive multimodal protocol. A838 The predictive value of emergency triage codes on the outcome of aneurysmal subarachnoid hemorrhage T. Urli1, E. Heusch Lazzeri2, R. Aspide1, M. Zanello1,2 1IRCCS Istituto delle Scienze Neurologiche, Anesthesia and Intensive Care, Bellaria Hospital, Bologna, Italy; 2University of Bologna, Bologna, Italy Correspondence: R. Aspide – IRCCS Istituto delle Scienze Neurologiche, Anesthesia and Intensive Care, Bellaria Hospital, Bologna, Italy Introduction: Outcome of patients with aneurysmal subarachnoid hemorrhage (SAH) was associated in different studies with different variables (baseline illness severity, physical status, treatments, complications), but the relationship between outcome and triage assessment in the emergency setting has never been evaluated. Emergency triage in Italy is carried out with color codes: red (immediate life-saving intervention needed), yellow (urgent intervention needed), green (delayed intervention is sufficient), white (not urgent).

, but the relationship between outcome and triage assessment in the emergency setting has never been evaluated. Emergency triage in Italy is carried out with color codes: red (immediate life-saving intervention needed), yellow (urgent intervention needed), green (delayed intervention is sufficient), white (not urgent). Objectives: To study the relationship between triage severity codes assigned to patients with SAH in an Italian emergency setting and the outcome expressed as modified Rankin Score (mRS) at hospital discharge (good outcome for mRS ≤ 3, poor outcome for mRS > 3).

, but the relationship between outcome and triage assessment in the emergency setting has never been evaluated. Emergency triage in Italy is carried out with color codes: red (immediate life-saving intervention needed), yellow (urgent intervention needed), green (delayed intervention is sufficient), white (not urgent). Objectives: To study the relationship between triage severity codes assigned to patients with SAH in an Italian emergency setting and the outcome expressed as modified Rankin Score (mRS) at hospital discharge (good outcome for mRS ≤ 3, poor outcome for mRS > 3). Methods: A retrospective clinical study included 52 patients with aneurysmatic SAH admitted to emergency departments of Bologna catchment area, and then to intensive care unit (ICU), from January 2014 to January 2015. Aneurysm coiling or clipping was performed after neuroradiological diagnosis and clinical stabilization, excluding patients too ill to benefit. Intensive care treatment was carried out according to current practical guidelines. Demographic, clinical and interventional data, complications, severity scores and outcome scores were recorded. The following parameters were considered in univariate analysis: age, sex, clinical condition on arrival in the emergency department (triage code, GCS, WFSN scale, vomiting and seizures) aneurysm clipping or coiling and other neurosurgical interventions, hydrocephalus, vasospasm, cerebral infarction (CT scan), fever, sepsis, acute respiratory failure with P/F ≤ 200, cardiovascular complications (hypotension requiring vasopressor therapy, acute cardiomyopathy, arrhythmias requiring treatment); the outcome variable was modified Rankin Score >3 at hospital discharge.

ons, hydrocephalus, vasospasm, cerebral infarction (CT scan), fever, sepsis, acute respiratory failure with P/F ≤ 200, cardiovascular complications (hypotension requiring vasopressor therapy, acute cardiomyopathy, arrhythmias requiring treatment); the outcome variable was modified Rankin Score >3 at hospital discharge. Results: Poor outcome (mRS > 3) was observed in 33 % of triage green codes, 53 % of yellow codes, 75 % of red codes. The univariate analysis showed the statistically significant (p < 0.05) association with mRS > 3 for the following variables: triage red code, WFSN scale >2, acute respiratory failure, cardiovascular complications, sepsis. On logistic regression analysis, the red code assigned in the emergency department, cardiovascular complications and sepsis were associated with poor outcome. Conclusions: The severity of general clinical conditions after subarachnoid hemorrhage needing immediate life-saving intervention, feature labelled “red code” in the emergency triage, was associated with poor outcome (mRS > 3), while the other triage codes did not show any significant correlation with outcome. Cardiovascular complications and sepsis during hospital stay were other variables associated with mRS > 3.

ediate life-saving intervention, feature labelled “red code” in the emergency triage, was associated with poor outcome (mRS > 3), while the other triage codes did not show any significant correlation with outcome. Cardiovascular complications and sepsis during hospital stay were other variables associated with mRS > 3. A839 Evaluation of intracerebral hemorrhage (ICH) score in patients admitted in intensive care by supratentorial brain hemorrhage L. Perez-Borrero1, J.M. Garcia-Alvarez1, M.D. Arias-Verdu2, E. Aguilar-Alonso3, R. Rivera-Fernandez1, J. Mora-Ordoñez2, C. De La Fuente-Martos3, E. Castillo-Lorente4, F. Guerrero-Lopez5 1Hospital Serrania, Ronda, Spain; 2Hospital Regional, Intensive Care, Malaga, Spain; 3Hospital Infanta Margarita, Intensive Care, Cabra, Spain; 4Hospital Neurotraumatologico, Jaen, Spain; 5Hospital Virgen de las Nieves, Intensive Care, Granada, Spain Correspondence: L. Perez-Borrero – Hospital Serrania, Ronda, Spain Introduction: Intracerebral hemorrhage is a stroke subtype with high mortality and significant disability among survivors. Objective: To evaluate in our area the Intracerebral Hemorrage (ICH) score in patients with spontaneous supratentorial brain hemorrhage.

A839 Evaluation of intracerebral hemorrhage (ICH) score in patients admitted in intensive care by supratentorial brain hemorrhage L. Perez-Borrero1, J.M. Garcia-Alvarez1, M.D. Arias-Verdu2, E. Aguilar-Alonso3, R. Rivera-Fernandez1, J. Mora-Ordoñez2, C. De La Fuente-Martos3, E. Castillo-Lorente4, F. Guerrero-Lopez5 1Hospital Serrania, Ronda, Spain; 2Hospital Regional, Intensive Care, Malaga, Spain; 3Hospital Infanta Margarita, Intensive Care, Cabra, Spain; 4Hospital Neurotraumatologico, Jaen, Spain; 5Hospital Virgen de las Nieves, Intensive Care, Granada, Spain Correspondence: L. Perez-Borrero – Hospital Serrania, Ronda, Spain Introduction: Intracerebral hemorrhage is a stroke subtype with high mortality and significant disability among survivors. Objective: To evaluate in our area the Intracerebral Hemorrage (ICH) score in patients with spontaneous supratentorial brain hemorrhage. Methods: Multicenter prospective observational study in three hospitals in Andalusia (Spain). We studied all patients with supratentorial brain hemorrhage admitted to the Regional Hospital of Malaga (between 2006 to 2011), Neurotraumatology Hospital of Jaen (between 2010 to 2012) and Virgen de las Nieves Hospital of Granada (between 2006 to 2011). ICH score was used: Glasgow Scale (GCS) (13–15: 0, 5–12: 1: 3–4: 2 points), age (<80: 0,> 80: 1 point), volume (<30 cc : 0,> 30 cc: 1 point), infratentorial location (No: 0, Yes: 1 point), intraventricular: (No: 0 Yes: 1 point). Data are expressed as mean ± standard deviation and percentage. Student t test to compare means, the Hosmer-Lemeshow to analyze the correlation between predicted and observed mortality and the area under the ROC curve for analyzing discrimination.

location (No: 0, Yes: 1 point), intraventricular: (No: 0 Yes: 1 point). Data are expressed as mean ± standard deviation and percentage. Student t test to compare means, the Hosmer-Lemeshow to analyze the correlation between predicted and observed mortality and the area under the ROC curve for analyzing discrimination. Results: N = 263 patients. Mean age 59.74 ± 14.14 years, Glasgow score (GCS) at admission 8 ± 4 points , APACHE-II 20.7 ± 7.68 points, ICH score 2.32 ± 1.04 points, intraventricular hemorrhage (IVH) was 62 % of patients. 78 patients were treated by surgery. The hospital mortality was 53.2 %. Patients who died in hospital were older 63.76 ± 12.27 vs 55.16 ± 14.76 (p < 0.001), lower GCS 6 ± 3. vs 10 ± 4 (p < 0.001) and higher APACHE II 23.91 ± 6.61 vs 17.06 ± 7.18 (p < 0.001). Mortality at 30 days was 51.3 % and predicted by the ICH score was 45.4 %. The standardized mortality ratio (SMR) was 1.13 (0.94-1.32) (the differences are not statistically significant). The Hosmer-Lemeshow test was 39.8 (p < 0.001), so there were statistically significant differences between the observed and predicted by the ICH. Discrimination by the area under the ROC curve was 0.74 (0.68-0.80). Conclusions: Patients admitted to ICU with spontaneous supratentorial brain hemorrhage have a high mortality. The ICH score has an acceptable discrimination. The calibration is inadequate, but the differences between predicted and observed mortality are low.Table 10 (abstract A839). Mortalities according ICH score

The Hosmer-Lemeshow test was 39.8 (p < 0.001), so there were statistically significant differences between the observed and predicted by the ICH. Discrimination by the area under the ROC curve was 0.74 (0.68-0.80). Conclusions: Patients admitted to ICU with spontaneous supratentorial brain hemorrhage have a high mortality. The ICH score has an acceptable discrimination. The calibration is inadequate, but the differences between predicted and observed mortality are low.Table 10 (abstract A839). Mortalities according ICH score ICH score N Predicted mortality at 30 days (%) Observed mortality (%) 0 10 10 % 0 % 1 43 18.6 % 13 % 2 101 45.5 % 26 % 3 74 67.6 % 72 % 4 33 84.8 % 97 % 5 2 100 % 100 %

Conclusions: Patients admitted to ICU with spontaneous supratentorial brain hemorrhage have a high mortality. The ICH score has an acceptable discrimination. The calibration is inadequate, but the differences between predicted and observed mortality are low.Table 10 (abstract A839). Mortalities according ICH score ICH score N Predicted mortality at 30 days (%) Observed mortality (%) 0 10 10 % 0 % 1 43 18.6 % 13 % 2 101 45.5 % 26 % 3 74 67.6 % 72 % 4 33 84.8 % 97 % 5 2 100 % 100 % A840 Predictors of mortality of spontaneous subarachnoid hemorrhage in patients admitted to an intensive care unit of a referral hospital S.P. Gómez Lesmes1, L.E. De la Cruz Rosario1, E.D. Díaz Pertuz2, A. Ansotegui Hernández1, J.C. García Romero3, M.J. Gómez Sánchez1, A.N. García Herrera1, J. Roldán Ramírez1, E. Regidor Sanz1, J. Barado Hualde1, J.P. Tirapu León1 1Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Spain; 2Complejo Hospitalario de Navarra, Neurology, Pamplona, Spain; 3Complejo Hospitalario de Navarra, Neurosurgery, Pamplona, Spain Correspondence: S.P. Gómez Lesmes – Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Spain Introduction: Spontaneous Subarachnoid haemorrhage (SSAH) accounts for only 5 % of strokes.Aneurysms are the cause of subarachnoid haemorrhage in 85 % of cases.Rebleeding is the most imminent danger. Other complications are vasospasm, delayed cerebral ischaemia and hydrocephalus.

ntensive Care Medicine, Pamplona, Spain Introduction: Spontaneous Subarachnoid haemorrhage (SSAH) accounts for only 5 % of strokes.Aneurysms are the cause of subarachnoid haemorrhage in 85 % of cases.Rebleeding is the most imminent danger. Other complications are vasospasm, delayed cerebral ischaemia and hydrocephalus. Objectives: To know mortality predictors of spontaneous subarachnoid hemorrhage (SSAH) in the intensive care unit at the Hospital of Navarra in a period of 10 years. Methods: Analytical retrospective study was carried out of patients admitted to the ICU at the hospital of Navarra between June 2004 and June 2014, with spontaneous subarachnoid hemorrhage diagnosis. Demographic , clinical , treatments and complications variables were studied with mortality at ICU discharge. Univariate analysis (Chi square test or Fisher where appropriate and T of Student) and multivariate logistic regression were used to identify independent predictors of mortality.

arachnoid hemorrhage diagnosis. Demographic , clinical , treatments and complications variables were studied with mortality at ICU discharge. Univariate analysis (Chi square test or Fisher where appropriate and T of Student) and multivariate logistic regression were used to identify independent predictors of mortality. Results: A total of 169 patients enrolled. Mortality was 36.1 %, of which 62.3 % were women. Average stay of patients who died was 7.34 days (4.23 to 10.46 days) and alives were 13.31 days (10.45 to 16.28 days), p = 0.006. In the univariate analysis, the variables were statistically related mortality: Glasgow coma score at admission (<9 vs > = 9) OR = 5.04 (95 % CI = 2.55 to 9.98), P = < 0.001; Fisher scale (I-II vs III-IV ), OR P = 0.042; Hunt & Hess scale (I, II, III vs IV, V), OR = 5.53 (95 % CI = 2.73 to 11.19), P = < 0.001; aneurysm endovascular embolization, OR = 0.18 (95 % CI = 0.089 to 0.36), P < 0.001; treatment with intravenous nimodipine, OR = 0.15 (95 % CI = 0.07 to 0.36), P = < 0.001; vasoactive treatment, OR = 6.15 (95 % CI = 2.83 to 13.35), P < 0.001; invasive mechanical ventilation, OR = 4.76 (95 % CI = 1.05 to 21.56), P = 0.028; presence of rebleeding, OR = 2.48 (95 % CI = 1.23 to 5.01), P = 0.010; vasospasm, OR = 0.409 (95 % CI = 0.17 to 1.01), P = 0.047; hydrocephalus, OR = 0.49 (95 % CI = 0.25 to 0.94), P = 0.030: respiratory infection, OR 0.5 (95 % CI = 0.26 to 0.99), P = 0.044; intracranial hypertension, OR = 3.57 (95 % CI = 1.62 to 7.84), P = 0.001.

rebleeding, OR = 2.48 (95 % CI = 1.23 to 5.01), P = 0.010; vasospasm, OR = 0.409 (95 % CI = 0.17 to 1.01), P = 0.047; hydrocephalus, OR = 0.49 (95 % CI = 0.25 to 0.94), P = 0.030: respiratory infection, OR 0.5 (95 % CI = 0.26 to 0.99), P = 0.044; intracranial hypertension, OR = 3.57 (95 % CI = 1.62 to 7.84), P = 0.001. Predictors of mortality: age, OR 1.06 (95 % CI 1.03 to 1.10), P = 0.001; rebleeding, OR 5.51 (95 % CI 1.97 to 15.38), P = 0.001; intracranial hypertension, OR 6.20 (95 % CI = 2.06 to 18.65), P = 0.001 and Glasgow on admission (<9 vs > = 9), OR = 6.83 (95 % CI 2.70 to 17, 30), P = 0.001). Protective variables included embolization, OR 0.10 (95 % CI 0.04 to 0.26), P < 0.001 and placement of an external ventricular shunt, OR 0.36 (95 % CI = 0.14 to 0.91), P = 0.032. Conclusions: Poor prognostic factors for mortality independently were older age, coma on admission, rebleeding and development of intracranial hypertension in evolution. As protective factors for mortality include embolization and placement of an external ventricular shunt. References: 1. Subarachnoid haemorrhage. Van Gijn J, Kerr RS, Rinkel GJ. Lancet. 2007 Jan 27; 369(9558):306–18.

Conclusions: Poor prognostic factors for mortality independently were older age, coma on admission, rebleeding and development of intracranial hypertension in evolution. As protective factors for mortality include embolization and placement of an external ventricular shunt. References: 1. Subarachnoid haemorrhage. Van Gijn J, Kerr RS, Rinkel GJ. Lancet. 2007 Jan 27; 369(9558):306–18. A841 Preventable and potentially preventable traumatic mortality in a neurotraumatologic ICU L. Navarro-Guillamón, S. Cordovilla-Guardia, A. Iglesias-Santiago, F. Guerrero-López, E. Fernández-Mondéjar Complejo Hospitalario de Granada, Granada, Spain Correspondence: L. Navarro-Guillamón – Complejo Hospitalario de Granada, Granada, Spain Objective: The reduction in mortality is a basic principle of health systems. Our goal is to quantify preventable and or potentially preventable traumatic mortality and analyze the mistakes that have contributed to the result to know our room for improvement. Methods: A retrospective, descriptive study of patients admitted to the ICU from January 2013 to December 2015 and who died in the hospital (both ICU and ward). It was considered as preventable mortality what occurs as a result of a potentially avoidable diagnostic or therapeutic mistake. Potencially preventable mortality was considered when a mistake was identified but death could have occurred even without him. And unavoidable mortality when it is as a result of extremely serious injuries incompatile with survival.

y what occurs as a result of a potentially avoidable diagnostic or therapeutic mistake. Potencially preventable mortality was considered when a mistake was identified but death could have occurred even without him. And unavoidable mortality when it is as a result of extremely serious injuries incompatile with survival. Quantitative variables were described by median and interquartile range and qualitative variables as percentages. Results: 365 patients were admitted to the ICU, dying 73 of them (20 %). Focusing on the death described: Age: 69 years [47–76], ISS: 27 [25–38] males 71.2 % (52 patients). From emergency room came 67.1 % (49) and transfer from another hospital 33 % (24). Cause of Death: neurologic in 41 % (30), multiorgan failure 30 % (22). In 20 patients (27.4 %) Limitation of treatment was decided. Of the 73 dead patients, 7 of them were classified as potentially preventable deaths and one preventable death (11 %) which represents 2.19 % of total patients. The detected errors associated with preventable or potentially preventable mortality were delayed correct interpretation of signs and symptoms (5 cases), delayed implementation of treatment (4 cases), incorrect treatment (2 case) and transfer for incorrect unit (1 case) Conclusions: The management of severe trauma is not free of mistakes that can be fatal, so training in this field must be maintained continuously. However, the room for improvement in this area is limited, therefore, the main efforts should be foccused towards prevention (primary and secondary).

The detected errors associated with preventable or potentially preventable mortality were delayed correct interpretation of signs and symptoms (5 cases), delayed implementation of treatment (4 cases), incorrect treatment (2 case) and transfer for incorrect unit (1 case) Conclusions: The management of severe trauma is not free of mistakes that can be fatal, so training in this field must be maintained continuously. However, the room for improvement in this area is limited, therefore, the main efforts should be foccused towards prevention (primary and secondary). A842 Have SAH prognostic factors changed in the era of endovascular intervention? A. Vidal1, M. Perez2, A. Juez2, N. Arias1, L. Colino1, J.L. Perez1, H. Pérez1, P. Calpe1, M.A. Alcala1, D. Robaglia1, C. Perez1 1Fundación Jiménez Díaz, Madrid, Spain; 2Hospital Rey Juan Carlos, Móstoles, Spain Correspondence: A. Vidal – Fundación Jiménez Díaz, Madrid, Spain, 2Hospital Rey Juan Carlos, Móstoles, Spain Introduction: Within the clinical importance of the SAH, there are factors described in the scientific literature that speak of an unfavorable evolution of the disease. Our hypothesis is based on trying to demonstrate if only one therapeutic intervention could alter the significance of these factors. Objectives: Analyze the sociodemographic, laboratory findings, clinical and radiological factors that influence prognosis at 6 months in discharged aneurysmal SAH patients treated with endovascular intervention.

A842 Have SAH prognostic factors changed in the era of endovascular intervention? A. Vidal1, M. Perez2, A. Juez2, N. Arias1, L. Colino1, J.L. Perez1, H. Pérez1, P. Calpe1, M.A. Alcala1, D. Robaglia1, C. Perez1 1Fundación Jiménez Díaz, Madrid, Spain; 2Hospital Rey Juan Carlos, Móstoles, Spain Correspondence: A. Vidal – Fundación Jiménez Díaz, Madrid, Spain, 2Hospital Rey Juan Carlos, Móstoles, Spain Introduction: Within the clinical importance of the SAH, there are factors described in the scientific literature that speak of an unfavorable evolution of the disease. Our hypothesis is based on trying to demonstrate if only one therapeutic intervention could alter the significance of these factors. Objectives: Analyze the sociodemographic, laboratory findings, clinical and radiological factors that influence prognosis at 6 months in discharged aneurysmal SAH patients treated with endovascular intervention. Methods: We performed a retrospective longitudinal observational study of all patients who were diagnosed with an aneurysmal SAH in ICU services of 2 hospitals between March 1st 2012 and November 10th 2015. They were treated by endovascular intervention. After being discharged from ICU and after 6 months of neurologic follow-up. Patients were divided into two groups, one formed by those who presented a favorable outcome (EF) and the other by those who didn´t (ED). The variables studied were age, sex, HBP, DM, smoking and dyslipidemia. At the time of admission PO2, PCO2, leukocytosis, hyperglycemia and hypertension was determined as well as sodium, magnesium and chlorine plasma levels.

After being discharged from ICU and after 6 months of neurologic follow-up. Patients were divided into two groups, one formed by those who presented a favorable outcome (EF) and the other by those who didn´t (ED). The variables studied were age, sex, HBP, DM, smoking and dyslipidemia. At the time of admission PO2, PCO2, leukocytosis, hyperglycemia and hypertension was determined as well as sodium, magnesium and chlorine plasma levels. The clinical status of patients on admission was assessed using the Hunt-Hess and WFNS scales. The severity of SAH was determined by CT using the Fischer scale. The aneurysm was located by four vessel angiography. The time between the SAH clinic presentation and the procedure was recorded, as well as if aneurysmal occlusion was complete or not. As for the complications, we took into account the presence of fever, hydrocephalus, vasospasm and infarction. Results: For the study, 39 patients who underwent acute endovascular SAH treatment using coils, were selected. Female sex was the predominant sex 80 % Vs 75.8 % between ED and EF, respectively. The age group most frequently found was between 45 and 65 years (60 % for ED and 62 % for EF). Logistic regression analysis determined as associated with a worse outcome factor: hyperglycemia on admission(OR 8.94, 95 % CI 1.76-45.3, p = 0.007), clinical status on admission determinated by Hunt-Hess (OR 14.58 CI 2.62-81 95 %, p = 0.0018) and WFNS scales (OR 5.75, 95 % CI 1.21-27.13, p = 0.02). The presence of fever on admission also has proven to be a poor prognostic factor (OR 7.33 95 % CI 1.48-36, p = 0.01).

admission(OR 8.94, 95 % CI 1.76-45.3, p = 0.007), clinical status on admission determinated by Hunt-Hess (OR 14.58 CI 2.62-81 95 %, p = 0.0018) and WFNS scales (OR 5.75, 95 % CI 1.21-27.13, p = 0.02). The presence of fever on admission also has proven to be a poor prognostic factor (OR 7.33 95 % CI 1.48-36, p = 0.01). Conclusions: Clinical factors for aneurysmal SAH patients treated with endovascular procedure that have shown relation with the clinical outcome at six months are: poor clinical grade on admission, hyperglycemia and fever. These data are similar to those found in the literature and support the idea that the therapeutic decision (surgical or endovascular) is not the determining factor for the evolution of these patients, however, the ones mentioned above could be. A843 Increased risk of ischemic stroke in patients with venous thromboembolism: a nationwide cohort study S.-K. Lan Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Chiayi, Taiwan, Province of China Background: Conflicting results have been obtained by studies attempting to assess the risks of ischemic stroke in patients with venous thromboembolism, while the long-term risk of stroke in survivors of venous thromboembolism remains unexplored. Objective: We evaluated whether the risk of ischemic stroke in patients hospitalized with venous thromboembolism is higher when compared to the general population.

A843 Increased risk of ischemic stroke in patients with venous thromboembolism: a nationwide cohort study S.-K. Lan Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Chiayi, Taiwan, Province of China Background: Conflicting results have been obtained by studies attempting to assess the risks of ischemic stroke in patients with venous thromboembolism, while the long-term risk of stroke in survivors of venous thromboembolism remains unexplored. Objective: We evaluated whether the risk of ischemic stroke in patients hospitalized with venous thromboembolism is higher when compared to the general population. Methods: One million patients from National Health Insurance beneficiaries in Taiwan were sampled. There were 2,145 patients who had been hospitalized with diagnosis of venous thromboembolism and 727,607 unexposed subjects. All adult patients were followed from 1 January 2005 to 31 December 2013 to evaluate if ischemic stroke was diagnosed. Cox regression models were applied to compare the hazards adjusted for potential confounders.

patients who had been hospitalized with diagnosis of venous thromboembolism and 727,607 unexposed subjects. All adult patients were followed from 1 January 2005 to 31 December 2013 to evaluate if ischemic stroke was diagnosed. Cox regression models were applied to compare the hazards adjusted for potential confounders. Results: After controlling for age, gender, urbanization level, socioeconomic status, diabetes, hypertension, coronary artery disease, hyperlipidemia, history of alcohol intoxication, malignancies, congestive heart failure, atrial fibrillation, smoking, peripheral artery disease and Charlson Comorbidity Index, the adjusted hazard ratio of ischemic stroke was significantly increased in patients with venous thromboembolism (2.47; 95 % CI, 2.16-2.83). A subgroup analysis based on patients who survived longer than 12 months in the cohort also revealed higher hazard ratio in the patients with venous thromboembolism. (1.32; 95 % CI, 1.05-1.66). Conclusion: The possible risk of ischemic stroke is significantly higher in patients hospitalized with venous thromboembolism than in the general population.

Results: After controlling for age, gender, urbanization level, socioeconomic status, diabetes, hypertension, coronary artery disease, hyperlipidemia, history of alcohol intoxication, malignancies, congestive heart failure, atrial fibrillation, smoking, peripheral artery disease and Charlson Comorbidity Index, the adjusted hazard ratio of ischemic stroke was significantly increased in patients with venous thromboembolism (2.47; 95 % CI, 2.16-2.83). A subgroup analysis based on patients who survived longer than 12 months in the cohort also revealed higher hazard ratio in the patients with venous thromboembolism. (1.32; 95 % CI, 1.05-1.66). Conclusion: The possible risk of ischemic stroke is significantly higher in patients hospitalized with venous thromboembolism than in the general population. A844 Stess as a predictor of status epilepticus' outcome M.M. Cunha, T. Moreira, F. Santos, E. Lafuente, M.J. Fernandes, J.G. Silva Centro Hospitalar Tâmega e Sousa, Intensive Care, Penafiel, Portugal Correspondence: M.M. Cunha – Centro Hospitalar Tâmega e Sousa, Intensive Care, Penafiel, Portugal Introduction: Status epilepticus (SE) is a common neurological emergency with considerable associated health-care costs, morbidity and mortality.1,2 SE is defined as a prolonged seizure or multiple seizures with incomplete return to baseline.1,2 The overall mortality of SE is around 20 % with convulsive status epilepticus representing about 45-74 % of all cases.1 Status Epilepticus Severity Score (STESS) is a prognostic score that relies on four outcome measures (age, history of seizures, seizure type and extent of consciousness impairment) determined before treatment institution that ranges between 0 and 6.3

nvulsive status epilepticus representing about 45-74 % of all cases.1 Status Epilepticus Severity Score (STESS) is a prognostic score that relies on four outcome measures (age, history of seizures, seizure type and extent of consciousness impairment) determined before treatment institution that ranges between 0 and 6.3 Objective: Evaluation of STESS as a prognostic measure of functional impairment, neurologic motor deficits and 28-day mortality. Methods: Retrospective observational study of patients with SE admitted at a general intensive care unit (ICU) from 2009 to 2015. Age, gender, SAPS II/III, type of SE, length of stay, number of anti-epileptic drugs, duration of SE, functional impairment, neurologic motor deficits and 28-day mortality were collected through the ICU informatics database - PICIS®. Data is presented as mean ± SD and we used logistic regression to correlate STESS with study variables. Statistical analysis was performed using XStat 2016®.

ileptic drugs, duration of SE, functional impairment, neurologic motor deficits and 28-day mortality were collected through the ICU informatics database - PICIS®. Data is presented as mean ± SD and we used logistic regression to correlate STESS with study variables. Statistical analysis was performed using XStat 2016®. Results: Sample included 29 patients, 72,6 % male, age 46,6 ± 12,5 years, SAPS II 42,3 ± 13,1, SAPS III 55,7 ± 16,6, ICU length of stay 4,4 ± 2,6 days and hospital length of stay 16,2 ± 13,3 days. Convulsive SE represented 79,3 % of cases. STESS score`s mean was 2,6 ± 0,8. SE lasted more than 1 day in 48,3 %. Electroencephalogram was performed in 51,6 % of the patients. 48,1 % of the patients needed two or more anti-epileptic drug for SE. At hospital discharge 20,7 % had functional impairment and 13,8 % had neurologic motor deficits. Mortality was 6,8 % at 28 days. There was a correlation between STESS and mortality (OR = 10,4; ROC = 0,963), functional impairment (OR = 1,93; ROC = 0,667) and neurologic motor deficits (OR = 4,32; ROC = 0,820). The number of anti-epileptic drugs and SE duration had no correlation significance. Conclusions: We found an excellent correlation between STESS and mortality in our study. Besides this, we also found this score to be a good prognostic tool for functional impairment and neurologic motor deficits. We consider our main limitations the sample size and lower mortality. Despite we recommend using STESS as an outcome predictor. References

Conclusions: We found an excellent correlation between STESS and mortality in our study. Besides this, we also found this score to be a good prognostic tool for functional impairment and neurologic motor deficits. We consider our main limitations the sample size and lower mortality. Despite we recommend using STESS as an outcome predictor. References 1- Lancet Neurol 2015; 14: 615–24; 2 - Epilesia 2015; 56(10):1515–1523; 3 - J Neurol 2008; 255:1561–1566. A845 Prognostic factors implicated in a good functional outcome one year after suffering a subarachnoid hemorrhaged L.E. de la Cruz Rosario1, S.P. Gómez Lesmes1, A.N. García Herrera1, J.C. García Romero2, E.D. Díaz Pertuz3, M.J. Gómez Sánchez1, E. Regidor Sanz1, J.G. Armando Echeverría1, A. Ansotegui Hernández1, J. Barado Hualde1 1Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Navarra, Spain; 2Complejo Hospitalario de Navarra, Neurosurgery, Pamplona, Navarra, Spain; 3Complejo Hospitalario de Navarra, Neurology, Pamplona, Navarra, Spain Correspondence: L.E. de la Cruz Rosario – Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Navarra, Spain Introduction: Aneurysmal subarachnoid hemorrhage (SAH) is a serious condition associated with high mortality rates and long-term disability. Prediction of outcome after spontaneous subarachnoid hemorrhage (SAH) lacks accuracy. We present some factors that can be associated with a good functional prognostic after one year of the event. Objectives: Identify predictors of good functional outcome a year to present spontaneous subarachnoid hemorrhage (SSAH).

A845 Prognostic factors implicated in a good functional outcome one year after suffering a subarachnoid hemorrhaged L.E. de la Cruz Rosario1, S.P. Gómez Lesmes1, A.N. García Herrera1, J.C. García Romero2, E.D. Díaz Pertuz3, M.J. Gómez Sánchez1, E. Regidor Sanz1, J.G. Armando Echeverría1, A. Ansotegui Hernández1, J. Barado Hualde1 1Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Navarra, Spain; 2Complejo Hospitalario de Navarra, Neurosurgery, Pamplona, Navarra, Spain; 3Complejo Hospitalario de Navarra, Neurology, Pamplona, Navarra, Spain Correspondence: L.E. de la Cruz Rosario – Complejo Hospitalario de Navarra, Intensive Care Medicine, Pamplona, Navarra, Spain Introduction: Aneurysmal subarachnoid hemorrhage (SAH) is a serious condition associated with high mortality rates and long-term disability. Prediction of outcome after spontaneous subarachnoid hemorrhage (SAH) lacks accuracy. We present some factors that can be associated with a good functional prognostic after one year of the event. Objectives: Identify predictors of good functional outcome a year to present spontaneous subarachnoid hemorrhage (SSAH). Methods: observational and retrospective study. SSAH admitted to the ICU in a period of ten year were analyzed. Association of demographic, clinical, therapeutic factors and complications, with good functional outcome a one year defined this as a Glasgow Outcome Scale (GOS) between 4 and 5. Poor prognosis was defined as GOS within 2 and 3.Statistics analysis by Chi square (Fisher test when appropriate) and Student T. The independent variables were analyzed by multivariate logistic regression.

plications, with good functional outcome a one year defined this as a Glasgow Outcome Scale (GOS) between 4 and 5. Poor prognosis was defined as GOS within 2 and 3.Statistics analysis by Chi square (Fisher test when appropriate) and Student T. The independent variables were analyzed by multivariate logistic regression. Results: 97 patients survived after one year, 39.1 % with good functional prognosis. The mean ICU stay was 8.3 days (SD 11.8), mean age 52.4 years (SD 12.6) and poor prognosis 21 days (SD 18,2). Without diabetes mellitus (DM) 98.5 % (OR 0.080 (CI95% 0.009-0.72), P 0.012) and blood pressure (HBP) 74.2 % (OR 0.42 (CI95% 0.172-1 03), P 0.05). Hunt and Hess (I, II and III) and 72.7 % (IV and V) 27.3 % (OR 0.35 (CI95% 0.15-0.86), P 0.02); Fisher (I and II) and 18.2 % (III and IV) 81.8 %, GCS (3–8) and 18.2 % (9–15) 81.8 %. Embolized within 72 hours 83.3 %; external ventricular drainage device (EVD) 24.6 % (OR 0.27 (95 % CI 0.109-0.665), P 0.004); 9.1 % angioplasty; 1.5 % craniectomy (OR 0.104 (95 % CI from 0.01-0.97, P 0.018) Nimodipine perfusion in 89.4 % and 12.1 % during intra-arterial embolization; 40.9 % vasoactive treatment (OR 0,44 (95 % CI 0.18-0.05), P 0.061) and mechanical ventilation 81.8 %; 13.2 % complicated with re-bleeding, vasospasm 25.8 %, 39.4 % hydrocephalus (OR 0.36 (CI95% 0.15- 0.86, P 0.021), 4.5 % intracranial hypertension (OR 0.16 (CI95% 0.04-0.68), P 0.011); 7.6 % stroke (OR 0.2 (CI95% 0.061- 0.66), P 0.01), 6.1 % ventriculitis (OR 0.19 (CI95% 0.05-0.68 ), P 0.016) and 31.8 % respiratory infection (OR 0.34 (95 % CI 0.14-0.81), P 0.014). Using logistic regression, the following variables results but no one of these was statistical significant: DM (OR 0.54 (CI95% 0.005-0.634), P 0.02), hypertension (OR 0.32 (CI95% 0.1-1.0), P 0.05), EVD (OR 0.26 (CI95% 0.07-0.92), P 0.04), craniectomy (OR 0.03 (CI95% 0.003-0.34), P 0.005) and length of stay (OR 0.95 (IC95% 0.90 to 0.99), P 0.02).

the following variables results but no one of these was statistical significant: DM (OR 0.54 (CI95% 0.005-0.634), P 0.02), hypertension (OR 0.32 (CI95% 0.1-1.0), P 0.05), EVD (OR 0.26 (CI95% 0.07-0.92), P 0.04), craniectomy (OR 0.03 (CI95% 0.003-0.34), P 0.005) and length of stay (OR 0.95 (IC95% 0.90 to 0.99), P 0.02). Conclusions: The presence of comorbidities and the intervention by craniectomy and EVD significantly increase the likelihood of functional impairment annually. References 1. Poor outcome is associated with less negative fluid balance in patients with aneurysmal subarachnoid hemorrhage treated with prophylactic vasopressor-induced hypertension.Sakr Y1,2, Dünisch P3, Santos C4, Ann Intensive Care. 2016 Dec;6(1):25 Grant acknowledgement None A846 Prognostic possibility of neurological evaluation scale in patients after fossa posterior surgery V. Podlepich, E. Sokolova, E. Alexandrova, K. Lapteva Burdenko Neurosurgery Institute, Moscow, Russian Federation Correspondence: V. Podlepich – Burdenko Neurosurgery Institute, Moscow, Russian Federation Introduction: In order to determine optimal airway protection measures in early postoperative period after fossa posterior surgery (PFS), it is important to carry out a prognosis of neurological dynamics based on the preoperative neurological exam. We have designed Neurological Evaluation Scale (NES). Objectives: Our study was aimed at determining the potential of NES to predict brain stem deterioration in early postoperative period after PFS based on the assessment of the preoperative neurological status.

A846 Prognostic possibility of neurological evaluation scale in patients after fossa posterior surgery V. Podlepich, E. Sokolova, E. Alexandrova, K. Lapteva Burdenko Neurosurgery Institute, Moscow, Russian Federation Correspondence: V. Podlepich – Burdenko Neurosurgery Institute, Moscow, Russian Federation Introduction: In order to determine optimal airway protection measures in early postoperative period after fossa posterior surgery (PFS), it is important to carry out a prognosis of neurological dynamics based on the preoperative neurological exam. We have designed Neurological Evaluation Scale (NES). Objectives: Our study was aimed at determining the potential of NES to predict brain stem deterioration in early postoperative period after PFS based on the assessment of the preoperative neurological status. Methods: The prospective study was carried out during the period from December 2013 to June 2014 and included 182 patients operated for fossa posterior tumors (FPT). To be included in the study, patients had to be over 18 years old and operated for fossa posterior non infiltrative paraxial tumors. We examined all patients before and after the operation, immediately after the extubation in ICU. NES provided complex neurological assessment with an emphasis on the brain stem function. Postoperative NES points were subtracted from the preoperative points - AB-criterion (ABc). Positive ABc corresponds to intensification of neurological deterioration. Negative or zero ABc corresponds to neurological improvement. All neurological symptoms were grouped in 7 NES blocks according to their relation to CNS.

Postoperative NES points were subtracted from the preoperative points - AB-criterion (ABc). Positive ABc corresponds to intensification of neurological deterioration. Negative or zero ABc corresponds to neurological improvement. All neurological symptoms were grouped in 7 NES blocks according to their relation to CNS. Results: We divided all patients in two groups depending on their ABc, which revealed that the patients with positive ABc had reliability less NES points before operation compared to the patients with negative or zero ABc. We found out the frequency of occurrence of each NES block for inclusion in the full neurologic status. We discovered that caudal stem affection occurred more frequently in the patients with more NES points. We evaluated the probability of neurological impairment or regression of neurological symptoms depending on ABc with sensitivity 90,7 % and specificity 95 %. We created a prognostic model, which could predict the discharge from clinic outcome on the basis of the NES blocks points assigned during the early postoperative period. Conclusions: We revealed neurological features of postoperative period in patient after FPS. Our data could predict neurological outcomes, and be useful in optimization tactic of airway protection. References 1. Matsui T. Therapeutic strategy and long-term outcome of meningiomas located in the posterior cranial fossa. Neurol Med Chir (Tokyo). 2012;52(10):704–13.

Conclusions: We revealed neurological features of postoperative period in patient after FPS. Our data could predict neurological outcomes, and be useful in optimization tactic of airway protection. References 1. Matsui T. Therapeutic strategy and long-term outcome of meningiomas located in the posterior cranial fossa. Neurol Med Chir (Tokyo). 2012;52(10):704–13. 2. Adachi K, Kazuhide Y. ABC Surgical Risk Scale for skull base meningioma: a new scoring system for predicting the extent of tumor removal and neurological outcome. J Neurosurg. 2009 Nov;111(5):1053–61 3. Wijdicks EF , Bamlet WR, Maramattom BV, Manno EM, McClelland RL. Validation of a new coma scale: The FOUR score. Ann Neurol. 2005 Oct; 58(4):585–93. 4. Harner SG, Laws ER Jr.Clinical findings in patients with acoustic neurinoma. Mayo Clin Proc. 1983 Nov;58(11):721–8. A847 Brain dysfunction in severe sepsis: an observational study P. Kurtz1,2, C. Shuinotsuka3, L. Rabello3, G. Vianna3, A. Reis3, C. Cairus3, J. Salluh3, F. Bozza3 1Fiocruz, Rio de Janeiro, Brazil; 2Paulo Niemeyer State Brain Institute, Rio de Janeiro, Brazil; 3D’Or Institute for Research and Education, Rio de Janeiro, Brazil Correspondence: P. Kurtz – Fiocruz, Rio de Janeiro, Brazil Introduction: Sepsis associated encephalopathy is a predictor of increased mortality and cognitive impairment in intensive care unit (ICU) patients. Electroencephalogram (EEG) abnormalities are common in the acute stages of sepsis and may correlate with the severity of systemic and brain dysfunction.

de Janeiro, Brazil Introduction: Sepsis associated encephalopathy is a predictor of increased mortality and cognitive impairment in intensive care unit (ICU) patients. Electroencephalogram (EEG) abnormalities are common in the acute stages of sepsis and may correlate with the severity of systemic and brain dysfunction. Objectives: The objective of this study was to establish the relationship between continuous EEG abnormalities, organ failure and mortality in patients with severe sepsis and septic shock. Methods: This is a prospective, multi center, observational study. Continuous EEG for 12 hours was performed in all consecutive patients acutely admitted in ICU for severe sepsis and sepstic shock from community-acquired infection, within 72 hours from admission. EEG was analyzed for its baseline rhythm, presence of epileptiform activity and according to Synek's classification. Quantitative analysis through compressed spectral array (CSA) was also applied. Neurological examination, including assessment of delirium, coma, brain stem reflexes and level of sedation were assessed daily for 14 days. Multiple logistic regression was used to analyze the relations between physiologic variables, EEG findings, delirium and mortality.

mpressed spectral array (CSA) was also applied. Neurological examination, including assessment of delirium, coma, brain stem reflexes and level of sedation were assessed daily for 14 days. Multiple logistic regression was used to analyze the relations between physiologic variables, EEG findings, delirium and mortality. Results: Nineteen patients were included, mean age was 79 (47–86) years and 11 (58 %) were female. Continuos EEG was performed for a duration of 12 (10–12) hours. Six patients (32 %) developed septic shock and 4 (21 %) were mechanically ventilated. Overall, 4 patients (21 %) developed delirium and 3 (16 %) patients died in the ICU. EEG recordings showed background slowing (theta or delta frerquency range) in 8 (42 %) patients. Absence of EEG reactivity was observed in 2 patients (11 %), periodic epileptiform discharges (PEDs) in 2 (11 %) and no electrographic seizures were recorded. Moreover Synek's score ≥3 was found in 6 (32 %) patients. Organ failure and derranged systemic physiologic parameters were more frequent in patients with altered EEG. In patients that developed delirium, EEG with a Synek's score ≥3 was more frequent than in those that did not (75 % vs 14 %, P = 0.04). Also among nonsurvivors Synek's score ≥3 was more common than among survivors (100 % vs 19 %, P = 0.02). Conclusions: In this pilot prospective study of 19 patients with severe sepsis or septic shock, continuous EEG with a Synek score ≥3 was more common in patients that developed delirium and among non survivors.

In patients that developed delirium, EEG with a Synek's score ≥3 was more frequent than in those that did not (75 % vs 14 %, P = 0.04). Also among nonsurvivors Synek's score ≥3 was more common than among survivors (100 % vs 19 %, P = 0.02). Conclusions: In this pilot prospective study of 19 patients with severe sepsis or septic shock, continuous EEG with a Synek score ≥3 was more common in patients that developed delirium and among non survivors. CONTEMPORARY ISSUES IN INFECTION AND SEPSIS II A848 Infections related to external ventricular drainage in critically ill patients J.C. Barrios Torres1, N.J. Fernández Araujo1, P. García-Olivares1, E. Keough1, M. Dalorzo1, L.K. Tang1, I. De Sousa1, M. Díaz1, L.J. Marcos-Zambrano2, J.E. Guerrero1 1Hospital Gregorio Marañon, Intensive Care Unit, Madrid, Spain; 2Hospital Gregorio Marañon, Microbiology, Madrid, Spain Correspondence: J.C. Barrios Torres – Hospital Gregorio Marañon, Intensive Care Unit, Madrid, Spain Introduction: External ventricular drains (EVDs) are used for the management of elevations in intracranial pressure secondary to acute hydrocephalus primarily caused by subarachnoid (SAH), intracerebral (ICH) or intraventricular hemorrhage (IVH) or traumatic brain injury (1). A serious complication of external cerebrospinal fluid (CSF) drainage is colonization of the catheter and posterior retrograde infection resulting in ventriculomeningitis. Rates of EVD associated ventriculitis range from 0 to 22 %(1,2,3).

erebral (ICH) or intraventricular hemorrhage (IVH) or traumatic brain injury (1). A serious complication of external cerebrospinal fluid (CSF) drainage is colonization of the catheter and posterior retrograde infection resulting in ventriculomeningitis. Rates of EVD associated ventriculitis range from 0 to 22 %(1,2,3). Objective: To assess the incidence and risk factors for EVD-related infections in critically ill patients with acute cerebral pathology. Methods: Retrospective study of patients admitted to the Intensive Care Unit (ICU) (2014–2015) requiring placement of an EVD due to acute cerebral pathology. Epidemiological and clinical data were collected, as well as data related to the EVD: insertion site, samples of CSF and microbiological isolates. EVD-related infection was defined as the presence of systemic signs of infection associated with an inflammatory CSF. Descriptive analysis was expressed as a mean (standard deviation) or median (interquartile range) for quantitative variables and percentages for qualitative. The association of factors related to EVD infections was performed using simple logistic regression, with the corresponding Odds Ratio.

with an inflammatory CSF. Descriptive analysis was expressed as a mean (standard deviation) or median (interquartile range) for quantitative variables and percentages for qualitative. The association of factors related to EVD infections was performed using simple logistic regression, with the corresponding Odds Ratio. Results: 30 patients. 56 % males. Age 55 ± 14 years. Charlson Index: 1 (IQR 0–4).Illness severity scores on admission: APACHE II 17 ± 5; SOFA 5 ± 3; GCS 7 (IQR 3–11). Reason for admission: SAH 60 %, ICH 28 % and ischaemic stroke 4 %. IVH was seen in 88 % of cases. EVD was inserted in the ICU in 88 % of patients, 10 % required replacement, lasting 17 ± 12 days. Two samples of CSF per patient (IQR 1–4) were obtained. Inflammatory CSF seen in 56 % of patients, with 35.7 % having a microbiological isolate (20 % of total), the most frequent being GRAM positive bacteria (62 % S. epidermidis). Elapsed time from placement of the EVD to infection was 11 ± 6 days. ICU stay 23 days (IQR 15–30), 16.2 % mortality. Length of hospital stay 52 days (IQR 35–77), 22.5 % mortality. A poor functional status (GOS 1–3) at discharge was found in 46 %. In the univariate analysis, factors associated with EVD-related infection were IVH (RR 2.75, 95 % CI 1.58-4.78), EVD placement in ICU (RR 2.75, 95 % CI 1.58-4.78) and EVD days (OR 1.10, 95 % CI 1.01-1.22). There was non significant trend towards an unfavorable GOS (OR 2.50, 95 % CI 0.8-16.42), increased ICU stay (18 ± 12 vs 29 ± 14 days, OR 1.03, 95 % CI 0.94-1.12) and longer hospital stay (44 ± 23 vs 82 ± 72, OR 1.02, 95 % CI 0.98-1.06).

ment in ICU (RR 2.75, 95 % CI 1.58-4.78) and EVD days (OR 1.10, 95 % CI 1.01-1.22). There was non significant trend towards an unfavorable GOS (OR 2.50, 95 % CI 0.8-16.42), increased ICU stay (18 ± 12 vs 29 ± 14 days, OR 1.03, 95 % CI 0.94-1.12) and longer hospital stay (44 ± 23 vs 82 ± 72, OR 1.02, 95 % CI 0.98-1.06). Conclusions: Our study showed a high incidence of EVD-related infection. Factors associated with its development include IVH, insertion outside the operating theatre and EVD placement days. References 1. Hagel et al. Interdiscip Perspect Infect Dis 2014; 2014: 708531. 2. Beer et al. Neurocrit Care 2009; 10: 363–7. 3. Lozier et al. Neurosurgery 2002; 51: 170–8 A849 Clinical and ventilatory differences between adult critical patients survivors and non survivors with moderate and severe ARDS due to 2009 AH1 N1 influenza virus S.E. Zamora Gomez, G.D. Hernandez Lopez, A.I. Vazquez Cuellar, O.R. Perez Nieto, J.A. Castanon Gonzalez Hospital Juarez de México, ICU, Mexico City, Mexico Correspondence: S.E. Zamora Gomez – Hospital Juarez de México, ICU, Mexico City, Mexico Introduction: The global rate mortality due to 2009 AH1 N1 influenza virus in critical care patients has not been well estimated, México is considered as the origin of this pandemic disease, and some reports of the CDC (Centers for Disease Control and Prevention) suggest the presentation in this country is usually more severe than in others, with a higher risk of death.

Objectives: The objective of this study was to determine the clinical and ventilatory differences betwen patients survivors and non survivors with diagnostic of moderate and severe ARDS due to confirmed 2009 AH1 N1 influenza virus, hospitalized in ICU at a mexican tertiary center. Methods: Observational, retrospective study including adult patients with moderate and severe ARDS, due to RT-PCR confirmed 2009 AH1 N1 influenza virus, hospitalized in the ICU during march 2009 to march 2016. Patients were analyzed studied in two groups:survivors and non survivors. Results: 27 patients were analyzed, 14 patients with moderate and 13 with severe ARDS,12 females and 15 males; mortality was of 52 %,13 survivors (sv) and 14 non survivors (nsv). The mean APACHE II score was 26 points (23 pts in sv vs 29 pts in nsv),mean SOFA score of 12 points (11 pts in sv vs13 pts in nsv). The median age was 46.07 years, with a median of 49.07 yrs in sv vs 43.28 yrs in nsv. The median body-mass index was 33.3,29.4 in sv vs 37 in nsv.The median of stay in ICU was 21.25 d, with a median of duration of mechanical ventilation of 18.44 d,23.81 d in sv vs 13.07 in nsv. Tracheostomy was performed in 19 patients (70 %),12 in the sv, 7 in the nsv.Mean PaO2/FiO2 ratio at entry in ICU was 91.5,100.1 in sv vs 82.3 in nsv. Prone position was performed in 20 patients (74 %), with a mean duration of 6.6 d, 7.99 d in sv vs 5.33 d in nsv. The mean highest PEEP was 14.35 cmH2O with a mean of 12 cm of H2O in sv vs 16.7 cm of H2O in nsv. 12 patients (44.4 %) received CRRT,3 (25 %) in the sv vs. 9 (75 %) in the nsv group.

1 in sv vs 82.3 in nsv. Prone position was performed in 20 patients (74 %), with a mean duration of 6.6 d, 7.99 d in sv vs 5.33 d in nsv. The mean highest PEEP was 14.35 cmH2O with a mean of 12 cm of H2O in sv vs 16.7 cm of H2O in nsv. 12 patients (44.4 %) received CRRT,3 (25 %) in the sv vs. 9 (75 %) in the nsv group. Conclusions: In this study the main differences in the group of non survivors patients in comparisson with survivors were a higher body-mass index, higher scores in mortality prediction scales, a minor age, a lower PaO2/FiO2 ratio at entry in ICU, the need of higher levels of PEEP, and a higher need of CRRT as reflect of progression to MODS. References 1.- Writing committee of the WHO consultation on clinical aspects of pandemic 2009 influenza.clinical aspects of pandemic 2009 influenza A(H1 N1) virus infection.N Engl J Med 2010;362:1708–19 2.- Chowell G, et al.Severe respiratory disease concurrent with the circulation of H1 N1 influenza. N Engl J Med 2009;361:674–9.

References 1.- Writing committee of the WHO consultation on clinical aspects of pandemic 2009 influenza.clinical aspects of pandemic 2009 influenza A(H1 N1) virus infection.N Engl J Med 2010;362:1708–19 2.- Chowell G, et al.Severe respiratory disease concurrent with the circulation of H1 N1 influenza. N Engl J Med 2009;361:674–9. A850 H1 N1 pneumonia related ARDS: encouraging results with early corticosteroids D. Bhasin1, S. Rai1, H. Singh1, O. Gupta1, M.K. Bhattal1, S. Sampley1, K. Sekhri2, R. Nandha2 1Max Superspeciality Hospital, Critical Care, Pulmonary and Sleep Medicine, Mohali, India; 2Dr. H.S.J Institute of Dental Sciences, Pharmacology, Chandigarh, India Correspondence: D. Bhasin – Max Superspeciality Hospital, Critical Care, Pulmonary and Sleep Medicine, Mohali, India Introduction: The emergence of H1 N1 pneumonia leading on to ARDS with its catastrophic course changed the very outlook of how to approach viral pneumonia in adults. The window of time period is very limited in which to act and rapidly deteriorating oxygen saturation despite high PEEP and Fio2 requirement necessitates expedient measures to halt the fast deteriorating process. In this scenario we attempted to use corticosteroids and got excellent results. Objectives: A minority of the patients who contract H1 N1 develops a rapidly progressive pneumonia leading to acute respiratory distress syndrome (ARDS). Role of corticosteroids is controversial in treatment of H1 N1 pneumonia related ARDS. The present study was undertaken to establish the use of corticosteroids in shaping the course of H1 N1 related ARDS.

who contract H1 N1 develops a rapidly progressive pneumonia leading to acute respiratory distress syndrome (ARDS). Role of corticosteroids is controversial in treatment of H1 N1 pneumonia related ARDS. The present study was undertaken to establish the use of corticosteroids in shaping the course of H1 N1 related ARDS. Methods: A retrospective observational study (May 2012 - March 2016) was conducted in a Medical Intensive Care Unit of a tertiary care hospital in Northern India. Total of 24 confirmed cases of H1 N1 pneumonia with ARDS were enrolled. The H1 N1 pneumonia was confirmed with real time reverse transcriptase polymerase chain reaction assay. The response to the administration of prolonged corticosteroid therapy (Methylprednisolone 1 mg/kg/day in a continuous infusion) along with oseltamivir (150 mg twice a day) was studied. Severity of illness and Multi Organ Dysfunction were quantified with Acute physiology and chronic health evaluation II (APACHE II) score at 24 hours of admission, Sequential organ failure assessment (SOFA) and Lung injury score (LIS) on day 1 & day 7 of hospitalisation. Results: On an average after one week of treatment, patients with confirmed H1 N1 pneumonia related moderate to severe ARDS, who required non invasive or invasive ventilation experienced a significant improvement in lung injury score. In 16 of 24 patients, Lung Injury Score was measured to be 3.35 ± 0.42 at the time of admission and was decreased to 1.8 ± 0.9 on day 7 of hospitalisation. In hospital mortality was found to be only 13 % ( 20 out of 23 patients survived)

r invasive ventilation experienced a significant improvement in lung injury score. In 16 of 24 patients, Lung Injury Score was measured to be 3.35 ± 0.42 at the time of admission and was decreased to 1.8 ± 0.9 on day 7 of hospitalisation. In hospital mortality was found to be only 13 % ( 20 out of 23 patients survived) Conclusions: In Patients of H1 N1 with confirmed ARDS, administration of early & prolonged low dose corticosteroids showed significant improvement in lung function with prevention of fibrosis and permanent pulmonary damage. It was also associated with significant low in hospital mortality. References 1. An expanded definition of the adult respiratory distress syndrome. Murray JF, Matthay MA, Luce JM, Flick MR. Am Rev Respir Dis. 1988 Sep;138(3):720–3. 2. Acute Respiratory Distress Syndrome: The Berlin Definition JAMA. 2012;307(23):2526–2533. 3. H1 N1 influenza A virus-associated acute lung injury: response to combination oseltamivir and prolonged corticosteroid treatment. Quispe-Laime, A. M.; Bracco, J. D.; Barberio, C. G.; Rolfo, V. E.; Umberger, R.; Meduri, G. U. J ICM 2010 Vol. 36 No. 1 pp. 33–41

2. Acute Respiratory Distress Syndrome: The Berlin Definition JAMA. 2012;307(23):2526–2533. 3. H1 N1 influenza A virus-associated acute lung injury: response to combination oseltamivir and prolonged corticosteroid treatment. Quispe-Laime, A. M.; Bracco, J. D.; Barberio, C. G.; Rolfo, V. E.; Umberger, R.; Meduri, G. U. J ICM 2010 Vol. 36 No. 1 pp. 33–41 A851 Viral induce acute respiratory distress syndrome (ARDS): characterization of patients in a tertiary hospital in South America F.A. Aliaga1,2, F. Olivares2, F. Appiani2, P. Farias2, F. Alberto2, A. Hernández2 1Instituto Nacional del Tórax, Fellow Respiratory Medicine, Santiago, Chile; 2Hospital Militar de Santiago, Intensive Care, Santiago, Chile Correspondence: F.A. Aliaga – Instituto Nacional del Tórax, Fellow Respiratory Medicine, Santiago, Chile Introduction: ARDS is an important cause of admission to intensive care unit and account a mortality of 40 % in it moderate form. The objective of this study is characterization of viral ARDS patients admitted to our intensive care unit Methods: Descriptive, retrospective study of patients admitted to a critical care unit in a Military Hospital of Santiago Chile with viral ARDS during year 2015. All patient have positive polymerase chain reaction (PCR) to a viral etiology. We analyze demographic, clinics, oxygenation, ventilator settings, prone positioning, neuromuscular blocking, 28 days mortality, intensive care stay and etiology study. Statistical analysis, measures of central tendency, percentages and chi square test.

have positive polymerase chain reaction (PCR) to a viral etiology. We analyze demographic, clinics, oxygenation, ventilator settings, prone positioning, neuromuscular blocking, 28 days mortality, intensive care stay and etiology study. Statistical analysis, measures of central tendency, percentages and chi square test. Results: We identify 15 patients. The median age 67 years, 25 % female and 80 % with comorbidities. 66 % need invasive mechanical ventilation (IMV) and 34 % non invasive. Mean PaO2/FIo2 138, with APACHE score of 17. Mean tidal volume 6.2 ml/kg with plateau pressure of 26.7 and driving pressure 13.8 %. 90 % use neuromuscular blockers and 60 % prone positioning ventilation. The mean time of IMV 8 days y ICU stay 18 days. In viral PCR 53 % positive for influenza, 3 adenovirus and 2 rhinovirus. The mortality at 28 days 26.6 %. We observe a statistic association between white cell count over 20.000 and 28 days mortality (p 0.012). Conclusions: We found less mortality than reported in LUNG SAFE study for moderate ARDS and observe and association between mortality and high white cell.

Results: We identify 15 patients. The median age 67 years, 25 % female and 80 % with comorbidities. 66 % need invasive mechanical ventilation (IMV) and 34 % non invasive. Mean PaO2/FIo2 138, with APACHE score of 17. Mean tidal volume 6.2 ml/kg with plateau pressure of 26.7 and driving pressure 13.8 %. 90 % use neuromuscular blockers and 60 % prone positioning ventilation. The mean time of IMV 8 days y ICU stay 18 days. In viral PCR 53 % positive for influenza, 3 adenovirus and 2 rhinovirus. The mortality at 28 days 26.6 %. We observe a statistic association between white cell count over 20.000 and 28 days mortality (p 0.012). Conclusions: We found less mortality than reported in LUNG SAFE study for moderate ARDS and observe and association between mortality and high white cell. A852 Early bacterial infectious complications following heart transplantation S. Pons1, R. Sonneville1, L. Bouadma1, M. Neuville1, E. Mariotte1, A. Radjou1, J. Lebut1, S. Chemam1, G. Voiriot1, M.-P. Dilly2, B. Mourvillier1, R. Dorent3, P. Nataf3, M. Wolff1, J.-F. Timsit1 1Bichat University Hospital, Medical ICU, Paris, France; 2Bichat University Hospital, Anaesthesiology, Paris, France; 3Bichat University Hospital, Cardiovascular surgery, Paris, France Correspondence: S. Pons – Bichat University Hospital, Medical ICU, Paris, France Introduction: National priority heart transplantation for severe acute heart failure has increased. The impact of early infectious complications has never been assessed in a large cohort of ICU patients.

rdiovascular surgery, Paris, France Correspondence: S. Pons – Bichat University Hospital, Medical ICU, Paris, France Introduction: National priority heart transplantation for severe acute heart failure has increased. The impact of early infectious complications has never been assessed in a large cohort of ICU patients. Objectives: We aimed to determine the characteristics, the determinants and the impact of infectious complications after heart transplantation. Methods: We retrospectively studied all consecutive heart-transplant recipients in Bichat Hospital, between January 1st, 2011 and June 6th, 2015. All infectious complications that occurred within six months after transplantation were considered for analysis. The primary endpoint was the rate of bacterial complications at 6 months. We used a multivariate logistic regression to identify risk factors for bacterial infections. Data are presented as median (IQR) or number (percentage).

lications that occurred within six months after transplantation were considered for analysis. The primary endpoint was the rate of bacterial complications at 6 months. We used a multivariate logistic regression to identify risk factors for bacterial infections. Data are presented as median (IQR) or number (percentage). Results: 113 patients (53 years (40–62), male n = 86 (75 %)) were included. At time of heart transplantation, 65 patients (57 %) were hospitalized in ICU with acute heart failure, 28 (25 %) were under extracorporeal membrane oxygenation (ECMO) support. Length of stay in ICU after transplantation was 16 [11–24] days. Seventeen (15 %) died in the ICU and 19 (17 %) died within 6 months. CMV mismatch was present in 15 (13 %) recipients. Twenty-two (19 %) patients had a multidrug-resistant bacteria (MRB) carriage before transplantation, and 6 (5,3 %) acquired one during the first week in ICU. Nine of those (32 %) developed at least one infection due to the same MRB during their stay in ICU.

in 6 months. CMV mismatch was present in 15 (13 %) recipients. Twenty-two (19 %) patients had a multidrug-resistant bacteria (MRB) carriage before transplantation, and 6 (5,3 %) acquired one during the first week in ICU. Nine of those (32 %) developed at least one infection due to the same MRB during their stay in ICU. Seventy-one (63 %) patients developed one or more bacterial infection within six months post-transplantation. The first one occurred in ICU for 66 (93 %) patients. Sixty-five pneumonia were diagnosed in ICU, among them 46 (70 %) before day 8 post-transplantation. Bloodstream (n = 26) and Scarpa infections (n = 17) were as frequent in the early post-operative period than after day 8. During the first post-operative week, most of bacteria were Gram negative bacteria (n = 51), including 6 ESBL and 7 Pseudomonas aeruginosa. After this period, infections in ICU were also predominantly due to Gram negative bacteria (n = 39) with higher rates of P. aeruginosa (n = 12) and ESBL (n = 10). Among Gram positive bacteria (n = 25), no Methicillin-resistant Staphylococcus aureus was found. Within 6 months after transplantation, 50 patients (44 %) developed a viral complication (36 CMV reactivations). Fungal infection was found in 16 patients (14 %), including 9 invasive aspergillosis.

Seventy-one (63 %) patients developed one or more bacterial infection within six months post-transplantation. The first one occurred in ICU for 66 (93 %) patients. Sixty-five pneumonia were diagnosed in ICU, among them 46 (70 %) before day 8 post-transplantation. Bloodstream (n = 26) and Scarpa infections (n = 17) were as frequent in the early post-operative period than after day 8. During the first post-operative week, most of bacteria were Gram negative bacteria (n = 51), including 6 ESBL and 7 Pseudomonas aeruginosa. After this period, infections in ICU were also predominantly due to Gram negative bacteria (n = 39) with higher rates of P. aeruginosa (n = 12) and ESBL (n = 10). Among Gram positive bacteria (n = 25), no Methicillin-resistant Staphylococcus aureus was found. Within 6 months after transplantation, 50 patients (44 %) developed a viral complication (36 CMV reactivations). Fungal infection was found in 16 patients (14 %), including 9 invasive aspergillosis. Bacterial infection in ICU was associated with a longer stay in ICU (20 days (13–29) vs. 12 days (8–17) p < 0,001), but not with increased mortality. After adjustment, ECMO following heart transplantation was identified as the only risk factor for bacterial infections in ICU (Odds ratio = 3.1 (1.4-6.9), p = 0.006). Conclusions: This study confirms the high rate of early infectious complications after heart transplantation. ECMO after transplantation is an independent factor associated with bacterial infections in ICU.

Bacterial infection in ICU was associated with a longer stay in ICU (20 days (13–29) vs. 12 days (8–17) p < 0,001), but not with increased mortality. After adjustment, ECMO following heart transplantation was identified as the only risk factor for bacterial infections in ICU (Odds ratio = 3.1 (1.4-6.9), p = 0.006). Conclusions: This study confirms the high rate of early infectious complications after heart transplantation. ECMO after transplantation is an independent factor associated with bacterial infections in ICU. A853 Characteristics of viral pneumonia patients admitted to the ICU due to hypoxemic respiratory failure O. Ediboglu, S. Ataman, H. Ozkarakas, C. Kirakli Dr. Suat Seren Chest Diseases and Surgery Training Hospital, Intensive Care Unit, Izmir, Turkey Correspondence: O. Ediboglu – Dr. Suat Seren Chest Diseases and Surgery Training Hospital, Intensive Care Unit, Izmir, Turkey Introduction: Hypoxemic respiratory failure due to viral pneumonia during epidemic seasons sometimes may require mechanical ventilation support and ICU stay. Mortality of these patients can be high despite young age (1,2). Objectives: To evaluate the characteristics and risk factors of viral pneumonia patients admitted to our ICU due to hypoxemic respiratory failure.

A853 Characteristics of viral pneumonia patients admitted to the ICU due to hypoxemic respiratory failure O. Ediboglu, S. Ataman, H. Ozkarakas, C. Kirakli Dr. Suat Seren Chest Diseases and Surgery Training Hospital, Intensive Care Unit, Izmir, Turkey Correspondence: O. Ediboglu – Dr. Suat Seren Chest Diseases and Surgery Training Hospital, Intensive Care Unit, Izmir, Turkey Introduction: Hypoxemic respiratory failure due to viral pneumonia during epidemic seasons sometimes may require mechanical ventilation support and ICU stay. Mortality of these patients can be high despite young age (1,2). Objectives: To evaluate the characteristics and risk factors of viral pneumonia patients admitted to our ICU due to hypoxemic respiratory failure. Methods: Patients admitted to our ICU between December 2015 and March 2016 who had hypoxemic respiratory failure due to clinically and radiologically suspected viral pneumonia were enrolled. Demographic characteristics, risk factors, need and type of mechanical ventilatory support, laboratory values, need of extracorporeal pulmonary (ECMO) and renal support (CRRT) and mortality rate were evaluated. Data are expressed as median (25th -75th percentile) or n (%) and compared using Mann Whitney U test or Chi-square where appropriate.

ics, risk factors, need and type of mechanical ventilatory support, laboratory values, need of extracorporeal pulmonary (ECMO) and renal support (CRRT) and mortality rate were evaluated. Data are expressed as median (25th -75th percentile) or n (%) and compared using Mann Whitney U test or Chi-square where appropriate. Results: Twenty two patients (11 male) met the enrollment criteria. Median age and APACHE II score was 45 (36–63) years and 19 (13–25) respectively. Time from the onset of symptoms to admission to the ICU was 7 (6–9) days. Most common symptoms were shortness of breath, cough and fever respectively. Risk factors were present at 8 (36 %) patients. Maximum set PEEP levels during mechanical ventilation was significantly lower in patients who survived [10 (8–10) vs 13 (10–16), p = 0.025)]. Deceased patients spent more time under a PaO2/FiO2 ratio below 100 [72 (24–90) vs. 0 (0–48) hours, p = 0.024]. Serum albumin levels were also lower in deceased patients [2.6 (2–3.1) vs. 3.5 (2.6-3.7) g/dl, p = 0.033]. Two patients needed ECMO support and 5 patients underwent CRRT due to acute renal failure. Survival rate was 88 % (7/8) in patients who had noninvasive ventilation (NIV) success while it was only 7 % (1/14) in patients who had undergone invasive mechanical ventilation (p < 0.001). Overall mortality was 64 %.

= 0.033]. Two patients needed ECMO support and 5 patients underwent CRRT due to acute renal failure. Survival rate was 88 % (7/8) in patients who had noninvasive ventilation (NIV) success while it was only 7 % (1/14) in patients who had undergone invasive mechanical ventilation (p < 0.001). Overall mortality was 64 %. Conclusions: Viral pneumonia may result in severe hypoxemic respiratory failure and ARDS especially during epidemic seasons. Mortality of these patients can be relatively high despite full support including ECMO and CRRT. NIV success, time spent under a PaO2/FiO2 ratio below 100 and low serum albumin levels at admission may be predictors of severity of the disease and mortality. References 1. Crotty MP, Meyers S, Hampton N, Bledsoe S, Ritchie DJ, Buller RS, Storch GA, Micek ST, Kollef MH. Epidemiology, Co-Infections, and Outcomes of Viral Pneumonia in Adults: An Observational Cohort Study. Medicine (Baltimore). 2015Dec; 94(50):e2332. 2. Kirakli C, Tatar D, Cimen P, Edipoglu O, Coskun M, Celikten E, Ozsoz A. Survival from severe pandemic H1 N1 in urban and rural Turkey: a case series. Respir Care. 2011 Jun;56(6):790–5.

1. Crotty MP, Meyers S, Hampton N, Bledsoe S, Ritchie DJ, Buller RS, Storch GA, Micek ST, Kollef MH. Epidemiology, Co-Infections, and Outcomes of Viral Pneumonia in Adults: An Observational Cohort Study. Medicine (Baltimore). 2015Dec; 94(50):e2332. 2. Kirakli C, Tatar D, Cimen P, Edipoglu O, Coskun M, Celikten E, Ozsoz A. Survival from severe pandemic H1 N1 in urban and rural Turkey: a case series. Respir Care. 2011 Jun;56(6):790–5. A854 Impact of FFP transfusion on percentage of ventilated patients developed VAP in ICU patients A. Vakalos, V. Avramidis Xanthi General Hospital, ICU, Xanthi, Greece Correspondence: A. Vakalos – Xanthi General Hospital, ICU, Xanthi, Greece Introduction: While plasma donation is still necessary as a unique source of human proteins and to treat coagulation disorders, FFP administration seems to have high rate of inappropriate indication. After all, FFP transfusion is not risk free, and is associated with lung injury, infectious disease, circulatory overload and immunosuppression in recipients. On the other hand, Ventilator Associated Pneumonia (VAP) is one of the most frequently seen infections in ICU setting and may have an impact not only to the length of ICU stay, but o ICU outcome as well. Objectives: The aim of our observation retrospective study was to test the hypothesis that a correlation exists between FFP transfusion and the percentage of ventilated patients developed VAP (% VP) in our both medical and surgical ICU served in community hospital.

A854 Impact of FFP transfusion on percentage of ventilated patients developed VAP in ICU patients A. Vakalos, V. Avramidis Xanthi General Hospital, ICU, Xanthi, Greece Correspondence: A. Vakalos – Xanthi General Hospital, ICU, Xanthi, Greece Introduction: While plasma donation is still necessary as a unique source of human proteins and to treat coagulation disorders, FFP administration seems to have high rate of inappropriate indication. After all, FFP transfusion is not risk free, and is associated with lung injury, infectious disease, circulatory overload and immunosuppression in recipients. On the other hand, Ventilator Associated Pneumonia (VAP) is one of the most frequently seen infections in ICU setting and may have an impact not only to the length of ICU stay, but o ICU outcome as well. Objectives: The aim of our observation retrospective study was to test the hypothesis that a correlation exists between FFP transfusion and the percentage of ventilated patients developed VAP (% VP) in our both medical and surgical ICU served in community hospital. Methods: From January 2006 to June 2014 admitted to our ICU 620 patients, mean age 64.8 years, mean length of ICU stay (LOS) 14.2 days, mean mechanical ventilation duration per ventilated patient (V. Days) 12.23 days, mean APACHE II score on admission 21.2, predicted mortality 38.9 %, actual mortality 31.45 %, Standardized Mortality Ratio (SMR) 0.80. From our database we looked for the percentage of ventilated patients developed VAP and the following values and indexes according FFP transfusion per year from 2006 to 2014 (mean values). Total, per patient, per hospitalization days (HD), per patient under mechanical ventilation (pts V) and per ventilation days (VD) Using linear correlation method, we looked for linear slope, correlation coefficient (r), and coefficient of determination (r2), and by linear regression method using ANOVA test we looked for p value, according % VP and FFP transfusion.

HD), per patient under mechanical ventilation (pts V) and per ventilation days (VD) Using linear correlation method, we looked for linear slope, correlation coefficient (r), and coefficient of determination (r2), and by linear regression method using ANOVA test we looked for p value, according % VP and FFP transfusion. Results: Conclusions: According to our data, there was no statistically significant correlation detected between the percentages of ventilated patients developed VAP and FFP transfusion indexes. Our data suggest that even though FFP transfusion may have an impact on immunosuppression and infection disease developing, the impact on the percentages of ventilated patients developed VAP is not statistically significant.Table 11 (abstract A854). Results FFP Slope r r2 St. Error L. CI U. CI p Value Total Tran −1.916 −0.210 0.044 3.369 −9.885 6.052 0.5873 Trans per pt −0.084 −0.544 0.296 0.048 −0.199 0.031 0.1294 Trans per H.D. −0.005 −0.450 0.184 0.004 −0.014 0.004 0.5480 Trans per Pt V. −0.084 −0.540 0.291 0.051 −0.210 0.034 0.1334 Trans per V. D. −0.008 −0.518 0.268 0.005 −0.020 0.003 0.1529

L. CI U. CI p Value Total Tran −1.916 −0.210 0.044 3.369 −9.885 6.052 0.5873 Trans per pt −0.084 −0.544 0.296 0.048 −0.199 0.031 0.1294 Trans per H.D. −0.005 −0.450 0.184 0.004 −0.014 0.004 0.5480 Trans per Pt V. −0.084 −0.540 0.291 0.051 −0.210 0.034 0.1334 Trans per V. D. −0.008 −0.518 0.268 0.005 −0.020 0.003 0.1529 A855 Sensitivity of biomarkers of sepsis in patients receiving anticancer therapy O. Obukhova1, I.A. Kurmukov1, S. Kashiya1, E. Golovnya2, V.N. Baikova2, T. Ageeva3, T. Haritydi3, E.V. Kulaga4 1N.N. Blokhin Russian Cancer Research Center, Medical ICU, Moscow, Russian Federation; 2N.N. Blokhin Russian Cancer Research Center, Express Laboratory, Moscow, Russian Federation; 3N.N. Blokhin Russian Cancer Research Center, Biochemical Laboratory, Moscow, Russian Federation; 4N.N. Blokhin Russian Cancer Research Center, Microbiology Department, Moscow, Russian Federation Correspondence: O. Obukhova – N.N. Blokhin Russian Cancer Research Center, Medical ICU, Moscow, Russian Federation Introduction: In patients receiving anti-cancer therapy, clinical and laboratory signs of treatment-related toxicity may coincide with the sepsis criteria, which makes early diagnosis difficult. Defining laboratory "markers of sepsis" - C-reactive protein (CRP), procalcitonin (PCT) and presepsin (PRS) is considered an additional method of diagnostics. Objectives: To estimate the sensitivity of laboratory markers of sepsis in patients receiving anticancer therapy.

A855 Sensitivity of biomarkers of sepsis in patients receiving anticancer therapy O. Obukhova1, I.A. Kurmukov1, S. Kashiya1, E. Golovnya2, V.N. Baikova2, T. Ageeva3, T. Haritydi3, E.V. Kulaga4 1N.N. Blokhin Russian Cancer Research Center, Medical ICU, Moscow, Russian Federation; 2N.N. Blokhin Russian Cancer Research Center, Express Laboratory, Moscow, Russian Federation; 3N.N. Blokhin Russian Cancer Research Center, Biochemical Laboratory, Moscow, Russian Federation; 4N.N. Blokhin Russian Cancer Research Center, Microbiology Department, Moscow, Russian Federation Correspondence: O. Obukhova – N.N. Blokhin Russian Cancer Research Center, Medical ICU, Moscow, Russian Federation Introduction: In patients receiving anti-cancer therapy, clinical and laboratory signs of treatment-related toxicity may coincide with the sepsis criteria, which makes early diagnosis difficult. Defining laboratory "markers of sepsis" - C-reactive protein (CRP), procalcitonin (PCT) and presepsin (PRS) is considered an additional method of diagnostics. Objectives: To estimate the sensitivity of laboratory markers of sepsis in patients receiving anticancer therapy. Methods: A total of 56 patients with a reliably diagnosed sepsis (SIRS + defined source of infection ± identification of the infectious agent) were examined. Blood samples for the determination of sepsis markers were carried out within the first hour of ICU hospitalization. CRP and PCT concentration in the serum was determined in all patients and PRC concentration - in 18 patients. The results for each marker are represented as Me (min-max). The calculation of 95 % confidence intervals (95%CI) was carried out by means of the Wilson method. For the diagnosis of sepsis, marker cut-off levels values exceeding 50 mg/l for CRP, 2 ng/ml for PCT and 600 pg/ml for PRS were taken into account.

8 patients. The results for each marker are represented as Me (min-max). The calculation of 95 % confidence intervals (95%CI) was carried out by means of the Wilson method. For the diagnosis of sepsis, marker cut-off levels values exceeding 50 mg/l for CRP, 2 ng/ml for PCT and 600 pg/ml for PRS were taken into account. Results: The medians of CRP, PCT and PRS were 102 mg/l (0.7 to 455.9); 4.15 ng/ml (0.5 to 200) and 789.5 pg/ml (343 to 6440), respectively. The sensitivity of CRP, PCT and PRS as positive markers of sepsis were 0.83 (95%CI 0.71-0.90); 0.61 (95 % CI 0.48-0.72); 0.39 (95%CI 0.20-0.61), respectively. Conclusions: CPR was the most sensitive sepsis marker in patients receiving anticancer therapy. In the examined population, values PCT and PRS within first few hours from sepsis development showed relatively low sensitivity. A856 Serum biomarkers and scores prognostic in severe sepsis/septic shock J.J. Rios-Toro1, L. Perez-Borrero1, E. Aguilar-Alonso2, M.D. Arias-Verdu3, J.M. Garcia-Alvarez1, C. Lopez-Caler3, C. De La Fuente-Martos2, S. Rodriguez-Fernandez1, M. Gomez Sanchez-Orézzoli1, F. Martin-Gallardo1 1Hospital Serrania, Ronda, Spain, 2Hospital Infanta Margarita, Intensive Care, Cabra, Spain, 3Hospital Regional, Intensive Care, Malaga, Spain Correspondence: L. Perez-Borrero – Hospital Serrania, Ronda, Spain Introduction: Mortality is elevated in patients with severe sepsis and, even more, in those with septic shock. Early diagnosis and appropriate treatment improve the survival in them.

Care, Cabra, Spain, 3Hospital Regional, Intensive Care, Malaga, Spain Correspondence: L. Perez-Borrero – Hospital Serrania, Ronda, Spain Introduction: Mortality is elevated in patients with severe sepsis and, even more, in those with septic shock. Early diagnosis and appropriate treatment improve the survival in them. Objective: To explore the prognostic value of procalcitonin (PCT) and C-reactive protein (CRP) serum levels for patients with severe sepsis and/or septic shock in a intensive care unit (ICU) and compare it with scores prognostic. Methods: 50 patients admitted at the intensive care unit (ICU) with the diagnosis of severe sepsis or septic shock were studied. SOFA and APACHE II scores as well as serum were measured at days 1,2 and 5. The influence of these variables on 28-day mortality was analyzed. 20 healthy individuals served as controls. Results: The sample is composed of 50 patients with severe sepsis and/or septic shock with an mean age 63.7 ± 14.25 years and APACHE II on admission of 19.14 ± 7.7 points. SOFA score of 7.62 ± 3.8 points. Mortality was 42 %. Patients who died on admission showed no statistically significant differences in APACHE-II 19.27 ± 6.08 vs 19.10 ± 8.16 points and SOFA score 8.55 ± 5.59 vs 7.36 ± 3.87 points. The inflammatory markers commonly analyzed had lower values of CRP 166 ± 100 vs 223 ± 117 mg/dL (p = 0.081) and PCT 10 ± 21 vs 24 ± 36.68 ng/mL (p = 0.144) but without being statistically significant differences.

ferences in APACHE-II 19.27 ± 6.08 vs 19.10 ± 8.16 points and SOFA score 8.55 ± 5.59 vs 7.36 ± 3.87 points. The inflammatory markers commonly analyzed had lower values of CRP 166 ± 100 vs 223 ± 117 mg/dL (p = 0.081) and PCT 10 ± 21 vs 24 ± 36.68 ng/mL (p = 0.144) but without being statistically significant differences. The changes between admission and the first day and its relation with mortality was analyzed, the inflammatory markers compared to surviving patients showed a decrease in CRP , −4.81 ± 55.97 vs −0.19 ± 73.48 (p = 0.891 ) and PCT −15.20 ± 22.65 vs 0.6 ± 14.76 (p = 0.745), and these decreases were not statistically significant. Clinical severity indicators show a statistically significant decline in the APACHE II between the second day and income −1.56 ± 4.72 vs −0.98 ± 5.13 (p < 0.001) and SOFA score 0.29 ± 2.55 vs 0.45 ± 3.70 (P < 0.001). The discrimination of changes regarding mortality was analyzed with the area for APACHE II changes of 0.771 (0.62-0.93) and SOFA evolution 0.68 (0.5-0.85). And the area for CRP changes and PCT was only 0.52 and 0.51. Multivariate analysis with logistic regression showed that mortality was statistically significant associated with the SOFA change wih OR: 1.76 (1.05-2.95) and without statistically significant relation with PCT, CRP and APACHE II changes. Conclusions: In septic patients admitted to the ICU, the improving of APACHE and SOFA score are more sensitive markers of survival than the evolution of inflammatory parameters more commonly used as CRP and PCT.

Multivariate analysis with logistic regression showed that mortality was statistically significant associated with the SOFA change wih OR: 1.76 (1.05-2.95) and without statistically significant relation with PCT, CRP and APACHE II changes. Conclusions: In septic patients admitted to the ICU, the improving of APACHE and SOFA score are more sensitive markers of survival than the evolution of inflammatory parameters more commonly used as CRP and PCT. A857 Correlates of perfusion markers with base line mortality prediction in icu? An elusive search for the best performer…… J. Nikhilesh1, V. Joshi2 1CHL Hospitals, Dept of Critical Care Services, Indore, India; 2Shalby Hospital, Dept of Critical Care Services, Indore, India Correspondence: J. Nikhilesh –CHL Hospitals, Dept of Critical Care Services, Indore, India Background: There have been various perfusion markers that have been evaluated in literature to implement strategies to optimize management in subsets of sepsis. However, a definitive answer to the same remains debatable. Objective: To determine performance of makers of tissue perfusion such as lactate, base excess and pCO2 gap (central venous to arterial carbon dioxide gap) at baseline in patients of sepsis and correlate them with predicting mortality in these subsets. Setting: Multidiscliplinary 60 bedded medical surgical ICUs of two tertiary care hospitals.

Objective: To determine performance of makers of tissue perfusion such as lactate, base excess and pCO2 gap (central venous to arterial carbon dioxide gap) at baseline in patients of sepsis and correlate them with predicting mortality in these subsets. Setting: Multidiscliplinary 60 bedded medical surgical ICUs of two tertiary care hospitals. Study Module: Consecutive patients presenting with sepsis getting admitted to ICU were included for a duration dated Jan 2015-Mar 2016 and data was collected at baseline with a focus on demographics, SOFA scores, Lactate, Base excess pCO2 gap, Length of stay (LOS) in days and mortality. Death/discharge from ICU were considered as end points. Patients with COPD/prior ILD, Hepatic failure and out of hospital cardiac arrests were excluded. ROC curves for association with mortality were constructed using SPSS version 22 and logistic regression analysis was done for the same data. Results: Fifty five patients were included (n = 55). Male: Female ratio was (M: F-37:18) with age 42.5 ± 12.3 years (range-22-69).The baseline lactate, Base excess and pCO2 gaps were 9.41 ± 3.43 (range 5–19), 7.23 ± 2.02 (range-4-13) and 6.21 ± 1.66 (range- 3.9-9.6) respectively. SOFA scores at baseline were 13.7 ± 2.08 (range-10-18).LOS was 4.32 ± 1.98 (range-1-9 days).In hospital mortality was 38.2 % (n = 21).ROC curves were drawn and AUC for lactate, Base excess and pCO2 gap was 0.74, 0.028 and 0.169 respectively.

5–19), 7.23 ± 2.02 (range-4-13) and 6.21 ± 1.66 (range- 3.9-9.6) respectively. SOFA scores at baseline were 13.7 ± 2.08 (range-10-18).LOS was 4.32 ± 1.98 (range-1-9 days).In hospital mortality was 38.2 % (n = 21).ROC curves were drawn and AUC for lactate, Base excess and pCO2 gap was 0.74, 0.028 and 0.169 respectively. Conclusion: In subsets of sepsis lactate levels score over base excess and pCO2 gap in terms of sensitivity in predicting mortality however base excess and pCO2 gap are more specific vis-a-vis lactate levels in prediction of mortality with base excess being the more specific marker. We need to review it further in terms of differentiating these subsets and validating this data in subsets of specific settings of multidiscliplinary critical care. Grant acknowledgement There are no grants for this project for all the authors.

Conclusion: In subsets of sepsis lactate levels score over base excess and pCO2 gap in terms of sensitivity in predicting mortality however base excess and pCO2 gap are more specific vis-a-vis lactate levels in prediction of mortality with base excess being the more specific marker. We need to review it further in terms of differentiating these subsets and validating this data in subsets of specific settings of multidiscliplinary critical care. Grant acknowledgement There are no grants for this project for all the authors. A858 Behavior of IgM antiendotoxin-core antibodies in septic shock. A preliminary study E. Villarreal1, J. Ruiz2, M. Gordon2, A. Quinza2, J. Gimenez2, M. Piñol2, A. Castellanos2, P. Ramirez2 1Hospital la Fe, Intensive Care Unit, Valencia, Spain; 2Hospital la Fe, Valencia, Spain Correspondence: E. Villarreal – Hospital la Fe, Intensive Care Unit, Valencia, Spain Introduction: The bacterial lipopolysaccharide (LPS) or endotoxin is a major component of the outer membrane of all Gram-negative bacteria. LPS plays an important role in the normal physiology, permeability and fluidity of the bacterial outer cell wall. Endotoxaemia is associated with infections caused by Gram-negative bacteria, but can occurs in absence of it. In sepsis, some of the protective mechanisms of gut barrier are disrupted and subsequently, bacterial products, like endotoxin, may cross the gut mucosa and spread to the mesenteric lymph nodes or more distant organs. This phenomenon is known as bacterial translocation (BT).

e bacteria, but can occurs in absence of it. In sepsis, some of the protective mechanisms of gut barrier are disrupted and subsequently, bacterial products, like endotoxin, may cross the gut mucosa and spread to the mesenteric lymph nodes or more distant organs. This phenomenon is known as bacterial translocation (BT). Objectives: Evaluate IgM antiendotoxin-core antibodies (IgM EndoCAb) and inflammatory biomarkers (C-reactive protein, procalcitonin and interleuquin-6), in septic shock. Methods: Prospective observational study of consecutive septic shock patients admitted to our intensive care unit (ICU). Patients were followed during the first three days after SS. Inflammatory biomarkers (CRP, PCT, IL-6), and IgM EndoCAb were measured daily. Clinical, hemodynamic and microbiological data were collected. Results: We included 25 patients. Urinary and abdominal were the infectious disease more diagnosed (32 %). Gram negative bacteria (GNB) was isolated in 40 % (n 12) and Gram-positive bacteria (GPB) in 32 % (n 9). Patients with infection caused by Gram-negative bacteria, showed lower IgM EndoCAb levels all days without statistical significance (p = 0.13). IgM EndoCAb were lower in patients with shock no resolved (4.9 ± 2; 5.7 ± 3 and ± 2 vs 4.8 ± 24.3 41; 28.3 ± 51 and 22.3 ± 41 respectively) and those who died (5.9 ± 3; 6.5 ± 2 and ± 5.6 25 ± 2 vs 42; 29.4 ± 52 and 23.1 ± 43 respectively) but difference was not significant.

Patients with infection caused by Gram-negative bacteria, showed lower IgM EndoCAb levels all days without statistical significance (p = 0.13). IgM EndoCAb were lower in patients with shock no resolved (4.9 ± 2; 5.7 ± 3 and ± 2 vs 4.8 ± 24.3 41; 28.3 ± 51 and 22.3 ± 41 respectively) and those who died (5.9 ± 3; 6.5 ± 2 and ± 5.6 25 ± 2 vs 42; 29.4 ± 52 and 23.1 ± 43 respectively) but difference was not significant. An inverse correlation between inflammatory biomarkers (PCT, CRP and IL-6) and IgM EndoCAb was detected. IL6 showed a higher correlation, but without statistically significant differences. ICU mortality rate was 8 %. Conclusion: IgM EndoCAb were detected in septic shock caused by GPB, it could be explain by a bacterial translocation. Patients with major endotoxaemia have higher consumption of antibodies and therefore lower levels of IgM EndoCAb that is associated with a worse prognosis.

An inverse correlation between inflammatory biomarkers (PCT, CRP and IL-6) and IgM EndoCAb was detected. IL6 showed a higher correlation, but without statistically significant differences. ICU mortality rate was 8 %. Conclusion: IgM EndoCAb were detected in septic shock caused by GPB, it could be explain by a bacterial translocation. Patients with major endotoxaemia have higher consumption of antibodies and therefore lower levels of IgM EndoCAb that is associated with a worse prognosis. A859 The relationship between the neutrophil/lymphocyte ratio and mortality in the severe sepsis patients Y.D. Jeon1, W.Y. Jeong2, M.H. Kim2, I.Y. Jeong2,3, M.Y. Ahn2, J.Y. Ahn2, S.H. Han2,3, J.Y. Choi2,3, Y.G. Song2,3, J.M. Kim2,3, N.S. Ku2,3 1Yonsei Wonju University College of Medicine, Department of Internal Medicine, Wonju, Republic of Korea; 2Yonsei University College of Medicine, Department of Internal Medicine, Seoul, Republic of Korea; 3Yonsei University College of Medicine, AIDS Research Institute, Seoul, Republic of Korea Correspondence: Y.D. Jeon – Yonsei Wonju University College of Medicine, Department of Internal Medicine, Wonju, Republic of Korea Introduction: The neutrophil/lymphocyte ratio (NLR) is known as the prognostic marker in the various malignancies and cardiovascular disease. Several studies reported about the relationship between NLR and mortality in the severe sepsis patients.1–3 But these studies have shown conflicting results. Objectives: So, we investigated the relationship between NLR and 28-day mortality in the severe sepsis patients.

A859 The relationship between the neutrophil/lymphocyte ratio and mortality in the severe sepsis patients Y.D. Jeon1, W.Y. Jeong2, M.H. Kim2, I.Y. Jeong2,3, M.Y. Ahn2, J.Y. Ahn2, S.H. Han2,3, J.Y. Choi2,3, Y.G. Song2,3, J.M. Kim2,3, N.S. Ku2,3 1Yonsei Wonju University College of Medicine, Department of Internal Medicine, Wonju, Republic of Korea; 2Yonsei University College of Medicine, Department of Internal Medicine, Seoul, Republic of Korea; 3Yonsei University College of Medicine, AIDS Research Institute, Seoul, Republic of Korea Correspondence: Y.D. Jeon – Yonsei Wonju University College of Medicine, Department of Internal Medicine, Wonju, Republic of Korea Introduction: The neutrophil/lymphocyte ratio (NLR) is known as the prognostic marker in the various malignancies and cardiovascular disease. Several studies reported about the relationship between NLR and mortality in the severe sepsis patients.1–3 But these studies have shown conflicting results. Objectives: So, we investigated the relationship between NLR and 28-day mortality in the severe sepsis patients. Methods: The electronic medical records of patients enrolled in Early Goal Directed Therapy (EGDT) between January 2010 and December 2011 at Severance Hospital, a 2000-bed tertiary care hospital in Seoul, Korea, were retrospectively reviewed. Analysis was performed to access the relationship between NLR and 28-day mortality.

s: The electronic medical records of patients enrolled in Early Goal Directed Therapy (EGDT) between January 2010 and December 2011 at Severance Hospital, a 2000-bed tertiary care hospital in Seoul, Korea, were retrospectively reviewed. Analysis was performed to access the relationship between NLR and 28-day mortality. Results: Total 198 patients of EGDT were enrolled in the study. 102 (51.5 %) patients were male and the mean age was 63.30 years. The primary focus of sepsis was the pneumonia in 56 (28.3 %), urinary tract infection in 56 (28.3 %), intra-abdominal infection in 23 (11.6 %) and gastroenteritis in 15 (7.6 %). The median NLR was 16.23 (interquartile range, 7.37 to 27.76) and the overall 28-day mortality rate was 11.1 %. NLR (p = 0.975) was not significantly associated with 28-day mortality. On mutlivariate analysis, SOFA score (p < 0.001) was only independent factor associated with mortality. Conclusions: In our study, NLR was not associated with 28-day mortality in the severe sepsis patients. References 1. Akilli NB, Yortanli M, Mutlu H, et al. Prognostic importance of neutrophil-lymphocyte ratio in critically ill patients: short- and long-term outcomes. The American journal of emergency medicine. 2014;32(12):1476–1480. 2. Riche F, Gayat E, Barthelemy R, Le Dorze M, Mateo J, Payen D. Reversal of neutrophil-to-lymphocyte count ratio in early versus late death from septic shock. Critical care (London, England). 2015;19:439.

1. Akilli NB, Yortanli M, Mutlu H, et al. Prognostic importance of neutrophil-lymphocyte ratio in critically ill patients: short- and long-term outcomes. The American journal of emergency medicine. 2014;32(12):1476–1480. 2. Riche F, Gayat E, Barthelemy R, Le Dorze M, Mateo J, Payen D. Reversal of neutrophil-to-lymphocyte count ratio in early versus late death from septic shock. Critical care (London, England). 2015;19:439. 3. Salciccioli JD, Marshall DC, Pimentel MA, et al. The association between the neutrophil-to-lymphocyte ratio and mortality in critical illness: an observational cohort study. Critical care (London, England). 2015;19:13. A860 Evaluating initiatives advocating appropriate fluid therapy in severe sepsis/septic shock according to the surviving sepsis campaign H. Shah, F. Kellner, F. Rezai, N. Mistry, P. Yodice, V. Ovnanian, K. Fless, E. Handler Saint Barnabas Medical Center, Department of Medicine, Livingston, United States Correspondence: H. Shah – Saint Barnabas Medical Center, Department of Medicine, Livingston, United States Introduction: Even under ideal scenarios and maximized medical treatment, severe sepsis/septic shock are amongst the leading causes of morbidity and mortality in hospitalized patients worldwide. The Surviving Sepsis Campaign guidelines emphasize early aggressive intravenous fluids (IVF) resuscitation and antibiotics administration in these patients to improve outcomes.

zed medical treatment, severe sepsis/septic shock are amongst the leading causes of morbidity and mortality in hospitalized patients worldwide. The Surviving Sepsis Campaign guidelines emphasize early aggressive intravenous fluids (IVF) resuscitation and antibiotics administration in these patients to improve outcomes. Objectives: Our objective is to assess compliance with dosage and timing of intravenous fluid therapy in severe sepsis/septic shock patients and the effect on hospital length of stay (LOS) after education of emergency department (ED) staff of the recommended treatment protocol in the surviving sepsis guidelines (30 mL/Kg within one hour of diagnosis).

ssess compliance with dosage and timing of intravenous fluid therapy in severe sepsis/septic shock patients and the effect on hospital length of stay (LOS) after education of emergency department (ED) staff of the recommended treatment protocol in the surviving sepsis guidelines (30 mL/Kg within one hour of diagnosis). Methods: Our 633-bed single center retrospective study evaluated patients 21 years of age older with severe sepsis/septic shock admitted to the intensive care unit (ICU) from the ED. We looked at two time frames, before (April 1 to December 31, 2012) and after (July 1 to December 21, 2014) an educational intervention from the ED pharmacist. Specifically, the ED pharmacist held educational group sessions along with one-on-one education with ED medical personnel. Collected data included patient demographics, blood pressures, lactic acid (LA), dosage and timing of IV fluids and hospital LOS. Patients who qualified for this intervention had least one of the following: LA greater than 4 mm/L, systolic blood pressure (SBP) less than 90 mm/Hg, or SBP lower than 40 mmHg from baseline. Data was collected to determine whether the educational intervention would affect patients receiving adequate treatment using Chi-square testing and if there was an effect on mean length of stay using student-t testing.

n 4 mm/L, systolic blood pressure (SBP) less than 90 mm/Hg, or SBP lower than 40 mmHg from baseline. Data was collected to determine whether the educational intervention would affect patients receiving adequate treatment using Chi-square testing and if there was an effect on mean length of stay using student-t testing. Results: Pre-intervention, forty five patients with severe sepsis/septic shock were admitted to the ICU from the ED. Forty-two qualified for the study. Only one received adequate fluid resuscitation. Post-education, 41 patients were admitted and thirty-six qualified for the study. Of these eighteen received adequate IVF resuscitation, yielding a P value of < 0.000001 and a significant difference following the educational intervention. However, there was not a significant decrease in mean LOS from 14.3 to 11.8 days between the two groups, with a p-value of 0.276, when excluding patients who expired or were discharged to hospice. Conclusions: In our study, there was a significant increase in patients with severe sepsis/septic shock receiving adequate IVF therapy after an extensive educational intervention was performed. Though there was a decrease in mean hospital LOS, it was not calculated to be significant. As patients in both populations frequently had additional comorbidities along with social and financial considerations affecting their hospital discharge, confounding factors affecting their LOS were likely to be present.Fig. 14 (abstract A860). Effect of education on appropriate treatment & LOS.

ted to be significant. As patients in both populations frequently had additional comorbidities along with social and financial considerations affecting their hospital discharge, confounding factors affecting their LOS were likely to be present.Fig. 14 (abstract A860). Effect of education on appropriate treatment & LOS. SPOTLIGHT ON ICU N&AHP PRACTICE A861 Efficacity of mechanical insufflation-exsufflation devices in intubates and ventilated ICU patients R. Martínez Alejos1, J.D. Martí Romeu2, D. González Antón1, A. Quinart1, A. Torres Martí2 1CHU Bordeaux, Bordeaux, France; 2Hospital Clínic, Barcelona, Spain Correspondence: R. Martínez Alejos – CHU Bordeaux, Bordeaux, France Introduction: Critically ill patients on invasive mechanical ventilation (IMV) often present retention of respiratory secretions, increasing the risk for morbidity. Endotracheal suctioning (ETS) is the main strategy to prevent mucus retention, but its effects are limited to the first bronchial bifurcation. Mechanical in-exsufflation (MI-E) is a non-invasive chest physiotherapy (CPT) technique aimed to generate high expiratory flows to simulate cough, and improve mucus clearance in the proximal airways. To date, no studies have assessed the effects of MI-E in intubated and sedated critically ill patients. Objectives:To evaluate the effect of CPT with or without MI-E device application on mucus clearance and to evaluate MI-E effect on pulmonary mechanics. To assess expiratory flows (PEF) generated during MI-E intervention.

Mechanical in-exsufflation (MI-E) is a non-invasive chest physiotherapy (CPT) technique aimed to generate high expiratory flows to simulate cough, and improve mucus clearance in the proximal airways. To date, no studies have assessed the effects of MI-E in intubated and sedated critically ill patients. Objectives:To evaluate the effect of CPT with or without MI-E device application on mucus clearance and to evaluate MI-E effect on pulmonary mechanics. To assess expiratory flows (PEF) generated during MI-E intervention. Methods: Controlled, randomized semi cross-over, single blind trial conducted at Bordeaux Universitary Hospital. We included intubated and sedated patients (>18 ys) connected to IMV > 48 h. We excluded patients with parenchyma damage, respiratory and/or hemodynamic instability, and high infectious risk. All patients received CPT followed by an ETS twice daily. Additionally, patients randomly received MI-E in one of the sessions (one day intervention).MI-E treatment consisted in 4 series of 5 in-expiratory cycles at +/− 40 cmH2O pressure, 3 sec inspiratory time, 2 sec expiratory time, and 1 sec pause between cycle. Mucus clearance was assessed through wet volume of sputum retrieved during ETS. Pulmonary mechanics was measured before, immediately after and 1 hr after intervention through a pneumotacograph (PNT). PEF during MI-E was continuously measured through a PNT. Results: 10 patients were enrolled in the study The volume of secretions recovered (Me, IQR) was 1.1 ml (0.76-2.29) and 0.41 ml (0–1.27) with or without MI-E respectively (p = 0.06).

Pulmonary mechanics was measured before, immediately after and 1 hr after intervention through a pneumotacograph (PNT). PEF during MI-E was continuously measured through a PNT. Results: 10 patients were enrolled in the study The volume of secretions recovered (Me, IQR) was 1.1 ml (0.76-2.29) and 0.41 ml (0–1.27) with or without MI-E respectively (p = 0.06). No significant differences in pulmonary mechanics were found before, immediately after or 1 hour after the use of MI-E or between groups. PEF generated during MI-E was 1.66 ± 0.42 l/s. Conclusion: MI-E may not be a useful technique to improve mucus clearance in intubated and sedated critically ill patients when conventional setting is used. Indeed, the MI-E setting applied in our study (ie: +/− 40 cmH2O) did not generate an efficient cough, which is considered to be PEF > 2.7 l/s in this population of patients. MI-E does not have significant short or long-effects on pulmonary mechanics. References Branson RD. Respir Care 2007; 52(10): 1328–42. Homnick D. Respir Care. 2007; 56: 1296–307. Guérin C et al. Respir Care. 2011 Aug; 56(8):1108–14.Table 12 (abstract A861). Anthropometric values of enrolled patients (X; SD) Male Female PATIENTS (n = 10) 4 6 AGE (years) 52.7 ± 23.5 59 ± 24.1 BODY MASS INDEX (kg/m2) 23.5 ± 2.9 28.6 ± 14.7 REASON FOR ICU ADMISSION: Abdominal or neurologic surgery 3 6 Cardiorrespiratory arrest 1 INTUBATION (hours) 61.6 ± 2.1 72.3 ± 17.8 APACHE II (%) 33.7 ± 11.8 33.3 ± 8.9 PaO2/FiO2 270.6 ± 62.6 207.5 ± 43.9 Fig. 15 (abstract A861). Vol of secretions with or without MI-E

24.1 BODY MASS INDEX (kg/m2) 23.5 ± 2.9 28.6 ± 14.7 REASON FOR ICU ADMISSION: Abdominal or neurologic surgery 3 6 Cardiorrespiratory arrest 1 INTUBATION (hours) 61.6 ± 2.1 72.3 ± 17.8 APACHE II (%) 33.7 ± 11.8 33.3 ± 8.9 PaO2/FiO2 270.6 ± 62.6 207.5 ± 43.9 Fig. 15 (abstract A861). Vol of secretions with or without MI-E Table 13 (abstract A861). Pulmonary mechanics values before, after and 1 hr Static compliance (Cst) before (ml/cmH2O) Cst after Cst 1 hr after Respiratory system resistance (Rrs) before (cmH2O/l/s) Rrs after Rrs 1 hr after Airway resistance (Raw) before (cmH2O/l/s) Raw after Raw 1 hr after CPT (X;SD) 53.7 ± 27.15 62.81 ± 45.1 53.71 ± 19.61 17.89 ± 6.54 18.33 ± 6.01 15.81 ± 3.85 2.48 ± 0.68 2.49 ± 0.69 2.14 ± 0.45 CPT + MI-E (X;SD) 58.33 ± 22.2 72.38 ± 32.55 59.87 ± 28.68 17.48 ± 2.75 18.62 ± 3.62 16.31 ± 3.29 2.27 ± 0.62 2.38 ± 0.58 2.19 ± 0.58

resistance (Raw) before (cmH2O/l/s) Raw after Raw 1 hr after CPT (X;SD) 53.7 ± 27.15 62.81 ± 45.1 53.71 ± 19.61 17.89 ± 6.54 18.33 ± 6.01 15.81 ± 3.85 2.48 ± 0.68 2.49 ± 0.69 2.14 ± 0.45 CPT + MI-E (X;SD) 58.33 ± 22.2 72.38 ± 32.55 59.87 ± 28.68 17.48 ± 2.75 18.62 ± 3.62 16.31 ± 3.29 2.27 ± 0.62 2.38 ± 0.58 2.19 ± 0.58 A862 Evaluation of semi-recumbent position compliance in patients with mechanical ventilation: comparison of three data collection methods M. Llaurado-Serra1, A. Lobo-Civico2, A. Ventura-Rosado3, A. Piñol-Tena4, M. Pi-Guerrero5, C. Paños-Espinosa6, M. Peralvo-Bernat7, J. Marine-Vidal7, R. Gonzalez-Engroba3, N. Montesinos-Cerro2, M. Treso-Geira4, A. Valeiras-Valero5, L. Martinez-Reyes6, A. Sandiumenge8, M.F. Jimenez-Herrera1, CAPCRI Study 1Rovira i Virgili University, Nursing Department, Tarragona, Spain; 2Dr Josep Trueta University Hospital, Intensive Care Unit, Girona, Spain; 3Quiron Salud-Hospital General de Catalunya, Intensive Care Unit, Barcelona, Spain; 4Verge de la Cinta University Hospital, Intensive Care Unit, Tortosa, Spain; 5Hospital de Sant Joan Despí Moissès Broggi, Intensive Care Unit, Barcelona, Spain; 6Sant Pau I Santa Tecla University Hospital, Intensive Care Unit, Tarragona, Spain; 7Joan XXIII University Hospital, Intensive Care Unit, Tarragona, Spain; 8University Hospital Vall d´Hebron, Medical Transplant Coordination Department, Barcelona, Spain Correspondence: M. Llaurado-Serra – Rovira i Virgili University, Nursing Department, Tarragona, Spain Background: Semi-recumbent position (SP) is defined as the head-of-bed elevation (HOBE) between 30-45°. Despite being included in guidelines for the prevention of ventilator-associated pneumonia, it is accomplished less than recommended.

Llaurado-Serra – Rovira i Virgili University, Nursing Department, Tarragona, Spain Background: Semi-recumbent position (SP) is defined as the head-of-bed elevation (HOBE) between 30-45°. Despite being included in guidelines for the prevention of ventilator-associated pneumonia, it is accomplished less than recommended. Objectives: To investigate the factors related to SP from three different perspectives: direct observation, questionnaire and asking the nurse when not complying with the recommendation. Methods: Descriptive, longitudinal and multicenter study in 6 Spanish intensive care units (ICU). Patients were included if they were under mechanical ventilation >48 h, ≥18 years old and had no contraindication for SP, were not mechanically ventilated during the previous 7 days or were not intubated in pre-hospital setting. HOBE was measured 3 times/day with BOSCH GLM80® device. When HOBE was < 30°, the investigator asked the staff nurse the underlying reason. At the end of patients' inclusion, CAPCRI-Q questionnaire was distributed to all ICU staff. Descriptive, bivariate and multivariate analyses were performed according to the source of data. Significance was set at p < 0.05.

GLM80® device. When HOBE was < 30°, the investigator asked the staff nurse the underlying reason. At the end of patients' inclusion, CAPCRI-Q questionnaire was distributed to all ICU staff. Descriptive, bivariate and multivariate analyses were performed according to the source of data. Significance was set at p < 0.05. Results: 276 patients were included (6894 measurements). Mean HOBE was 30.1(6.7)° and 45.9 % of the measurements were < 30° with significant differences among ICUs. In multivariate analysis the most relevant factors related to a decreased SP compliance according to measurements were the ICU(p < 0.001), agitation [OR 0.39 (CI95% 0.28-0.54)], renal replacement therapy [OR 0.56 (CI95% 0.47-0.66)], abdominal vacuum therapy [OR 0.59 (CI95% 0.37-0.95)] and open abdomen [OR 0.65 (CI95% 0.50-0.85)]. From staff nurses, 2146 reasons for non-compliance were obtained. The most frequent causes were incorrect visual perception (n = 568) and that the HOBE measuring device from the bed indicated 30° (n = 460). Patient's clinical conditions represented 33.2 % (n = 742) of the causes for non-compliance whilst factors related to patient care represented 66.3 %(n = 1484). Finally, 223 questionnaires were obtained (77.8 % response rate). According to the results, the most relevant factors with a negative effect to SP compliance were clinical contraindications for SP (98.7 %), patient discomfort (89.2 %), abdominal surgery with open abdomen (85.7 %), unavailability of beds with HOBE measuring device (80.7 %), lack of awareness of the importance of the measure (79.8 %) and lack of experience in critical care (77.6 %).

ve effect to SP compliance were clinical contraindications for SP (98.7 %), patient discomfort (89.2 %), abdominal surgery with open abdomen (85.7 %), unavailability of beds with HOBE measuring device (80.7 %), lack of awareness of the importance of the measure (79.8 %) and lack of experience in critical care (77.6 %). Conclusions: SP compliance is below recommended but the mean HOBE reaches the lower limit of the recommendation. The factors affecting SP compliance differ according to the method used for data collection and include other factors than patient's clinical condition. Politics targeting to increase its compliance should address various areas of care such as team and professionals, resources and equipment and re-consider clinical indications for SP. Grant acknowledgement The project was funded by the 14° National Award of Nursing Research from Marques de Valdecilla Hospital (Spain).

Conclusions: SP compliance is below recommended but the mean HOBE reaches the lower limit of the recommendation. The factors affecting SP compliance differ according to the method used for data collection and include other factors than patient's clinical condition. Politics targeting to increase its compliance should address various areas of care such as team and professionals, resources and equipment and re-consider clinical indications for SP. Grant acknowledgement The project was funded by the 14° National Award of Nursing Research from Marques de Valdecilla Hospital (Spain). A863 A comparison of staff perceptions/attitudes, supportive care standards and patient pathophysiology between the resuscitation room, operating rooms and critical care units of a central London university hospital S. Helyar, P. Riozzi, A. Noon, G. Hallows, H. Cotton, J. Keep, P.A. Hopkins King's College Hospital, ACET Research Team, London, United Kingdom Correspondence: P. Riozzi – King’s College Hospital, ACET Research Team, London, United Kingdom Introduction: Critically ill patients presenting to the emergency department (ED) frequently go on to being managed within the operating rooms (OR) and intensive care units (ICU) of the same institution. Here we describe an observational analysis of airway (A), breathing (B), and circulation (C) pathophysiology and supportive care in these three locations to test whether there are differences in practice. Objectives:To compare the perceptions and attitudes of professionals (ED vs OR vs ICU) for supportive care in A, B and C.

A863 A comparison of staff perceptions/attitudes, supportive care standards and patient pathophysiology between the resuscitation room, operating rooms and critical care units of a central London university hospital S. Helyar, P. Riozzi, A. Noon, G. Hallows, H. Cotton, J. Keep, P.A. Hopkins King's College Hospital, ACET Research Team, London, United Kingdom Correspondence: P. Riozzi – King’s College Hospital, ACET Research Team, London, United Kingdom Introduction: Critically ill patients presenting to the emergency department (ED) frequently go on to being managed within the operating rooms (OR) and intensive care units (ICU) of the same institution. Here we describe an observational analysis of airway (A), breathing (B), and circulation (C) pathophysiology and supportive care in these three locations to test whether there are differences in practice. Objectives:To compare the perceptions and attitudes of professionals (ED vs OR vs ICU) for supportive care in A, B and C. To compare supportive care standards (ED vs OR vs ICU) in A,B and C. To compare patient pathophysiology, in the first 24 hours of hospital admission, in A,B and C.

A863 A comparison of staff perceptions/attitudes, supportive care standards and patient pathophysiology between the resuscitation room, operating rooms and critical care units of a central London university hospital S. Helyar, P. Riozzi, A. Noon, G. Hallows, H. Cotton, J. Keep, P.A. Hopkins King's College Hospital, ACET Research Team, London, United Kingdom Correspondence: P. Riozzi – King’s College Hospital, ACET Research Team, London, United Kingdom Introduction: Critically ill patients presenting to the emergency department (ED) frequently go on to being managed within the operating rooms (OR) and intensive care units (ICU) of the same institution. Here we describe an observational analysis of airway (A), breathing (B), and circulation (C) pathophysiology and supportive care in these three locations to test whether there are differences in practice. Objectives:To compare the perceptions and attitudes of professionals (ED vs OR vs ICU) for supportive care in A, B and C. To compare supportive care standards (ED vs OR vs ICU) in A,B and C. To compare patient pathophysiology, in the first 24 hours of hospital admission, in A,B and C. Methods: Institutional/patient & public approval was gained to collect staff opinions and anonymised patient data. 90 ED, 30 OR and 110 ICU professionals were surveyed following a pilot to determine options, ranking and scoring criteria a priori where needed. Anonymous patient data from 36 intubated patients who were cared for in the ED, OR and ICU within their first 24 hours were collected. This included physiological observations and supportive care standards around A, B and C.

yed following a pilot to determine options, ranking and scoring criteria a priori where needed. Anonymous patient data from 36 intubated patients who were cared for in the ED, OR and ICU within their first 24 hours were collected. This included physiological observations and supportive care standards around A, B and C. Results: The most striking differences in staff opinion involved the preferential use of artificial colloid-based fluid resuscitation in sepsis (17 % ED staff; 50 % OR staff; 8 % critical care staff); the value and significance of recording End Tidal CO2 (4 % ED staff; 97 % OR staff; 69 % critical care staff); and the potential preferential use of flow-directed fluid boluses rather than pressure-directed fluid boluses in critically ill patients (22 % ED staff; 90 % anaesthetic staff; 75 % critical care staff). When observing supportive care standards the largest differences were in the use of stress ulcer prophylaxis (only prescribed in critical care); patient positioning (head-up: 94 % patients in CC; 6 % in ED); the recording of sedation level (0 % ED; 6 % anaesthetic; 78 % critical care) and the recording of ventilator parameters-tidal volume, peak pressure and ET-CO2-(3 % ED; 69 % anaesthetic; 100 % critical care) . Finally, in respect to patient pathophysiology, all groups were under ventilated and over oxygenated. Mean arterial pressure was most divergent from baseline in the ED. However, changes in pathophysiology were related to interventions (fluid boluses, analgesia, surgical interventions, inotropes, pressors) rather than location. Despite the divergent views regarding the relative value of flow monitoring, observed fluid boluses were predominantly triggered by pressure changes in all three locations.

nges in pathophysiology were related to interventions (fluid boluses, analgesia, surgical interventions, inotropes, pressors) rather than location. Despite the divergent views regarding the relative value of flow monitoring, observed fluid boluses were predominantly triggered by pressure changes in all three locations. Conclusions: Differences in staff attitudes; application of standards and patient pathophysiology were identified between care locations. The influence of variation in resources and professional composition of teams (nurses:doctors) on these results requires further work. It remains uncertain whether more uniform approaches would improve patient outcomes. A864 Lung comet score (LCS) for evaluation of extravascular lung water (EVLW) in intensive care unit (ICU) patients undergoing renal replacement therapies (RRT) A. Taggu1, S. Renuka2, S. Sampath3 1St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India; 2St. Johns Medical College Hospital, Bangalore, India; 3St. Johns Medical College Hospital, Bangalore, India Correspondence: St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India Introduction: Lung ultrasound (LUS) is now considered as a valid and fast method that allow quantification and monitoring of EVLW. There is paucity of such studies in ICU patients. Bio-impedance analysis (BIA) is a validated tool for EVLW but is not easily available. Objectives: To test the hypothesis that lung comet score (LCS) is a reliable surrogate of EVLW in ICU patients undergoing RRT.

A864 Lung comet score (LCS) for evaluation of extravascular lung water (EVLW) in intensive care unit (ICU) patients undergoing renal replacement therapies (RRT) A. Taggu1, S. Renuka2, S. Sampath3 1St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India; 2St. Johns Medical College Hospital, Bangalore, India; 3St. Johns Medical College Hospital, Bangalore, India Correspondence: St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India Introduction: Lung ultrasound (LUS) is now considered as a valid and fast method that allow quantification and monitoring of EVLW. There is paucity of such studies in ICU patients. Bio-impedance analysis (BIA) is a validated tool for EVLW but is not easily available. Objectives: To test the hypothesis that lung comet score (LCS) is a reliable surrogate of EVLW in ICU patients undergoing RRT. Methods: A prospective observational study was conducted on 450 patients in ICU needing RRT. Exclusion criteria were Age < 18 years, pregnant, amputees, cardiac pacemakers, pre-existing lung diseases and ascites. Lung comet score as per validated technique , BIA measurements and baseline data were collected pre and post dialysis. Lung comet score and other covariates were fitted into a regression model using BIA as the standard test. Based on BIA delta hydration relative (HS rel), patients were divided into normohydration and hyperhydration using a cut-off of 15 %.

nique , BIA measurements and baseline data were collected pre and post dialysis. Lung comet score and other covariates were fitted into a regression model using BIA as the standard test. Based on BIA delta hydration relative (HS rel), patients were divided into normohydration and hyperhydration using a cut-off of 15 %. Results: A linear regression model in predialysis state showed that only LCS could significantly predict lung water (const 8.79, Coef.0.203, p value 0.00 ). In the postdialysis state LCS perfectly predicted lung water (const 9.98, Coef.,0.261; p value 0.00 ).Bland Altman plots showed good agreement between LCS and hydration status (BIA) pre and post dialysis. The LCS >15 nearly perfectly predicted hydration status in both pre and post dialysis states. Conclusions: Lung comet score is a good surrogate of EVLW and reliably predicts reflects hydration status pre and post dialysis in ICU patients. References 1. Wizemann V, Wabel P, Chamney P. The mortality risk of over hydration in haemodialysis patients. Nephrol Dial Transplant 2009;24:1574–1579. 2. Noble VE, Murray AF, Capp R, Sylvia-Reardon MH, Steele DJ, Liteplo A. Ultrasound assessment for extravascular lung water in patients undergoing hemodialysis. Time course for resolution. Chest 2009;135:1433–9. Grant acknowledgement NoneTable 14 (abstract A864). Baseline characteristics

1. Wizemann V, Wabel P, Chamney P. The mortality risk of over hydration in haemodialysis patients. Nephrol Dial Transplant 2009;24:1574–1579. 2. Noble VE, Murray AF, Capp R, Sylvia-Reardon MH, Steele DJ, Liteplo A. Ultrasound assessment for extravascular lung water in patients undergoing hemodialysis. Time course for resolution. Chest 2009;135:1433–9. Grant acknowledgement NoneTable 14 (abstract A864). Baseline characteristics PARAMETERS normohydration n = 213 hyperhydration n = 237 p -value Male/Female 190/23 207/30 0.074 Age (years) 50.4 (sd 11.48) 50.95 (sd12.30) 0.811 APACHE II 14.25 (sd 4.8) 14.6 (sd 6.8) 0.082 Body mass index(kg/m2) 20.4 (sd 2.52) 19.95 (sd 6.2) 0.643 Mean Arterial Pressure (mmHg) 90.80 (sd 5.7) 102.52 (sd.6.8) 0.001 Heart Rate (beats/min) 99.8 (sd 7.02) 103.5 (sd 4.8) 0.001 Diabetes mellitus 9.7 % 14.6 % 0.042 Hypertension 14.5 % 16.2 % 0.081 acute febrile illness 21.3 % 14.6 % 0.08 Table 15 (abstract A864). Pre and post dialysis comparison of LCS and BIA Dialysis state Hydration status (as per BioImpedance Analysis) LCS (lung comet score) < 15 LCS >15 Fischer exact test PREDIALYSIS Normohydration 210 0 P = 0.00 Hyperhydration 3 237 POSTDIALYSIS Normohydration 446 0 P = 0.00 Hyperhydration 0 4 Fig. 16 (abstract A864). Bland Altman graphs for LCS BIA

PARAMETERS normohydration n = 213 hyperhydration n = 237 p -value Male/Female 190/23 207/30 0.074 Age (years) 50.4 (sd 11.48) 50.95 (sd12.30) 0.811 APACHE II 14.25 (sd 4.8) 14.6 (sd 6.8) 0.082 Body mass index(kg/m2) 20.4 (sd 2.52) 19.95 (sd 6.2) 0.643 Mean Arterial Pressure (mmHg) 90.80 (sd 5.7) 102.52 (sd.6.8) 0.001 Heart Rate (beats/min) 99.8 (sd 7.02) 103.5 (sd 4.8) 0.001 Diabetes mellitus 9.7 % 14.6 % 0.042 Hypertension 14.5 % 16.2 % 0.081 acute febrile illness 21.3 % 14.6 % 0.08 Table 15 (abstract A864). Pre and post dialysis comparison of LCS and BIA Dialysis state Hydration status (as per BioImpedance Analysis) LCS (lung comet score) < 15 LCS >15 Fischer exact test PREDIALYSIS Normohydration 210 0 P = 0.00 Hyperhydration 3 237 POSTDIALYSIS Normohydration 446 0 P = 0.00 Hyperhydration 0 4 Fig. 16 (abstract A864). Bland Altman graphs for LCS BIA A865 Daily sleep evaluation in the ICU: a one year prospective cohort study P.J.T. Rood, T. Frenzel, R. Verhage, M. Bonn, P. Pickkers, J.G. van der Hoeven, 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: Sleep is a fundamental need for recovery, and lack of good sleep is associated with adverse effects. ICU patients have increased risk for disturbed sleep architecture. The importance of sleep is internationally recognized. Recently, we determined that the optimal cut-off score for sufficient sleep was a numeric rating score (NRS sleep) of ≥6(1). Subsequently, this sleep NRS was implemented in our ICU practice, supplemented with the Richards Campbell Sleep Questionnaire (RSCQ) when sleep was deemed insufficient.

nternationally recognized. Recently, we determined that the optimal cut-off score for sufficient sleep was a numeric rating score (NRS sleep) of ≥6(1). Subsequently, this sleep NRS was implemented in our ICU practice, supplemented with the Richards Campbell Sleep Questionnaire (RSCQ) when sleep was deemed insufficient. Objective: To determine quality of sleep in daily ICU practice in a large cohort of ICU patients. Methods: A prospective cohort study was performed at the ICU of a university medical center from January to December 2015. In non-comatose patients (RASS −2 to +1), nurses daily asked patients to rate their perceived sleep of the last night on a numeric rating scale of 0–10. When sleep quality was insufficient (NRS sleep < 6), subsequently the RCSQ was performed. Subsequently tailored, non-pharmacological protocolized sleep enhancement interventions were applied, like improving the circadian rhythm by providing stimuli (i.e. light, mobilisation) during daytime, and reducing these (i.e. noise, light, alarms) and clustering of care activities at night. If necessary, medication strategies were added/adjusted. Compliance to the assessments of NRS sleep was measured and defined as compliant when sleep quality was assessed at least once per day after a full night ICU presence.

ime, and reducing these (i.e. noise, light, alarms) and clustering of care activities at night. If necessary, medication strategies were added/adjusted. Compliance to the assessments of NRS sleep was measured and defined as compliant when sleep quality was assessed at least once per day after a full night ICU presence. Results: A total of 1603 ICU patients ranked their sleep during their ICU stay. The age was mean 63 ± 14 (mean ± SD), the mean APACHE II score was 16 ± 6, and 62.3 % patients were male. In total 1062 (66 %) were surgical patients, 331 (21 %) medical , and 201 (13 %) were neurological patients. Most patients (77 %) were mechanically ventilated for median 1 day [IQR 1–2]. ICU length of stay was median 1 day [IQR 1–2], and in-hospital stay was median 9 days [IQR 5–17]. The compliance of the sleep measurements was 59 %. Of 4,532 unique daily measurements, 3,199 (71 %) NRS sleep was rated as sufficient by the patients. Overall, the NRS sleep was 6 ± 2. No differences in sleep scores were found between surgical, medical and neurological patients. When sleep was rated as insufficient, multiple RCSQ themes were marked as important factors (Table 16). Conclusions: With this two step-approach of assessing sleep in ICU practice, we determined that 7 out of 10 ICU patients experienced adequate sleep. However, the mean NRS sleep was just at the cut-off value of sufficient sleep which was determined earlier. No differences between specific patients groups were found. Sleep quality and depth of sleep were the most important themes for insufficient sleep measured with the RCSQ. References

Conclusions: With this two step-approach of assessing sleep in ICU practice, we determined that 7 out of 10 ICU patients experienced adequate sleep. However, the mean NRS sleep was just at the cut-off value of sufficient sleep which was determined earlier. No differences between specific patients groups were found. Sleep quality and depth of sleep were the most important themes for insufficient sleep measured with the RCSQ. References 1. Rood P, Pickkers P, vander Hoeven J, vanden Boogaard M. Sufficient sleep quality easily measured: a multicenter centre study in dutch ICUS.Intensive Care Medicine Experimental. 2015;3(1):1–2.Table 16 (abstract A865). Measurements of insufficient sleep (NRS sleep <6) N = 1949 NRS score, median [IQR] 4 [3–5] RCSQ theme as important reason for not slepping, n (%) Sleep depth, 1539 (79) Sleep Latency 987 (51) Awakenings 1186 (61) Returning to sleep 1156 (59) Sleep quality 1616 (83)

1. Rood P, Pickkers P, vander Hoeven J, vanden Boogaard M. Sufficient sleep quality easily measured: a multicenter centre study in dutch ICUS.Intensive Care Medicine Experimental. 2015;3(1):1–2.Table 16 (abstract A865). Measurements of insufficient sleep (NRS sleep <6) N = 1949 NRS score, median [IQR] 4 [3–5] RCSQ theme as important reason for not slepping, n (%) Sleep depth, 1539 (79) Sleep Latency 987 (51) Awakenings 1186 (61) Returning to sleep 1156 (59) Sleep quality 1616 (83) A866 Nurse-performed chest ultrasonography for the localization of central venous catheter and detection of post-procedural malpositions and complications as an alternative to chest radiography F. Corradi, L. Melnyk, F. Moggia, R. Pienovi, G. Adriano, C. Brusasco, L. Mariotti, M. Lattuada Ente Ospedaliero Ospedali Galliera, Anestesia e Rianimazione, Genova, Italy Correspondence: L. Melnyk – Ente Ospedaliero Ospedali Galliera, Anestesia e Rianimazione, Genova, Italy Introduction: Central venous catheterization (CVC) is considered commonplace in the care of critically ill patients. It has been demonstrated that ultrasounds performed by an expert physician can detect accurately CVC´s mechanical complications and misplacements1. By contrast, there is still no consensus regarding their diagnostic power in the identification of silent CVC complications, when performed by a trained nurse.

ll patients. It has been demonstrated that ultrasounds performed by an expert physician can detect accurately CVC´s mechanical complications and misplacements1. By contrast, there is still no consensus regarding their diagnostic power in the identification of silent CVC complications, when performed by a trained nurse. Objectives: To determine the usefulness of nurse-performed ultrasound to evaluate CVC misplacements and detection of pneumothorax (PNX), thus obviating post-procedural radiograph (CXR). After the insertion of a central venous catheter, CXR is usually obtained to ensure correct positioning of the catheter tip and to detect post-procedural complications. Recently published current guidelines for appropriate use of bedside general and cardiac ultrasonography in the evaluation of critically ill patients, suggest a detailed post-cannulation ultrasound examination to confirm CVC location and exclude PNX in adult patients2. Methods: A prospective study of 47 CVC procedures, was conducted in an adult mixed intensive care unit. At the end of the procedure, a B-mode ultrasonography was first performed to assess catheter position and detect PNX. Then, contrast enhanced ultrasonography (CEUS) was used to facilitate visualization of catheter tip1, avoiding unknown right atrium positioning or artifacts. A post-procedural CXR was obtained for all patients and was considered as a reference technique

hy was first performed to assess catheter position and detect PNX. Then, contrast enhanced ultrasonography (CEUS) was used to facilitate visualization of catheter tip1, avoiding unknown right atrium positioning or artifacts. A post-procedural CXR was obtained for all patients and was considered as a reference technique Results: Internal Jugular Vein malposition was detected in two patients by B-mode ultrasonography and right atrium positioning was detected in 4 patients by CEUS. CEUS yielded a 100 % sensitivity and 98 % specificity in detecting catheter misplacement. k value between the two methods was 0.91 (p < .001). PNX was detected in one patient by ultrasonography and in none by CXR. The mean time required to perform ultrasonography plus CEUS was 10 ± 3 mins vs. 25 ± 8 mins for CXR (p < .001). Conclusions: The close concordance between CEUS and CXR justifies the use of nurse-performed sonography to ensure the correct positioning of the catheter tip and to detect PNX after CVC cannulation to optimize use of hospital resources and minimize time consumption and radiation. CXR will be necessary when sonographic examination is impossible to perform by technical limitations. References

Conclusions: The close concordance between CEUS and CXR justifies the use of nurse-performed sonography to ensure the correct positioning of the catheter tip and to detect PNX after CVC cannulation to optimize use of hospital resources and minimize time consumption and radiation. CXR will be necessary when sonographic examination is impossible to perform by technical limitations. References 1) Ultrasound localization of central vein catheter and detection of postprocedural pneumothorax: an alternative to chest radiography. Vezzani A et al. Crit Care Med. 2010 Feb;38(2):533–8 2) Guidelines for the Appropriate Use of Bedside General and Cardiac Ultrasonography in the Evaluation of Critically Ill Patients-Part I: General Ultrasonography. Frankel HL et al. Crit Care Med. 2015 Nov;43(11):2479–502

mothorax: an alternative to chest radiography. Vezzani A et al. Crit Care Med. 2010 Feb;38(2):533–8 2) Guidelines for the Appropriate Use of Bedside General and Cardiac Ultrasonography in the Evaluation of Critically Ill Patients-Part I: General Ultrasonography. Frankel HL et al. Crit Care Med. 2015 Nov;43(11):2479–502 A867 The art of nurse-family communication: a qualitative descriptive study of how nurses initiate and manage communication with families during treatment withdrawal processes in intensive care M.J. Bloomer1,2, M. Coombs3, K. Ranse4, R. Endacott5,6,7 1Deakin University, School of Nursing and Midwifery, Melbourne, Australia; 2Deakin University, Centre for Quality and Patient Safety, Melbourne, Australia; 3Victoria University, Wellington, Graduate School of Nursing Midwifery and Health, Wellington, New Zealand; 4University of Canberra, Faculty of Health, Canberra, Australia; 5Royal Devon and Exeter Hospital, Royal Devon and Exeter Clinical School, Exeter, United Kingdom; 6Plymouth University, Faculty of Health and Human Sciences, Plymouth, United Kingdom; 7Monash University, School of Nursing and Midwifery, Melbourne, Australia Correspondence: M.J. Bloomer – Deakin University, School of Nursing and Midwifery, Melbourne, Australia Introduction: Treatment withdrawal in intensive care is common(1). Whilst significant research attention has focused on how treatment is withdrawn and what information is communicated to families(2), little is known about how critical care nurses initiate and manage family communication needs at this time.

e, Australia Introduction: Treatment withdrawal in intensive care is common(1). Whilst significant research attention has focused on how treatment is withdrawn and what information is communicated to families(2), little is known about how critical care nurses initiate and manage family communication needs at this time. Objectives: The purpose of this study was to explore how nurses initiate and manage communication with family during treatment withdrawal in the ICU. Methods: A qualitative descriptive approach was used. Semi-structured focus group interviews were conducted with adult critical care nurses from four Intensive Care Units, two in Australia and two in New Zealand. Results: Twenty-one critical care nurses participated in the study. The focus groups ranged from 39–58 minutes. Participants had a mean of 13 years experience in critical care and all had experience of treatment withdrawal and providing end-of-life care. Inductive content analysis was used to identify five key themes. READING THE ROOM describes the need to look for non-verbal and other cues to help guide initiating communication. ESTABLISHING THE WHO was seen as important in building rapport and aided in establishing trust. NAVIGATING HOW to communicate was considered just as important as the content. Similarly, JUDGING WHEN to communicate was equally essential to aid family understanding and coping. ASSESSING WHAT to say was individually tailored to each family. In terms of WHERE, participants reported that the art of nurse-family communication was learnt on the job.

was considered just as important as the content. Similarly, JUDGING WHEN to communicate was equally essential to aid family understanding and coping. ASSESSING WHAT to say was individually tailored to each family. In terms of WHERE, participants reported that the art of nurse-family communication was learnt on the job. Conclusions: Initiating and managing communication with family during treatment withdrawal is a complex and multi-faceted nursing activity that can have significant impact on families. There is need for support and education that develops both the art, as well as the science, in this frequently encountered aspect of end-of-life care. References 1. Bloomer, M.J., et al., End of life management of adult patients in an Australian metropolitan intensive care unit: a retrospective observational study. Australian Critical Care, 2010. 23(1): p. 13–19. 2. Truog, R.D., et al., Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College of Critical Care Medicine. Critical Care Medicine, 2008. 36(3): p. 953–963. Grant acknowledgement This study was partially funded by an Experience Researcher Grant from the Australian College of Critical Care Nurses, Australia

2. Truog, R.D., et al., Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College of Critical Care Medicine. Critical Care Medicine, 2008. 36(3): p. 953–963. Grant acknowledgement This study was partially funded by an Experience Researcher Grant from the Australian College of Critical Care Nurses, Australia A868 Taper-shaped endotracheal cuffs in the prevention of endotracheal tube-associated pneumonia (ETAP): a systematic review and meta-analysis of randomized controlled clinical trials B. Maertens1, K. Blot1, S. Blot2 1Ghent University, Faculty of Medicine and Health Sciences, Ghent, Belgium; 2Ghent University, Dept. of Internal Medicine, Ghent, Belgium Correspondence: B. Maertens – Ghent University, Faculty of Medicine and Health Sciences, Ghent, Belgium Introduction: Micro-aspiration of subglottic secretions is considered a major pathogenic mechanism of endotracheal tube-associated pneumonia (ETAP), either postoperative pneumonia or ventilator-associated pneumonia. Endotracheal tubes (ETs) with taper-shaped cuffs have been proposed to provide a better seal of the extraluminal airway, thereby preventing micro-aspiration and possibly ETAP. Objectives: To perform a systematic review and meta-analysis to assess the efficacy of ETs with taper-shaped cuffs in the prevention of ETAP.

A868 Taper-shaped endotracheal cuffs in the prevention of endotracheal tube-associated pneumonia (ETAP): a systematic review and meta-analysis of randomized controlled clinical trials B. Maertens1, K. Blot1, S. Blot2 1Ghent University, Faculty of Medicine and Health Sciences, Ghent, Belgium; 2Ghent University, Dept. of Internal Medicine, Ghent, Belgium Correspondence: B. Maertens – Ghent University, Faculty of Medicine and Health Sciences, Ghent, Belgium Introduction: Micro-aspiration of subglottic secretions is considered a major pathogenic mechanism of endotracheal tube-associated pneumonia (ETAP), either postoperative pneumonia or ventilator-associated pneumonia. Endotracheal tubes (ETs) with taper-shaped cuffs have been proposed to provide a better seal of the extraluminal airway, thereby preventing micro-aspiration and possibly ETAP. Objectives: To perform a systematic review and meta-analysis to assess the efficacy of ETs with taper-shaped cuffs in the prevention of ETAP. Methods: A systematic search of MEDLINE, EMBASE and CENTRAL/CCTR was conducted in March 2016. Eligible trials were randomized controlled clinical trials (RCTs) comparing taper-shaped cuffs with standard, cylindrical-shaped cuffs in intubated patients. All studies reporting the incidence of ETAP were included. Inclusion of trials was irrespective of publication status, date of publication or language. Random-effects meta-analysis calculated the risk ratio (RR) and 95 % confidence interval (CI) for the incidence of ETAP between both groups using the Mantel-Haenszel method.

tudies reporting the incidence of ETAP were included. Inclusion of trials was irrespective of publication status, date of publication or language. Random-effects meta-analysis calculated the risk ratio (RR) and 95 % confidence interval (CI) for the incidence of ETAP between both groups using the Mantel-Haenszel method. Results: Three RCTs, given a total of 855 patients, met the inclusion criteria. One trial was published as a conference abstract only (1), while the others were published in full (2,3). None of the trials was blinded for the intervention. 440 patients were allocated to the intervention arm and 415 to the control arm. 87 ETAP episodes occurred in the intervention group and 83 in the control group. The pooled RR for the incidence of ETAP was 1.01 (95 % CI, 0.78-1.31; z = 0.11 p = 0.91). Conclusions: The use of endotracheal tubes with taper-shaped cuffs did not show to reduce the incidence of ETAP. However, the number of available studies is small, and there is an inherent risk of bias due to the unblinded designs. References 1. Saito N, et al. 33rd International Symposium on Intensive Care and Emergency Medicine 2013 2. Philippart F, et al. Am J Respir Crit Care 2015 3. Monsel A, et al. Anesthesiology 2016 Grant acknowledgement We received no grants for this study.Fig. 17 (abstract A868). Meta-analysis Forest plot; incidence of ETAP

Conclusions: The use of endotracheal tubes with taper-shaped cuffs did not show to reduce the incidence of ETAP. However, the number of available studies is small, and there is an inherent risk of bias due to the unblinded designs. References 1. Saito N, et al. 33rd International Symposium on Intensive Care and Emergency Medicine 2013 2. Philippart F, et al. Am J Respir Crit Care 2015 3. Monsel A, et al. Anesthesiology 2016 Grant acknowledgement We received no grants for this study.Fig. 17 (abstract A868). Meta-analysis Forest plot; incidence of ETAP A869 Predictability of nursing workload of patients admitted to the ICU after IHCA/OHCA, assessed with the nursing activities score (NAS) in a university hospital M.P. van Nieuw Amerongen1, E.S. van der Heiden1, J.W.R. Twisk2, A.R.J. Girbes1, J.J. Spijkstra1 1VU University Medical Center, Intensive Care, Amsterdam, Netherlands; 2VU University Medical Center, Epidemiology and Biostatistics, Amsterdam, Netherlands Correspondence: M.P. van Nieuw Amerongen – VU University Medical Center, Intensive Care, Amsterdam, Netherlands Background: Understaffing of ICU's can have serious adverse consequences both for patients and for nurses, and therefore it is important to have an adequate number of nurses on the ward. Nurses however are in short demand and resources are scarce. Being able to predict the nursing workload for a certain group of patients may help to allocate nursing capacity as efficiently as possible and thus to reduce costs, without endangering the patients safety and nurses' health. NAS is a validated tool for the measurement of nursing workload in an intensive care unit.

ce. Being able to predict the nursing workload for a certain group of patients may help to allocate nursing capacity as efficiently as possible and thus to reduce costs, without endangering the patients safety and nurses' health. NAS is a validated tool for the measurement of nursing workload in an intensive care unit. Goal: This study was conducted to investigate whether it is possible to predict the nursing workload for a homogeneous group of patients, admitted after an in or out of hospital cardiac arrest and to assess the effects of baseline characteristics, vital parameters and admittance time on this workload. Method: We performed a retrospective analysis of NAS scores of all IHCA and OHCA patients admitted to our ICU from October 2012 until September 2015 during the first 48 hours of stay. The NAS was recorded per patient per nursing shift. We furthermore recorded patient characteristics and vital parameters.

Goal: This study was conducted to investigate whether it is possible to predict the nursing workload for a homogeneous group of patients, admitted after an in or out of hospital cardiac arrest and to assess the effects of baseline characteristics, vital parameters and admittance time on this workload. Method: We performed a retrospective analysis of NAS scores of all IHCA and OHCA patients admitted to our ICU from October 2012 until September 2015 during the first 48 hours of stay. The NAS was recorded per patient per nursing shift. We furthermore recorded patient characteristics and vital parameters. Results: During this period 386 patients, 273 males and 113 females, were admitted to the ICU after cardiac arrest. The mean age at admission was 63.4 years (SD = 14.8). The mean NAS at admission was 71.4 (SD = 22.4). Patients admitted in the evening shift had a significantly higher NAS compared to patients admitted in the night shift (74.3; SD = 21.2 vs 67.4; SD = 22.3 (p = 0.037)), but no significant difference was found with the day shift (mean NAS day shift: 71.7; SD = 23.3). After admission the workload decreased in all patients by a mean of 18.3 points (SD = 23.9; p < 0.001). A higher SOFA score, a higher PEEP and a lower pH at admittance resulted in a higher NAS score on average over time (p < 0.001).

nificant difference was found with the day shift (mean NAS day shift: 71.7; SD = 23.3). After admission the workload decreased in all patients by a mean of 18.3 points (SD = 23.9; p < 0.001). A higher SOFA score, a higher PEEP and a lower pH at admittance resulted in a higher NAS score on average over time (p < 0.001). Conclusions: The nursing workload at admission of patients after cardiac arrest is fairly predictable, with no clinically significant difference between shifts , necessitating a nurse-to-patient ratio of at least 1:1. After the first shift it is almost always possible to decrease the nurse-to-patient ratio to 1:2. The NAS was influenced by severity of illness. These results can be used to assess the needed nursing staff for the treatment of these patients for the first days after admittance. References Armstrong E et al. Using Nursing Activities Score to Assess Nursing Workload on a Medium Care Unit. Anesthesia and Analgesia. 2015;121:1274–1280 McGahan M et al. Nurse staffing levels and the incidence of mortality and morbidity in the adult intensive care unit: A literature review. Australian Critical Care.2012;25:64–77 Grillo Padilha K et al. Nursing Activities Score in the intensive care unit: Analysis of related factors. Intensive and Critical Care Nursing. 2008;24:197–204 Grant acknowledgement None.

McGahan M et al. Nurse staffing levels and the incidence of mortality and morbidity in the adult intensive care unit: A literature review. Australian Critical Care.2012;25:64–77 Grillo Padilha K et al. Nursing Activities Score in the intensive care unit: Analysis of related factors. Intensive and Critical Care Nursing. 2008;24:197–204 Grant acknowledgement None. A870 A comparison of the epidemiology, pathophysiology and clinical management of intubated patients treated in a central london emergency department resuscitation room with major trauma versus severe sepsis/septic shock P. Riozzi1, S. Helyar1, H. Cotton1, G. Hallows1, A. Noon1, C. Bell2, K. Peters2, A. Feehan2, J. Keep1, P.A. Hopkins1 1King's College Hospital, ACET Research Team, London, United Kingdom; 2King's College Hospital, Audit Team, King's Critical Care, London, United Kingdom Correspondence: P. Riozzi – King's College Hospital, ACET Research Team, London, United Kingdom Introduction: Major trauma and severe sepsis are both leading causes of admission to the resuscitation rooms in emergency departments across the world. Despite obvious differences in precipitating mechanism, there are surprising similarities between subsequent pathophysiology: both disorders lead to disorders of the macrocirculation, microcirculation and host inflammatory response (1,2).

f admission to the resuscitation rooms in emergency departments across the world. Despite obvious differences in precipitating mechanism, there are surprising similarities between subsequent pathophysiology: both disorders lead to disorders of the macrocirculation, microcirculation and host inflammatory response (1,2). Objectives: Here we compare the baseline epidemiology, pathophysiology, operational and clinical management of intubated resuscitation room patients with these two critical illness syndromes (Major Trauma/Septic shock). The results will be used to facilitate the design and planning of a study to test the feasibility/effectiveness of advanced monitoring systems (thromboelastography, oesophageal doppler flow monitoring, echocardiography, and microcirculatory monitoring) in the resuscitation room management of critically ill patients with these conditions. Methods: Institutional approval was gained to collect anonymised patient data over a 6-month period from a mixture of written and electronic records. Where appropriate, significance was tested by Mann Whitney U (Sigmaplot 11.0).

Objectives: Here we compare the baseline epidemiology, pathophysiology, operational and clinical management of intubated resuscitation room patients with these two critical illness syndromes (Major Trauma/Septic shock). The results will be used to facilitate the design and planning of a study to test the feasibility/effectiveness of advanced monitoring systems (thromboelastography, oesophageal doppler flow monitoring, echocardiography, and microcirculatory monitoring) in the resuscitation room management of critically ill patients with these conditions. Methods: Institutional approval was gained to collect anonymised patient data over a 6-month period from a mixture of written and electronic records. Where appropriate, significance was tested by Mann Whitney U (Sigmaplot 11.0). Results: 162 patients, intubated pre-hospital or in ED resus, were identified with trauma or sepsis diagnoses. Trauma patients were commoner (n = 128; 79 %) and more likely to be intubated pre-hospital (107/128; 83.6 % vs 27/34; 79 %). Lactate profiles were similar in the two groups at start and end of resus episode (Figure 18). Patients with major trauma were more hypertensive but equally tachycardic when compared with patients with severe sepsis/septic shock. Patients with septic shock/severe sepsis (median 202 minutes vs 128 minutes) spent longer in resuscitation room, but received less documented consultant-level review (90.6 % vs 20.6 %). Imaging of major trauma patients was with CT (128/128; 100 %) and ultrasound (44/128; 34.3 %) in contrast to septic patients (11/34 CT; 32.4 %; 2/34, 1 % US). ICU and hospital mortality was higher in patients with sepsis (38.2 % vs 14.8 %), but death in resus only occurred in the trauma population. Only two patients, both with facial trauma, would have had a relative contraindication to the proposed advanced monitoring.

st to septic patients (11/34 CT; 32.4 %; 2/34, 1 % US). ICU and hospital mortality was higher in patients with sepsis (38.2 % vs 14.8 %), but death in resus only occurred in the trauma population. Only two patients, both with facial trauma, would have had a relative contraindication to the proposed advanced monitoring. Conclusions: This novel preliminary work has highlighted some important differences between the epidemiology, outcomes, pathophysiology and clinical/operational management of intubated patients with severe sepsis versus major trauma. These will influence the conduct and outcome measures of any trial of advanced monitoring in this setting. However, contraindications to any of the advanced monitoring technologies being considered were rare and no obvious barriers to the planned study of advanced monitoring were identified. References (1) Lord JM et al. 2014. The systemic immune response to trauma: an overview of pathophysiology and treatment. Lancet 384: 1455–65. (2) Dewar D et al. 2009. Post injury multiple organ failure. Injury 40: 912–8.Fig. 18 (abstract A870). Lactate (Sepsis vs Trauma)

Conclusions: This novel preliminary work has highlighted some important differences between the epidemiology, outcomes, pathophysiology and clinical/operational management of intubated patients with severe sepsis versus major trauma. These will influence the conduct and outcome measures of any trial of advanced monitoring in this setting. However, contraindications to any of the advanced monitoring technologies being considered were rare and no obvious barriers to the planned study of advanced monitoring were identified. References (1) Lord JM et al. 2014. The systemic immune response to trauma: an overview of pathophysiology and treatment. Lancet 384: 1455–65. (2) Dewar D et al. 2009. Post injury multiple organ failure. Injury 40: 912–8.Fig. 18 (abstract A870). Lactate (Sepsis vs Trauma) A871 Impact of an education programme for novice nurses on staff work engagement in a UK ICU K. Churchill1, K. Hawkins1, R. Brook1, N. Paver1, R. Endacott2,3 1Royal Devon and Exeter Hospital, ICU, Exeter, United Kingdom; 2Plymouth University/Royal Devon and exeter Hospital Clinical School, Exeter, United Kingdom; 3Monash University, School of Nursing and Midwifery, Melbourne, Australia Correspondence: K. Churchill – Royal Devon and Exeter Hospital, ICU, Exeter, United Kingdom Introduction: In our 15-bedded GICU, demand for beds has increased while recruitment of ICU trained nurses has decreased. There is enduring evidence of links between workload and stress [1, 2], with high levels of burnout reported in ICU nurses [3]. There has been a shift towards measuring what makes people positive and engaged [4], rather than why people reach the extreme state of burn out. It is important to understand factors that affect work engagement to develop strategies that enhance nurse retention and improve the quality of ICU patient care.

U nurses [3]. There has been a shift towards measuring what makes people positive and engaged [4], rather than why people reach the extreme state of burn out. It is important to understand factors that affect work engagement to develop strategies that enhance nurse retention and improve the quality of ICU patient care. Objectives: To examine the impact of an education initiative for novice ICU nurses on work engagement for the ICU nursing staff and organisational resource use. Methods: A pre - post design was used to collect data from all ICU nurses at the start of the education programme and at 6 months following he intervention. Work engagement was measured using the self-report 17 item Utrecht Work Engagement Scale (UWES) [4] with an open question to capture staff experiences. Organisational impact was measured using levels of sick leave, % staff turnover, use of agency nurses and staff recruitment.

and at 6 months following he intervention. Work engagement was measured using the self-report 17 item Utrecht Work Engagement Scale (UWES) [4] with an open question to capture staff experiences. Organisational impact was measured using levels of sick leave, % staff turnover, use of agency nurses and staff recruitment. Results: Fifty three ICU nurses completed the pre-intervention survey (61 % response) and 42 completed the post-intervention survey (46 % response). Respondents had reasonable years of ICU experience (Mean7.57, SD 8.1) and time in current post (Mean6.21, SD 8.2). Internal consistency for the UWES was high (alpha .91). Levels of work engagement (Mean [SD]) increased (3.94, [0.64] vs 4.03 [0.69]) but did not reach significance and remained in the 'average' band as judged by the scale authors [4]. When examined by senior and junior nurses, the increase was similar. Organisational measures showed decrease in sick leave, turnover, agency use and increase in recruitment of experienced ICU nurses. Qualitative feedback was positive, with perceptions of improvement in unit morale due to time being invested in the individual and reduced stress and workload for shift leaders. Conclusions: Providing education for the newest ICU recruits can have benefits for the whole ICU team. However, it is important to examine how work engagement might be further improved. References 1. Schaufeli W, et al. In : Miranda DR, et al. (eds) Organisation and Management of Intensive Care : a prospective study in 12 European countries. 1998; Springer, Berlin

Conclusions: Providing education for the newest ICU recruits can have benefits for the whole ICU team. However, it is important to examine how work engagement might be further improved. References 1. Schaufeli W, et al. In : Miranda DR, et al. (eds) Organisation and Management of Intensive Care : a prospective study in 12 European countries. 1998; Springer, Berlin 2. Gurses AP et al. Impact of performance obstacles on intensive care nurses' workload, perceived quality and safety of care and quality of working life. Health Services Research 2009; 44(2): 422–443 3. Poncet MC et al. Burnout syndrome in critical care nursing staff. Am J Resp Crit Care Med 2007; 175: 698–704 4. Bakker, A. B. et al. Work engagement: An emerging concept in occupational health psychology. Work and Stress 2008, 22(3), 187–200. Grant acknowledgement None

2. Gurses AP et al. Impact of performance obstacles on intensive care nurses' workload, perceived quality and safety of care and quality of working life. Health Services Research 2009; 44(2): 422–443 3. Poncet MC et al. Burnout syndrome in critical care nursing staff. Am J Resp Crit Care Med 2007; 175: 698–704 4. Bakker, A. B. et al. Work engagement: An emerging concept in occupational health psychology. Work and Stress 2008, 22(3), 187–200. Grant acknowledgement None A872 The incidence of silent aspiration on intensive care N. Maistry Royal Brompton and Harefield NHS Foundation Trust, Rehabilitation and Therapies, London, United Kingdom Introduction: The incidence of Dysphagia on intensive care is an area of growing research. Dysphagia is associated with aspiration pneumonia and increased ICU bed days. In general, Speech and Language Therapy (SLT), makes recommendations based on the results of a clinical bedside swallowing evaluation, despite the unreliability of this method 1. This is largely due to the difficulty accessing gold standard assessment methods such as Videofluoroscopy (VF) and Fibreoptic Endoscopic Evaluation of Swallowing (FEES)3. Referral for these assessment methods are based on a local defined criteria. This study evaluates the incidence of silent aspiration identified by VF and FEES in a 20 bedded tertiary Cardio-Respiratory Intensive Care Unit.

methods such as Videofluoroscopy (VF) and Fibreoptic Endoscopic Evaluation of Swallowing (FEES)3. Referral for these assessment methods are based on a local defined criteria. This study evaluates the incidence of silent aspiration identified by VF and FEES in a 20 bedded tertiary Cardio-Respiratory Intensive Care Unit. Objectives: To determine the incidence of silent aspiration, defined as “aspiration before, during, or after swallowing in the absence of cough or visible signs of choking and distress 2," in ICU patients assessed by VF or FEES between July 2014 and June 2015. Method: Data was retrospectively reviewed for 12 month period from all ICU referrals made to SLT for swallowing evaluation. All patients received a clinical bedside swallowing evaluation. The results are presented as percentages and counts for patients receiving VF and FEES that silently aspirated. Results: A total of 97 patients were referred for swallowing assessment and 35 % (34/97) had a VF or FEES. There were 27 males and ages were 44.1 ± 15.7 years. In this group, 22 patients had Videofluoroscopic assessments and 12 patients had FEES. In the VF group 77 % silently aspirated whilst in the FEES group the values were 83 %. 27 patients (79 %) silently aspirated during objective assessment, impacting on how and when oral feeding was commenced. Conclusion: This study suggests that silent aspiration is highly prevalent in this population group. Consequently, VF and FEES should be part of standard routine assessment in the management of critically ill patients. References

Results: A total of 97 patients were referred for swallowing assessment and 35 % (34/97) had a VF or FEES. There were 27 males and ages were 44.1 ± 15.7 years. In this group, 22 patients had Videofluoroscopic assessments and 12 patients had FEES. In the VF group 77 % silently aspirated whilst in the FEES group the values were 83 %. 27 patients (79 %) silently aspirated during objective assessment, impacting on how and when oral feeding was commenced. Conclusion: This study suggests that silent aspiration is highly prevalent in this population group. Consequently, VF and FEES should be part of standard routine assessment in the management of critically ill patients. References 1. Romano M, Schultz T, Tai A. Diagnostic Accuracy of clinical swallow assessment for oropharyngeal aspiration: a systematic review. JBI Library of Systematic Reviews, 2012 Vol 10 2. Bruce K. Rubin. The Cruelest Lies Are Often Told in Silence. Chest September 2011, Vol 140, No. 3 3. Jeronimo, B. Critical Review: Is the endoscopic swallowing assessment more sensitive than the videofluoroscopic swallowing assessment at identifying penetration or aspiration in adults with dysphagia 2013

1. Romano M, Schultz T, Tai A. Diagnostic Accuracy of clinical swallow assessment for oropharyngeal aspiration: a systematic review. JBI Library of Systematic Reviews, 2012 Vol 10 2. Bruce K. Rubin. The Cruelest Lies Are Often Told in Silence. Chest September 2011, Vol 140, No. 3 3. Jeronimo, B. Critical Review: Is the endoscopic swallowing assessment more sensitive than the videofluoroscopic swallowing assessment at identifying penetration or aspiration in adults with dysphagia 2013 A873 Delerium related incidents at the ICU and nursing aspects A. van Wijk1, N. Rouw1, T. van Galen2, S. Evelein-Brugman3 1Senior ICU Nurse Quality and Safety VU University Medical Center, Amsterdam, Netherlands; 2ICU Nursing,VU University Medical Center, Amsterdam, Netherlands; 3VU University Medical Center, Amsterdam, Netherlands Correspondence: A. van Wijk – Senior ICU Nurse Quality and Safety VU University Medical Center, Amsterdam, Netherlands Introduction: At the ICU of VU University Medical Center (VUmc) nurses are frequently confronted with delirium1,2. Delirium is known to be present in 60-80 % of mechanical ventilated patients and 20-50 % in non-ventilated patients. Immediate consequences are falling incidents or for patients to remove tubes and IV lines that are necessary for treatment. In literature, this is stated as a result of treatment, but often data is missing. Consequences of removal are increased risk of complications3, prolonged mechanical ventilation, LOS and increased morbidity/mortality4.

are falling incidents or for patients to remove tubes and IV lines that are necessary for treatment. In literature, this is stated as a result of treatment, but often data is missing. Consequences of removal are increased risk of complications3, prolonged mechanical ventilation, LOS and increased morbidity/mortality4. Objectives: To measure the frequency of removing tubes, lines and falling incidents related to delirium. Methods: A multidisciplinary focus group was formed (2014) in order to properly diagnose, prevent and/or treat delirium due to the high prevalence. The first steps were increasing awareness and implementing the CAM-ICU score. To clarify delirium-related incidents a one year period was set in which the dedicated senior nurse informed and trained the nursing staff regarding delirium and potential risks. To register delirium-related incidents a modified report button was built in the EPR (Metavision, IMD Soft) and used beside the regular incident reporting system5. Results: After one year, 52 individual patient incidents were reported concerning falling or tube or IV line removal. This included 14 gastric tubes, 15 airway tubes, 14 IV/CVC/arterial lines, 5 other lines and 4 fall incidents. In 83 % of the cases the patient was diagnosed with delirium. 4 out of 5 patients received medication or were fixated before the incident despite a 40 % CAM-ICU registration rate. Because the focus group doubted about underreporting 10 nurses were interviewed if the results corresponded with their experience. They were unanimous that there was hardly any underreporting.

delirium. 4 out of 5 patients received medication or were fixated before the incident despite a 40 % CAM-ICU registration rate. Because the focus group doubted about underreporting 10 nurses were interviewed if the results corresponded with their experience. They were unanimous that there was hardly any underreporting. Discussion: Despite therapy or fixation delirium-related incidents occur on a weekly basis at our ICU, causes harm and increases nursing workload. Although the incidence rate is presumed to be low, there is no feeling of satisfaction. Further improvement is necessary due to the high risks for the patient. Therefore, we need to be able to diagnose incidents faster so we can start treatment sooner. Although the CAM-ICU score was implemented, compliance is insufficient. Increasing compliance is the first step to further improvement. The follow-up question is whether delirium-related injury can be reduced when CAM-ICU compliance improves. Second step is to investigate the effectiveness of our fixation protocol.

er. Although the CAM-ICU score was implemented, compliance is insufficient. Increasing compliance is the first step to further improvement. The follow-up question is whether delirium-related injury can be reduced when CAM-ICU compliance improves. Second step is to investigate the effectiveness of our fixation protocol. A874 Prospective study to determine the predictors of extubation success A. Taggu1, B. Krishna2, S. Sampath3 1St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India; 2St. Johns Medical College Hospital, Bangalore, India; 3St. Johns Medical College, Bangalore, India Correspondence: A. Taggu – St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India Introduction: Timely extubation is crucial in critically ill patients. Traditional indices like rapid shallow breathing index are considered as accurate during the spontaneous breathing trial. Multiple other proposed parameters like diaphragm thickness, fluid balance and cardiac indices have been shown to predict succesful extubation in the recent years. Objectives: To assess the reliability of the parameters in predicting successful extubation. Methods: A prospective observational study done on 220 adult patients eligible for extubation as decided by the attending intensivists.

A874 Prospective study to determine the predictors of extubation success A. Taggu1, B. Krishna2, S. Sampath3 1St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India; 2St. Johns Medical College Hospital, Bangalore, India; 3St. Johns Medical College, Bangalore, India Correspondence: A. Taggu – St. Johns Medical College Hospital, Critical Care Medicine, Bangalore, India Introduction: Timely extubation is crucial in critically ill patients. Traditional indices like rapid shallow breathing index are considered as accurate during the spontaneous breathing trial. Multiple other proposed parameters like diaphragm thickness, fluid balance and cardiac indices have been shown to predict succesful extubation in the recent years. Objectives: To assess the reliability of the parameters in predicting successful extubation. Methods: A prospective observational study done on 220 adult patients eligible for extubation as decided by the attending intensivists. Exclusion criteria: Pregnant and tracheostomised patients. Along with baseline parameters, following measurements were taken pre and post extubation.Cardiac parameters including Simpsons method for ejection fraction, E/A, E/e' (lateral) for diastolic function, TAPSE and TAD for Right ventricular function. All recordings were taken just before extubation and within six hours post extubation.

eline parameters, following measurements were taken pre and post extubation.Cardiac parameters including Simpsons method for ejection fraction, E/A, E/e' (lateral) for diastolic function, TAPSE and TAD for Right ventricular function. All recordings were taken just before extubation and within six hours post extubation. Just before extubation,high frequency linear ultrasound probe was used to measure the right sided DT at the zone of apposition (ZOA) between 8th to 10th intercostal spaces in mid-axillary line.The change in DT fraction(Δdtfrac_pre%) was calculated as DT(end-inspiration)-DT(end-expiration)/DT (end-expiration)x100.RSBI was simultaneously recorded. Fluid balance 24 hours were recorded. Statistics: Logistic regression was used to develop a model with extubation failure as the outcome and change in DT(delta fraction), RSBI, E/e', E/A , fluid balance 24 hrs and other patient covariates as predictors. Extubation failure was defined as re-intubation within 48 hrs of extubation. Results: In the logistic model, delta fraction was the only significant predictor of extubation success among all covariates (const 8.985, Coef.-0.2715, p value 0.00). The model showed very good discrimination (receiver operating curve, ROC area of 0.94.4) but poor calibration (Hosmer- Lemeshow chi2 (3) = 206.53, Prob > chi2 = 0.000). Conclusions: Among all the parameters studied, diaphragm thickness change just before extubation reliably predicted the extubation success. References

Results: In the logistic model, delta fraction was the only significant predictor of extubation success among all covariates (const 8.985, Coef.-0.2715, p value 0.00). The model showed very good discrimination (receiver operating curve, ROC area of 0.94.4) but poor calibration (Hosmer- Lemeshow chi2 (3) = 206.53, Prob > chi2 = 0.000). Conclusions: Among all the parameters studied, diaphragm thickness change just before extubation reliably predicted the extubation success. References Matamis D, Soilemezi E, Tsagourias M, Akoumianaki E, Dimassi S, Boroli F, Richard JC, Brochard L (2013) Sonographic evaluation of the diaphragm in critically ill patients. Technique and clinical applications. Intensive Care Med 39:801–810Table 17 (abstract A874). Baseline characteristics Parameters Extubation success n = 186 Extubation failure n = 34 P value Age (years) 52.40(sD 16.14) 56.10((sD 16.20) 0.56 Male 108/78 25/9 0.60 Rapid shallow breathing index- median( IQR) 48(39–82) 54(40–82) 0.06 Mean arterial pressure 69.4((sD2.46) 67.56((sD2.78) 0.34 Mechanical ventilation days-median (IQR) 4(6–11) 7(9–14) 0.49 Fluid balance 24 hrs-median (IQR) ml 110(1200–760) 320(900–1500) 0.05 E/e’ (lateral) 7.5((sD 3.8) 9.2((sD3.16) 0.001 E/A 1.92((sD1.2) 1.10((sD0.98) 0.001 Delta fraction of diaphragm thickness 46.35((sD11.7) 30.4((sD12.4) 0.000 Table 18 (abstract A874). Logistic regression diaphragm thickness fraction

6–11) 7(9–14) 0.49 Fluid balance 24 hrs-median (IQR) ml 110(1200–760) 320(900–1500) 0.05 E/e’ (lateral) 7.5((sD 3.8) 9.2((sD3.16) 0.001 E/A 1.92((sD1.2) 1.10((sD0.98) 0.001 Delta fraction of diaphragm thickness 46.35((sD11.7) 30.4((sD12.4) 0.000 Table 18 (abstract A874). Logistic regression diaphragm thickness fraction Delta fraction (Diaphragm thickness change) Extubation success Extubation failure Significance >25 % 183(98.4 %) 2(5.8 %) Fisher’s exact test <25 % 3(1.6 %) 32(94.2 %) p = 0.000 Total 186 34 220 Fig. 19 (abstract A874). ROC curves of variables predicting extubation Grant acknowledgement None

6–11) 7(9–14) 0.49 Fluid balance 24 hrs-median (IQR) ml 110(1200–760) 320(900–1500) 0.05 E/e’ (lateral) 7.5((sD 3.8) 9.2((sD3.16) 0.001 E/A 1.92((sD1.2) 1.10((sD0.98) 0.001 Delta fraction of diaphragm thickness 46.35((sD11.7) 30.4((sD12.4) 0.000 Table 18 (abstract A874). Logistic regression diaphragm thickness fraction Delta fraction (Diaphragm thickness change) Extubation success Extubation failure Significance >25 % 183(98.4 %) 2(5.8 %) Fisher’s exact test <25 % 3(1.6 %) 32(94.2 %) p = 0.000 Total 186 34 220 Fig. 19 (abstract A874). ROC curves of variables predicting extubation Grant acknowledgement None ADVANCES IN GENERAL INTENSIVE CARE A875 Blood purification with continuous venovenous hemofiltration in septic and ARDS patients: a meta-analysis of randomized trials A. Putzu1, M. Fang2,3, M. Boscolo Berto1, A. Belletti2, T. Cassina1, L. Cabrini2 1Fondazione Cardiocentro Ticino, Cardiac Anesthesia and Intensive Care, Lugano, Switzerland; 2IRCCS San Raffaele Scientific Institute, Department of Anesthesia and Intensive Care, Milan, Italy; 33rd Hospital of HeBei Medical University, Department of Intensive Care Medicine, Shi Jiazhuang, China Correspondence: A. Putzu – Fondazione Cardiocentro Ticino, Cardiac Anesthesia and Intensive Care, Lugano, Switzerland Introduction: Severe inflammatory conditions, as sepsis and acute respiratory distress syndrome (ARDS), are common among critically ill patients. Despite healthcare improvement, mortality still remains high and few therapies are nowadays available. Attenuating inflammation through direct removal of inflammatory mediators from circulation appears to be intuitively. Theoretically, continuous venovenous hemofiltration (CVVH) could remove part of the inflammatory mediators and/or bacterial toxins.

mortality still remains high and few therapies are nowadays available. Attenuating inflammation through direct removal of inflammatory mediators from circulation appears to be intuitively. Theoretically, continuous venovenous hemofiltration (CVVH) could remove part of the inflammatory mediators and/or bacterial toxins. Objectives: The objective of this systematic review and meta-analysis of randomized trials was to evaluate if CVVH reduces mortality in critically ill patients with sepsis or ARDS. Methods: PubMed, EMBASE, and the Cochrane Central Register were searched for pertinent studies (last updated February 1st, 2016). We included randomized trials on the use of CVVH as blood purification technique in comparison to standard therapy in adult critical ill patients with sepsis or ARDS. Trials in which primary indication for CVVH was acute kidney injury needing renal replacement therapy were excluded. Primary endpoint was mortality at longest follow-up available. Results: Ten studies published between 2002 and 2015, randomizing 623 patients, were included in the analysis. The included trials had moderate to high risk of bias. Patients who received blood purification with CVVH showed significantly lower mortality compared with those who received conventional therapy (86/323 [26.63 %] patients in the CVVH group vs. 120/300 [40 %] in the conventional therapy group, OR 0.54 [95 % CI, 0.38, 0.77], p for effect = 0.0006, I2 = 4 %, Number Needed to Treat = 8) (Figure 20).

od purification with CVVH showed significantly lower mortality compared with those who received conventional therapy (86/323 [26.63 %] patients in the CVVH group vs. 120/300 [40 %] in the conventional therapy group, OR 0.54 [95 % CI, 0.38, 0.77], p for effect = 0.0006, I2 = 4 %, Number Needed to Treat = 8) (Figure 20). The results were confirmed when including only studies performed in a septic setting (7 trials and 413 patients, OR 0.57 [95 % CI, 0.36, 0.89], p for effect = 0.01, I2 = 34 %) or in ARDS patients (3 trials and 210 patients, OR 0.49 [95 % CI, 0.28, 0.88], p for effect = 0.02, I2 = 0 %) (Figure 20). We found no significant difference on mortality when limiting the results just to trial employing CVVH at hemofiltration rate lower or higher then 50 ml kg−1 h−1. Conclusions: Blood purification with CVVH might be associated with a significant reduction in mortality when performed in patients with sepsis or ARDS. This is the first meta-analysis suggesting beneficial effects of CVVH on mortality and we could suppose that the beneficial effects of CVVH in these inflammatory conditions could arise from the immunomodulatory properties of hemofiltration. Further high-quality randomized controlled trials adequate powered for mortality are needed to clarify the impact of CVVH on these inflammatory conditions. Grant acknowledgement The authors declare no support or funding and no potential conflict of interest.Fig. 20 (abstract A875). Forest plot for the risk of mortality at longest follow up available

Conclusions: Blood purification with CVVH might be associated with a significant reduction in mortality when performed in patients with sepsis or ARDS. This is the first meta-analysis suggesting beneficial effects of CVVH on mortality and we could suppose that the beneficial effects of CVVH in these inflammatory conditions could arise from the immunomodulatory properties of hemofiltration. Further high-quality randomized controlled trials adequate powered for mortality are needed to clarify the impact of CVVH on these inflammatory conditions. Grant acknowledgement The authors declare no support or funding and no potential conflict of interest.Fig. 20 (abstract A875). Forest plot for the risk of mortality at longest follow up available A876 Clinical significance of circulating nucleosomes in disseminated intravascular coagulation M. Mistry1, Y. Alhamdi2, I. Welters2,3, S.T. Abrams2, C.H. Toh2 1University of Liverpool, Institute of Ageing and Chronic Disease, Liverpool, United Kingdom; 2University of Liverpool, Institute of Infection and Global Health, Liverpool, United Kingdom; 3Royal Liverpool University Hospital, Liverpool, United Kingdom Correspondence: M. Mistry – University of Liverpool, Institute of Ageing and Chronic Disease, Liverpool, United Kingdom Introduction: Disseminated intravascular coagulation (DIC) is a life-threatening condition that occurs secondary to many underlying diseases which initiate a massive inflammatory response. This damages the microvascular endothelium and ultimately leads to increased cellular damage. During cell death, chromatin is cleaved into nucleosomes, which are released extracellularly and further degraded into individual histones and free DNA (1). Recent studies have shown circulating histones can be extremely toxic by increasing thrombin generation, platelet aggregation and endothelial damage. Nucleosomes are also elevated in numerous critical illnesses associated with the development of DIC.

larly and further degraded into individual histones and free DNA (1). Recent studies have shown circulating histones can be extremely toxic by increasing thrombin generation, platelet aggregation and endothelial damage. Nucleosomes are also elevated in numerous critical illnesses associated with the development of DIC. Objectives: This study was designed to elucidate the role of nucleosomes in DIC and to analyse their potential value for diagnosing DIC and predicting patient outcome. Methods: A retrospective analysis of 170 patients admitted to the intensive care unit (ICU) of Royal Liverpool University Hospital (Liverpool, UK) was performed. Citrated blood samples were collected for four days after ICU admission and centrifuged to obtain plasma. Routine coagulation markers were measured. Nucleosomes were measured by enzyme-linked immunosorbent assay. Patients with a score of ≥5 by the ISTH Sub-committee scoring system were diagnosed as having overt-DIC. Statistical analysis was performed using SPSS 22.0 for Windows. Two sided p-values of < 0.05 were considered to be statistically significant.

measured. Nucleosomes were measured by enzyme-linked immunosorbent assay. Patients with a score of ≥5 by the ISTH Sub-committee scoring system were diagnosed as having overt-DIC. Statistical analysis was performed using SPSS 22.0 for Windows. Two sided p-values of < 0.05 were considered to be statistically significant. Results: According to the ISTH DIC scoring criteria, 28 out of the total 106 patients (26.4 %) were diagnosed with overt-DIC and 26 patients (24.5 %) were classified as non survivors. The circulating levels of nucleosomes over the four days from intensive care unit (ICU) admission were not significantly elevated in overt-DIC patients compared to ICU controls (P = 0.58). There was also no significant difference observed in nucleosome levels between survivors and non survivors, indicating no prognostic value. Nucleosomes did not appear to correlate with any of the clotting tests or variables predictive for disease severity or patient outcome. Nucleosomes also showed little predictive value in the diagnosis of DIC (AUC = 0.535 (95 % CI 0.405-0.666, P = 0.583)) Conclusions: Patients with overt DIC do not have elevated plasma levels of nucleosomes. Nucleosomes levels did not appear to change significantly in patients with overt DIC, showing no independent prognostic significance or predictive value in DIC. These findings provide an interesting insight into the differing roles of histones and nucleosomes in DIC pathogenesis; histones may not be toxic when bound in a complex as nucleosomes. References

Conclusions: Patients with overt DIC do not have elevated plasma levels of nucleosomes. Nucleosomes levels did not appear to change significantly in patients with overt DIC, showing no independent prognostic significance or predictive value in DIC. These findings provide an interesting insight into the differing roles of histones and nucleosomes in DIC pathogenesis; histones may not be toxic when bound in a complex as nucleosomes. References (1) Holdenrieder S, Stieber P, Bodenmuller H, Fertig G, Furst H, Schmeller N, et al. Nucleosomes in serum as a marker for cell death. Clinical chemistry and laboratory medicine. 2001;39(7):596–605. A877 Which factor dose affect the recovery from extreme hyperbilirubinemia in critically ill patients? H.-S. Han1, E.M. Gil2, D.-S. Lee2, C.-M. Park1 1Samsung Medical Center, Surgery, Seoul, Republic of Korea; 2Samsung Medical Center, Critical Care Medicine, Seoul, Republic of Korea Correspondence: H.-S. Han – Samsung Medical Center, Surgery, Seoul, Republic of Korea Introduction: Extreme hyperbilirubinemia occurs rarely in critically ill patients. It is usually caused due to hepatic failure or cholestasis with various causes and has a dreadful clinical course without recovery. However, only a few reports are available regarding the clinical course or prognostic factors of extreme hyperbilirubinemia in critically ill patients. Objectives: We evaluated the clinical course and various factors affecting the recovery and survival from extreme hyperbilirubinemia in critically ill patients.

A877 Which factor dose affect the recovery from extreme hyperbilirubinemia in critically ill patients? H.-S. Han1, E.M. Gil2, D.-S. Lee2, C.-M. Park1 1Samsung Medical Center, Surgery, Seoul, Republic of Korea; 2Samsung Medical Center, Critical Care Medicine, Seoul, Republic of Korea Correspondence: H.-S. Han – Samsung Medical Center, Surgery, Seoul, Republic of Korea Introduction: Extreme hyperbilirubinemia occurs rarely in critically ill patients. It is usually caused due to hepatic failure or cholestasis with various causes and has a dreadful clinical course without recovery. However, only a few reports are available regarding the clinical course or prognostic factors of extreme hyperbilirubinemia in critically ill patients. Objectives: We evaluated the clinical course and various factors affecting the recovery and survival from extreme hyperbilirubinemia in critically ill patients. Methods: A retrospective study was performed at a single center from 2006 to 2015. We defined extreme hyperbilirubinemia as a state of total bilirubin above 20 mg/dl and selected all patients whose serum total bilirubin increased above 20 mg/dl at least once during their stay in the intensive care unit. We investigated the overall clinical course of the patients and compared the differences between one group with normalization of total bilirubin (recovery group) and the other group without normalization (non-recovery group). Furthermore, we evaluated the association between prognosis and various clinical factors, including the peak total bilirubin levels, increasing rate of total bilirubin (Vi), results of laboratory analyses related to hepatic function, and clinical features at the time of extreme hyperbilirubinemia. These data were analyzed using Chi-square test and Cox and logistic regression analyses.

ious clinical factors, including the peak total bilirubin levels, increasing rate of total bilirubin (Vi), results of laboratory analyses related to hepatic function, and clinical features at the time of extreme hyperbilirubinemia. These data were analyzed using Chi-square test and Cox and logistic regression analyses. Results: In total, 610 patients experienced extreme hyperbilirubinemia. The mean age was 56.4 ± 14.0 years. The number of recovery group was 46 (7.5 %) and non-recovery group was 564 (92.5 %). A high Vi, young age, and use of hepatotonic agents were identified as influential factors for the recovery from extreme hyperbilirubinemia, whereas obesity, diabetes mellitus, hypertension, and dyslipidemia were the unfavorable factors. The mortality rate was 86.89 % of total patients, and there were no significant differences in mortality rate according to the highest level of total bilirubin. The Vi (95 % CI 1.070-1.221, p < 0.001), serum level of aspartate aminotransferase (95 % CI 1.000-1.001, p = 0.005) and lactate (95 % CI 1.125-1.298, p < 0.001), and use of vasoactive drugs (95 % CI 1.014-1.943, p = 0.04) were identified as significant risk factors for mortality, whereas the use of hepatotonic agents decreased the mortality risk (95 % CI 0.610-0.887, p = 0.001).

tate aminotransferase (95 % CI 1.000-1.001, p = 0.005) and lactate (95 % CI 1.125-1.298, p < 0.001), and use of vasoactive drugs (95 % CI 1.014-1.943, p = 0.04) were identified as significant risk factors for mortality, whereas the use of hepatotonic agents decreased the mortality risk (95 % CI 0.610-0.887, p = 0.001). Conclusions: The mortality rate of patients with extreme hyperbilirubinemia was exceedingly high. This study identified the various factors related to the recovery and mortality from extreme hyperbilirubinemia. We expect that the results of this study will be helpful in treating and predicting the prognosis of patients with extreme hyperbilirubinemia. References 1. Vishal Bansal, MD. et al. Jaundice in the Intensive Care Unit. Surg Clin N Am 86 2006;1495–1502 Grant acknowledgement Thanks for H-R Jung, W-S Han, S-H Jung

Conclusions: The mortality rate of patients with extreme hyperbilirubinemia was exceedingly high. This study identified the various factors related to the recovery and mortality from extreme hyperbilirubinemia. We expect that the results of this study will be helpful in treating and predicting the prognosis of patients with extreme hyperbilirubinemia. References 1. Vishal Bansal, MD. et al. Jaundice in the Intensive Care Unit. Surg Clin N Am 86 2006;1495–1502 Grant acknowledgement Thanks for H-R Jung, W-S Han, S-H Jung A878 Are intravascular thermoregulation devices associated with the development of venous thrombo-embolic phenomenon? S. Winder-Rhodes, R. Lotay, J. Doyle Royal Surrey County Hospital NHS Foundation Trust, Department of Intensive Care Medicine and Surrey Peri-Operative Anaesthesia and Critical Care Collaborative Research Group (SPACER), Guildford, United Kingdom Correspondence: S. Winder-Rhodes – Royal Surrey County Hospital NHS Foundation Trust, Department of Intensive Care Medicine and Surrey Peri-Operative Anaesthesia and Critical Care Collaborative Research Group (SPACER), Guildford, United Kingdom Introduction: Intravascular thermoregulation devices (ITDs) have become increasingly popular in recent years and have proved effective in critically ill patients for achieving targeted temperature management1. Previous reports, however, have suggested an association between ITDs and the development of venous thrombosis2.

troduction: Intravascular thermoregulation devices (ITDs) have become increasingly popular in recent years and have proved effective in critically ill patients for achieving targeted temperature management1. Previous reports, however, have suggested an association between ITDs and the development of venous thrombosis2. Objectives: The aim of this quality improvement initiative was to determine the prevalence of venous thrombosis amongst patients in the 28-bed Intensive Care Unit at the Royal Surrey County Hospital (RSCH, UK) who had received ITDs (Thermoguard XP temperature management system3). Methods: Patients with ITDs inserted on the Intensive Care Unit between April 2014 and April 2016 were identified retrospectively from a departmental record of procedures. The electronic patient records, discharge summaries and radiological imaging for each patient were reviewed for evidence of a venous thrombus following ITD insertion. Results: Of twenty-five patients who received ITDs between April 2014 and April 2016, three had subsequent radiological evidence of venous thrombosis. Of these, one had the ITD inserted to achieve therapeutic hypothermia (target 33 °C) and the others, for the prevention of pyrexia (target 37.5 - 38 °C). In all three patients, ITDs were sited in femoral veins. Thrombi were identified in the femoral vein local to the site of ITD placement in one patient and involved the IVC in the other two; one patient later had evidence of pulmonary embolism. Two of the three cases had ITDs in situ for more than 72 hours.

.5 - 38 °C). In all three patients, ITDs were sited in femoral veins. Thrombi were identified in the femoral vein local to the site of ITD placement in one patient and involved the IVC in the other two; one patient later had evidence of pulmonary embolism. Two of the three cases had ITDs in situ for more than 72 hours. Conclusions: At least 12 % of patients managed with ITDs in the Intensive Care Unit at the RSCH (UK) had radiological evidence of venous thrombosis. This association justifies a high index of clinical suspicion and low threshold for investigating venous thrombosis in patients with ITDs, and warrants further exploration in larger patient groups. References 1. Cooling techniques for targeted temperature management post-cardiac arrest. Vaity C, Al-Subaie N, Cecconi M. Critics Care. 2015; 19(1): 103. 2. Thermoregulatory catheter-associated inferior vena cava thrombus. Gierman JL, Shutze WP Sr, Pearl GJ, Foreman ML, Hohmann SE, Shutze WP Jr. Proc (Bayl Univ Med Cent). 2013;26(2):100–2. 3. IVTM Intravascular Temperature Management Catheter Specifications. http://www.zoll.com/uploadedFiles/Public_Site/Products/Catheters/Catheters_spec_sheet.pdf. Accessed April 2016.

2. Thermoregulatory catheter-associated inferior vena cava thrombus. Gierman JL, Shutze WP Sr, Pearl GJ, Foreman ML, Hohmann SE, Shutze WP Jr. Proc (Bayl Univ Med Cent). 2013;26(2):100–2. 3. IVTM Intravascular Temperature Management Catheter Specifications. http://www.zoll.com/uploadedFiles/Public_Site/Products/Catheters/Catheters_spec_sheet.pdf. Accessed April 2016. A879 Pro-inflammatory macrophage regulates MMP1 and MMP10 expression in pulmonary arterial hypertension via the ERK and JNK signal pathways M.-W. Ke1, W.-C. Huang1, C.-H. Chiang2, W.-T. Hung1, C.-C. Cheng2, K.-C. Lin1, S.-C. Lin1, K.-R. Chiou2, S.-R. Wann1, C.-W. Shu3, P.-L. Kang2, G.-Y. Mar2, C.-P. Liu2 1Kaohsiung Veterans General Hospital, Critical Care Division, Kaohsiung City, Taiwan, Province of China; 2Kaohsiung Veterans General Hospital, Cardiovascular Division, Kaohsiung City, Taiwan, Province of China; 3Kaohsiung Veterans General Hospital, Department of Medical Education and Research, Kaohsiung City, Taiwan, Province of China Correspondence: C.-H. Chiang – Kaohsiung Veterans General Hospital, Cardiovascular Division, Kaohsiung City, Taiwan, Province of China Introduction: Pulmonary arterial hypertension (PAH) is a fatal disease characterized by vascular remodeling of pulmonary arteries. The remodeling is associated with abnormal proliferation of pulmonary arterial smooth muscle cells (PASMCs), deposition of extracellular matrix proteins and perivascular inflammation. Matrix metalloproteinases (MMPs) are critical to the maintenance of the homeostasis of extracellular environment.

of pulmonary arteries. The remodeling is associated with abnormal proliferation of pulmonary arterial smooth muscle cells (PASMCs), deposition of extracellular matrix proteins and perivascular inflammation. Matrix metalloproteinases (MMPs) are critical to the maintenance of the homeostasis of extracellular environment. Objectives: The goal of this study is to explore the role of MMP1 and MMP10 in PAH, and to investigate the molecular mechanisms involved in the regulation of MMP1 and MMP10 expression. Methods: This study first compared the expression levels of MMP1 and MMP10 in the serum between healthy donors and PAH patients. The primary cultured human monocyte-derived macrophages from PAH patients or healthy donors were subjected to the induction of MMP1 or MMP10 expression by LPS/IFNg stimulation. The expression of mRNA and protein of MMP1 or MMP10 was analyzed by semi-quantitative PCR and western blot, respectively. The human monocyte cell line, THP-1, in combined with chemical inhibitors, including PD98059, LY294002, SB203580 and SP600125, were used to investigate the signaling pathways involved in the regulation of MMP1 and MMP10 expression. For in vivo study, the lung specimens from monocrotaline (MCT) induced PAH rat were examined for the expression of MMP1 and MMP10 by semi-qPCR and immunohistochemistry, respectively.

LY294002, SB203580 and SP600125, were used to investigate the signaling pathways involved in the regulation of MMP1 and MMP10 expression. For in vivo study, the lung specimens from monocrotaline (MCT) induced PAH rat were examined for the expression of MMP1 and MMP10 by semi-qPCR and immunohistochemistry, respectively. Results: In human serum, the level of MMP1 and MMP10 is significantly higher in PAH patients. The elevated expression MMP1 and MMP10 was confirmed by in vitro cultured primary human macrophages. Under the stimulation of LPS/IFNg, macrophages derived from PAH patients showed robust upregulated MMP1 and MMP10 expression compared to healthy donors. The human THP-1 cell line, which was polarized by LPS/IFNg stimulation, provided more convincible data of the up-regulation of MMP1 and MMP10 expression in pro-inflammatory macrophages. Following the treatment of chemical inhibitors, the expression of MMP1 and MMP10 was significantly suppressed when the ERK signaling or JNK signaling was interfered, while the inhibition of p38 or PI3K signaling had no effects. For in vivo evidence, the expression level of MMP1 and MMP10 protein was critically elevated in the lung tissue of MCT-induced PAH rats compared to the PBS control group. Immunohistochemical staining further revealed the co-localization of MMP1 or MMP10 molecule with the COX-2-positive macrophages, which meant the expression of MMP1 and MMP10 was highly associated with the pro-inflammatory macrophage phenotype.

n the lung tissue of MCT-induced PAH rats compared to the PBS control group. Immunohistochemical staining further revealed the co-localization of MMP1 or MMP10 molecule with the COX-2-positive macrophages, which meant the expression of MMP1 and MMP10 was highly associated with the pro-inflammatory macrophage phenotype. Conclusions: MMP1 and MMP10 were significantly up-regulated in pro-inflammatory macrophages in vitro and in vivo in both primary cultured human macrophages, THP-1 cell line and MCT-induced PAH rat model. Blockade of JNK and ERK signaling pathways may be a novel strategy for the treatment of PAH. Grant acknowledgement None.

Conclusions: MMP1 and MMP10 were significantly up-regulated in pro-inflammatory macrophages in vitro and in vivo in both primary cultured human macrophages, THP-1 cell line and MCT-induced PAH rat model. Blockade of JNK and ERK signaling pathways may be a novel strategy for the treatment of PAH. Grant acknowledgement None. A880 An early rise in central venous pressure (CVP) predicts weaning failure during a spontaneous breathing trial: preliminary results of the 4P study S. Dubó1, A. Aquevedo2, M. Jibaja3, D. Berrutti4, C. Labra5, R. Lagos5, M.F. García3, V. Ramirez3, M. Tobar3, F. Picoita3, C. Peláez3, D. Carpio5, L. Alegría5, C. Hidalgo6, K. Godoy6, J. Bakker5, G. Hernández5 1Universidad de Concepción, Facultad de Medicina, Departamento de Kinesiología, Concepción, Chile; 2Hospital Dr. Sótero del Río, Unidad de Pacientes Críticos, Santiago, Chile; 3Hospital Eugenio Espejo, Unidad de Cuidados Intensivos, Quito, Ecuador; 4Universidad de la República, Hospital de Clínicas, Centro de Terapia Intensiva, Montevideo, Uruguay; 5Pontificia Universidad Catolica de Chile, Facultad de Medicina, Departamento de Medicina Intensiva, Santiago, Chile; 6Hospital Guillermo Grant Benavente, Concepción, Chile Correspondence: G. Hernández – Pontificia Universidad Catolica de Chile, Facultad de Medicina, Departamento de Medicina Intensiva, Santiago, Chile Background: Prediction of weaning failure through a spontaneous breathing trial (SBT) is a major task for intensivists since a failed weaning attempt exposes the patient to the risk of cardiovascular and respiratory complications, among others, that could increase morbidity or mortality. Several tests have been used to assess risk incorporating ventilatory mechanics and drive, and some cardiovascular parameters with acceptable predictive values. However, all have limitations and in general are more centered in ventilatory aspects. We hypothesized that a multimodal monitoring during a SBT addressing the capacity of the heart to manage increased venous return as reflected by CVP changes, metabolic effort as reflected by changes in central venous O2 saturation (ScvO2), and adrenergic stress as reflected by changes in capillary refill time (CRT) during a 1 or 2 h T-tube trial could predict weaning failure in critically ill patients.

the heart to manage increased venous return as reflected by CVP changes, metabolic effort as reflected by changes in central venous O2 saturation (ScvO2), and adrenergic stress as reflected by changes in capillary refill time (CRT) during a 1 or 2 h T-tube trial could predict weaning failure in critically ill patients. Objective The aim of the study was to evaluate changes in CVP, ScvO2 and CRT during a SBT and their association to weaning failure. Methods: Prospective multicenter observational study including patients in mechanical ventilation for at least 24 h for any cause, excluding only neurological or DNR patients. Usual clinical parameters and the study variables were recorded at baseline, 2, 30, and 120 min during the T-tube trial. Peripheral perfusion was evaluated with CRT and mottling score. Attending physicians were informed about changes in CVP, ScvO2 and CRT during SBT but decisions about extubation after the test were taken by them without any written guideline and based only in their usual practice. Weaning failure was defined as need of reintubation in less than 48 h after extubation. The results of a first interim analysis after completing half of the sample size calculation are presented here. Statistical analysis included t-test.

ken by them without any written guideline and based only in their usual practice. Weaning failure was defined as need of reintubation in less than 48 h after extubation. The results of a first interim analysis after completing half of the sample size calculation are presented here. Statistical analysis included t-test. Results: A total of 114 patients from 5 centers were enrolled (age: 53 ± 22 y, APACHE II: 19.7 ± 9.6), and 93 pts were extubated after the SBT. Weaning failure occurred occurred in 7 (8 %). Usual clinical parameters such as HR, RR, SBP, and FiO2 were not different between successful vs failed weaning procedures. Reintubated patients presented an early (2 min) significant increase in CVP (9 to 11.8 mmHg; p 0.007) and a late (120 min) decrease in ScvO2 (60 % vs 43 %; p 0.05), as compared to succesfully weaned pts. Conclusions: An early rise in central venous pressure appears to predict weaning failure during a spontaneous breathing trial.

Results: A total of 114 patients from 5 centers were enrolled (age: 53 ± 22 y, APACHE II: 19.7 ± 9.6), and 93 pts were extubated after the SBT. Weaning failure occurred occurred in 7 (8 %). Usual clinical parameters such as HR, RR, SBP, and FiO2 were not different between successful vs failed weaning procedures. Reintubated patients presented an early (2 min) significant increase in CVP (9 to 11.8 mmHg; p 0.007) and a late (120 min) decrease in ScvO2 (60 % vs 43 %; p 0.05), as compared to succesfully weaned pts. Conclusions: An early rise in central venous pressure appears to predict weaning failure during a spontaneous breathing trial. A881 Disseminated intravascular coagulation during early phase of out-of-hospital cardiac arrest and resuscitation belongs to the fibrinolytic phenotype Y. Sadamoto, K. Katabami, T. Wada, Y. Ono, K. Maekawa, M. Hayakawa, A. Sawamura, S. Gando Hokkaido University Graduate School of Medicine, Emergency and Critical Care Medicine, Sapporo, Japan Correspondence: Y. Sadamoto – Hokkaido University Graduate School of Medicine, Emergency and Critical Care Medicine, Sapporo, Japan Background: Hypoxia and ischemia influence blood coagulation and fibrinolysis, if sufficiently severe like cardiac arrest and resuscitation, cause disseminated intravascular coagulation (DIC). We investigated characteristics of DIC diagnosed during the first 24 hours after cardiac arrest and return of spontaneous circulation.

und: Hypoxia and ischemia influence blood coagulation and fibrinolysis, if sufficiently severe like cardiac arrest and resuscitation, cause disseminated intravascular coagulation (DIC). We investigated characteristics of DIC diagnosed during the first 24 hours after cardiac arrest and return of spontaneous circulation. Methods: Patients with established out-of-hospital cardiac arrest (OHCA) who underwent cardiopulmonary resuscitation with subsequent return of spontaneous circulation were retrospectively enrolled. Patients were divided two subgroups with DIC (208) and without DIC (180) by the Japanese Association for Acute Medicine (JAAM) DIC criteria. Platelet count, global markers of coagulation and fibrinolysis were measured 4 times after admission to emergency department (T1, 0–6; T2, 6–12; T3, 12–18; T4, 18–24 hr). A FDP/D-dimer ratio was used as a surrogate marker of fibrin(ogen)olysis. Results: DIC patients showed significantly lower platelet counts, fibrinogen and antithrombin levels, and more prolonged prothrombin time throughout the study period. FDP and D-dimer levels as well as FDP/D-dimer ratio of DIC patients were extremely higher than those without DIC. Mean values of FDP/D-dimer ratio of DIC patients exceed 2.0, which suggests fibrin(ogen)olysis. Higher sequential organ failure assessment score (SOFA) (9 vs. 6, p < 0.001) associated with increased prevalence of multiple organ dysfunction (23.6 % vs. 3.9 %, p < 0.001) in DIC patients were observed, which leads to higher hospital mortality of DIC patients (54.8 % vs. 23.9 %, p < 0.001).

ggests fibrin(ogen)olysis. Higher sequential organ failure assessment score (SOFA) (9 vs. 6, p < 0.001) associated with increased prevalence of multiple organ dysfunction (23.6 % vs. 3.9 %, p < 0.001) in DIC patients were observed, which leads to higher hospital mortality of DIC patients (54.8 % vs. 23.9 %, p < 0.001). Conclusion: DIC immediately after cardiac arrest and resuscitation shows increased fibrin(ogen)olysis, which suggest that the DIC belongs to fibrinolytic phenotype. This type of DIC may affects OHCA patient's outcome. A882 Experience in circulatory and respiratory support with extracorporeal membrane oxigenation (ECMO) in complicated postoperative course after pulmonary endarterectomy H. Marin-Mateos, J.L. Perez-Vela, R. Garcia-Gigorro, M.A. Corres Peiretti, M.J. Lopez-Gude, S. Chacon-Alves, E. Renes-Carreño, J.C. Montejo-González Hospital 12 de Octubre, Medicina Intensiva, Madrid, Spain Correspondence: H. Marin-Mateos – Hospital 12 de Octubre, Medicina Intensiva, Madrid, Spain Introduction: Pulmonary endarterectomy is an effective solution for chronic thromboembolic pulmonary hypertension. However outcome at the ICU is not always easy due to important changes in respiratory and cardiovascular systems after surgery. Objectives: To describe our experience in circulatory and respiratory assistance with ECMO in complicated postoperative course after bilateral thromboendarterectomy in patients with pulmonary hipertension cause by Thromboembolic Chronic Disease.

A882 Experience in circulatory and respiratory support with extracorporeal membrane oxigenation (ECMO) in complicated postoperative course after pulmonary endarterectomy H. Marin-Mateos, J.L. Perez-Vela, R. Garcia-Gigorro, M.A. Corres Peiretti, M.J. Lopez-Gude, S. Chacon-Alves, E. Renes-Carreño, J.C. Montejo-González Hospital 12 de Octubre, Medicina Intensiva, Madrid, Spain Correspondence: H. Marin-Mateos – Hospital 12 de Octubre, Medicina Intensiva, Madrid, Spain Introduction: Pulmonary endarterectomy is an effective solution for chronic thromboembolic pulmonary hypertension. However outcome at the ICU is not always easy due to important changes in respiratory and cardiovascular systems after surgery. Objectives: To describe our experience in circulatory and respiratory assistance with ECMO in complicated postoperative course after bilateral thromboendarterectomy in patients with pulmonary hipertension cause by Thromboembolic Chronic Disease. Methods: An observational research about a prospective data basis is made. All patients with this type of assistance requirements, admitted in the ICU for five years, from January 2010 to December 2015 are included. Variables as demographic and hemodynamics data, resource consumption in days of ICU stay, hours of mechanical ventilation, hours of ECMO therapy, complications, outcomes and mortality are analysed.

this type of assistance requirements, admitted in the ICU for five years, from January 2010 to December 2015 are included. Variables as demographic and hemodynamics data, resource consumption in days of ICU stay, hours of mechanical ventilation, hours of ECMO therapy, complications, outcomes and mortality are analysed. Results: In total, eight patients requiered ECMO support in the postoperative course after endarterctomy. The median age was 54 years (ICR 39–67), 62,5 % women. Peripheral cannulation in 100 % cases. The type of asisstance was veno-venous in 3 cases and veno-arterial in 5 cases. Indication for assistance was respiratory falilure alone in 37 % of patients, and combination of respiratory and cardiogenic shock (mainly caused by right ventricle failure) in 63 %. Underlying patology for respiratory and circulatory failure was reperfusión pulmonary edema alone 25 %, combination of reperfusión edema and right ventricle failure 50 %, anda diffuse alveolar haemorrhage in 25 % cases. A median of 204 hours (ICR 151–334) of therapy with ECMO is observed. Three patients presented bleeding complication (two minor pericannula bleeding and one major complication with intracraneal bleeding). Five patients developed acute renal failure, four of them with renal replacement therapy requirement. 62 % patients required norepinefrine and dobutamine infusion before cannulation. Hours of mechanical ventilation median 624 (ICR 451–1110), days of ICU stay median 26.5 (ICR 12–51) and in-hospital stay median 31 days (ICR 13–82). Mortality in ICU was 50 % of cases. Two-years survival was 100 % in patients where succesful weaning from ECMO therapy was performed.

butamine infusion before cannulation. Hours of mechanical ventilation median 624 (ICR 451–1110), days of ICU stay median 26.5 (ICR 12–51) and in-hospital stay median 31 days (ICR 13–82). Mortality in ICU was 50 % of cases. Two-years survival was 100 % in patients where succesful weaning from ECMO therapy was performed. Conclusions: Sometimes, pulmonary endarterectomy may present a complicated postoperative course with severe respiratory insufficiency and/or right ventricle failure (cardiogenic shock), that implies high morbidity and mortality, and high resource consumption. In these patients ECMO assistance is a useful support for ICU management and patients survival.

ndarterectomy may present a complicated postoperative course with severe respiratory insufficiency and/or right ventricle failure (cardiogenic shock), that implies high morbidity and mortality, and high resource consumption. In these patients ECMO assistance is a useful support for ICU management and patients survival. A883 Lung ultrasound detects pulmonary complications earlier than chest x-ray in patients after cardiac surgery: a prospective single centre observational study K.L. Parlevliet1, H.R.W. Touw2, M. Beerepoot1, C. Boer2, P.W.G. Elbers1, P.R. Tuinman1 1VU University Medical Center, Intensive Care Medicine, Amsterdam, Netherlands; 2VU University Medical Center Amsterdam, Anaesthesiology, Amsterdam, Netherlands Correspondence: K.L. Parlevliet – VU University Medical Center, Intensive Care Medicine, Amsterdam, Netherlands Introduction: Postoperative pulmonary complications (PPCs) are common after cardiac surgery and especially clinically relevant PPCs (cr-PPCs) are associated with adverse outcomes. Early detection is important as it may lead to improved outcome (1). The accuracy of chest X-rays (CXRs) in the detection of PPCs is limited. Lung ultrasound (LUS) is emerging as an accurate diagnostic tool for pulmonary pathology within the intensive care unit (ICU) (2, 3). In literature little is known about the value of LUS for the detection of PPCs in patients after cardiac surgery. Objectives: The primary aim of this study was to assess whether routine LUS can detect cr-PPCs earlier than routine CXR in patients after cardiac surgery.

A883 Lung ultrasound detects pulmonary complications earlier than chest x-ray in patients after cardiac surgery: a prospective single centre observational study K.L. Parlevliet1, H.R.W. Touw2, M. Beerepoot1, C. Boer2, P.W.G. Elbers1, P.R. Tuinman1 1VU University Medical Center, Intensive Care Medicine, Amsterdam, Netherlands; 2VU University Medical Center Amsterdam, Anaesthesiology, Amsterdam, Netherlands Correspondence: K.L. Parlevliet – VU University Medical Center, Intensive Care Medicine, Amsterdam, Netherlands Introduction: Postoperative pulmonary complications (PPCs) are common after cardiac surgery and especially clinically relevant PPCs (cr-PPCs) are associated with adverse outcomes. Early detection is important as it may lead to improved outcome (1). The accuracy of chest X-rays (CXRs) in the detection of PPCs is limited. Lung ultrasound (LUS) is emerging as an accurate diagnostic tool for pulmonary pathology within the intensive care unit (ICU) (2, 3). In literature little is known about the value of LUS for the detection of PPCs in patients after cardiac surgery. Objectives: The primary aim of this study was to assess whether routine LUS can detect cr-PPCs earlier than routine CXR in patients after cardiac surgery. Methods: This is a prospective observational single centre study performed in a tertiary ICU. 40 cardiac surgical patients were included. LUS was performed on the same days as routine CXR; at ICU admission (day 0), day 2 and day 3 post-operative. CXR and LUS findings on day 0, 2 and 3 were compared to cr-PPCs. Cr-PPCs were defined as PPCs requiring an intervention by the attending ICU-physician and considered the composite reference standard. Physicians were blinded for LUS findings. To analyse earlier detection of cr-PPCs the Wilcoxon Signed-Rank test was used. Detection of a cr-PPC was ranked per day and rankings between methods were compared per patient.

iring an intervention by the attending ICU-physician and considered the composite reference standard. Physicians were blinded for LUS findings. To analyse earlier detection of cr-PPCs the Wilcoxon Signed-Rank test was used. Detection of a cr-PPC was ranked per day and rankings between methods were compared per patient. Results: Overall PPC (≥1) incidence was 37 (92.5 %) for CXR and 40 (100 %) for LUS. Fifteen cr-PPCs were identified in 14 patients (35 %). Patients with a cr-PPC had a prolonged ICU stay (82 ± 45 vs. 53 ± 24 hours; p = 0.004) and a lower PaO2/FiO2 ratio (199 ± 56 vs. 277 ± 97 mmHg; p = 0.010) compared to patients without cr-PPCs. CXR detected 5 (33.3 %) and LUS detected 11 (78.6 %) cr-PPCs. These cr-PPCs were detected earlier with LUS than CXR as presented in Table 19 (p = 0.024). Conclusions: LUS detects the more PPCs and cr-PPCs than CXR and in an earlier stage. Further research needs to determine which specific LUS findings are associated with the development of clinically relevant pulmonary complications. References 1. Canet J, Gallart L, Gomar C, Paluzie G, Vallès J, Castillo J, et al. Prediction of postoperative pulmonary complications in a population-based surgical cohort. Anesthesiology. 2010 Dec;113(6):1338–50. 2. Touw HRW, Tuinman PR, Gelissen HPMM, Lust E, Elbers PWG. Lung ultrasound: routine practice for the next generation of internists. Neth J Med. 2015;73(3):100–7. 3. Lichtenstein DA. Relevance of Lung Ultrasound in the Diagnosis of Acute Respiratory Failure*. CHEST J. 2008 Jul 1;134(1):117. Grant acknowledgement

1. Canet J, Gallart L, Gomar C, Paluzie G, Vallès J, Castillo J, et al. Prediction of postoperative pulmonary complications in a population-based surgical cohort. Anesthesiology. 2010 Dec;113(6):1338–50. 2. Touw HRW, Tuinman PR, Gelissen HPMM, Lust E, Elbers PWG. Lung ultrasound: routine practice for the next generation of internists. Neth J Med. 2015;73(3):100–7. 3. Lichtenstein DA. Relevance of Lung Ultrasound in the Diagnosis of Acute Respiratory Failure*. CHEST J. 2008 Jul 1;134(1):117. Grant acknowledgement This study is not supported by any grant.Table 19 (abstract A883). Cumulative number of detected cr-PPCs by CXR, LUS and the composite reference standard (cr-PPC) per day after cardiac surgery Day 0 Day 2 Day 3 CXR LUS cr-PPC CXR LUS cr-PPC CXR LUS cr-PPC Atelectasis (n = 2) 1 2 2 1 2 2 1 2 2 Pulmonary edema (n = 9) 1 6 0 2 7 4 3 7 9 Other (n = 4) 1 2 1 1 2 4 1 2 4 Total (n = 15) 3 10 3 4 11 10 5 11 15 A884 Predictors of severe hypotension in ICU patients sedated with propofol S.A. Abdelmonem1, T.A. Helmy1, I. El Sayed2, S. Ghazal1 1University of Alexandria, Critical Care, Alexandria, Egypt; 2Medical Research Institute, Biomedical Informatics & Medical Statistics, Alexandria, Egypt Correspondence: S.A. Abdelmonem – University of Alexandria, Critical Care, Alexandria, Egypt Introduction: Propofol is widely used in critical care sedation due to its pharmacological properties which allow serial neurological examination(1).Hypo tension is a common side effect of propofol infusion, which affect patient outcome.

ence: S.A. Abdelmonem – University of Alexandria, Critical Care, Alexandria, Egypt Introduction: Propofol is widely used in critical care sedation due to its pharmacological properties which allow serial neurological examination(1).Hypo tension is a common side effect of propofol infusion, which affect patient outcome. Objectives: Aimed to identify predictors of severe hypotension in patients sedated with propofol Methods: This prospective study included 105 patientsthat were randomly selected by simple random sampling from population of patients admitted to the critical care medicine department in Alexandria Main University Hospital (49 normotensiveand 56 hypotensive).Bivariate analyses were performed to identify risk factors of severe hypotension during propofol infusion.Multivariate logistic regression analysis was performed to identify variables independently associated with hypotension after adjustment of other factors

Main University Hospital (49 normotensiveand 56 hypotensive).Bivariate analyses were performed to identify risk factors of severe hypotension during propofol infusion.Multivariate logistic regression analysis was performed to identify variables independently associated with hypotension after adjustment of other factors Results: Multivariable logistic regression identified age, baseline MAP, Serum bicarbonate and serum creatinine as factors independently associated with and account for 86 % of the hypotension variance.The odds of hypotension increases by 1.17(95 % CI: 1.06 to 1.29) per one year increase in age. One unit increase in HCO3 results in decrease chance to develop hypotension by 30 % (OR:.705; 95 % CI:.523 to .951). One unit increase in MAP at beginning of treatment decrease the odds of hypotension by 7.6 % (OR:.924; 95 % CI:.869 to .982). Both Age and serum biacarbonate have fair diagnostic accuracy to differentiate cases with and without hypertension (AUC (SE): .746(.050), .785 (.047)) respectively. Baseline MAP < 98 mmHg has good diagnostic accuracy to differentiate cases with hypotension (AUC (SE): .837(.040)). Conclusions: Age and serum bicarbonate can predict hypotension during propofol infusion. We should use propofol with caution in elderly and patients with baseline MAP < 98 mmHg. References 1- Shah NK, Harris M, Govindugari K, Rangaswamy HB, Jeon H. Effect of propofol titration v/s bolus during induction of anesthesia on hemodynamics and bispectral index. Middle East J Anesthesiol. 2011;21(2):275–81. 2. Grant acknowledgement

Conclusions: Age and serum bicarbonate can predict hypotension during propofol infusion. We should use propofol with caution in elderly and patients with baseline MAP < 98 mmHg. References 1- Shah NK, Harris M, Govindugari K, Rangaswamy HB, Jeon H. Effect of propofol titration v/s bolus during induction of anesthesia on hemodynamics and bispectral index. Middle East J Anesthesiol. 2011;21(2):275–81. 2. Grant acknowledgement Thanks to all staff of the Department of Critical Care Medicine, University of Alexandria A885 Interlukins and post coronary artery bypass delirium in intensive care unit, an experimental clinical study S.H. Akhlagh1, M. Masjedi1,2, K. Hozhabri1, E. Kamali3 1Shiraz University of Medical Sciences, Anesthesia and Intensive Care Department, Shiraz, Islamic Republic of Iran; 2Anesthesiology and Critical Care Research Center, Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran; 3Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran Correspondence: M. Masjedi – Shiraz University of Medical Sciences, Anesthesia and Intensive Care Department, Shiraz, Islamic Republic of Iran Introduction: Delirium is a common Complication in elderly patients after Cardiac surgery. The pathophysiology of delirium remains poorly understood. Several conditions associated with delirium are characterized by activation of inflammatory Cascade. Objectives: The purpose of this study was to find out the correlation between serum interleukins 6 and 10 with delirium in intensive care unit following coronary artery bypass surgery.

A885 Interlukins and post coronary artery bypass delirium in intensive care unit, an experimental clinical study S.H. Akhlagh1, M. Masjedi1,2, K. Hozhabri1, E. Kamali3 1Shiraz University of Medical Sciences, Anesthesia and Intensive Care Department, Shiraz, Islamic Republic of Iran; 2Anesthesiology and Critical Care Research Center, Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran; 3Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran Correspondence: M. Masjedi – Shiraz University of Medical Sciences, Anesthesia and Intensive Care Department, Shiraz, Islamic Republic of Iran Introduction: Delirium is a common Complication in elderly patients after Cardiac surgery. The pathophysiology of delirium remains poorly understood. Several conditions associated with delirium are characterized by activation of inflammatory Cascade. Objectives: The purpose of this study was to find out the correlation between serum interleukins 6 and 10 with delirium in intensive care unit following coronary artery bypass surgery. Methods: In an exploratory observational study we included 20 patients with delirium that has undergone coronary artery bypass surgery (CABG). Serum samples were taken, postoperatively, before, during and after development of delirium in cardiac intensive care unit . Delirium was diagnosed by using the Confusion- Assessment Method- ICU (CAM- ICU). IL6 and IL10 were determined by Enzyme Linked Immunosorbent Assay.

dergone coronary artery bypass surgery (CABG). Serum samples were taken, postoperatively, before, during and after development of delirium in cardiac intensive care unit . Delirium was diagnosed by using the Confusion- Assessment Method- ICU (CAM- ICU). IL6 and IL10 were determined by Enzyme Linked Immunosorbent Assay. Results: Delirium was detected in 20 of 380 patients (5.2 %) which were admitted in cardiac ICU during a 17 month period. In delirious patients Plasma IL6 Level were higher in delirious stage than post delirium period (p = 0/002). Patients with the hyperactive Subtype of delirium had higher IL6 Levels in delirious stage than post delirium period (p = 0/009). There was no difference in Serum IL10 levels in delirious and post delirious period. Conclusions: Interleukin 6 may contribute to the pathogenesis of post coronary artery bypass delirium. Its role is more prominent in the hyperactive behavior of delirium. No correlation was found between serum level of interleukin10 and delirium. Further studies are needed to elucidate the relationship of other cytokines with the pathogenesis of delirium. References 1. De Rooij SE, Van Munster BC, Korevaar JC, Levi M. Cytokines and acute phase response in delirium. J Psychosom Res. 2007 May; 62 (5): 521–5. 2. Boogaard M, Kox M, Quinn K, Achterberg T, Hoeven J, Schoonhoven L, et al. Biomarkers associated with delirium in critically ill patients and their relation with long- term subjective cognitive dysfunction; indications for different pathways governing delirium in inflamed and noninflamed patients. Crit Care. 2011; 15 (6): R 297.

M, Kox M, Quinn K, Achterberg T, Hoeven J, Schoonhoven L, et al. Biomarkers associated with delirium in critically ill patients and their relation with long- term subjective cognitive dysfunction; indications for different pathways governing delirium in inflamed and noninflamed patients. Crit Care. 2011; 15 (6): R 297. 3. Miyazaki S, Yoshitani K, Miura N, Irie T, Inatomi Y, Ohnishi Y, et al. Risk factors of stroke and delirium after off- pump Coronary artery bypass surgery. Interact Cardiovasc Thorac Surg. 2011 Mar; 12 (3): 379–83. Grant acknowledgement This study was financially supported by the vice chancellery for research affairs of Shiraz university of medical sciences, Shiraz, Iran. Many thanks to all nurses of cardiac ICU of Nemazee hospital , Shiraz , Iran.

3. Miyazaki S, Yoshitani K, Miura N, Irie T, Inatomi Y, Ohnishi Y, et al. Risk factors of stroke and delirium after off- pump Coronary artery bypass surgery. Interact Cardiovasc Thorac Surg. 2011 Mar; 12 (3): 379–83. Grant acknowledgement This study was financially supported by the vice chancellery for research affairs of Shiraz university of medical sciences, Shiraz, Iran. Many thanks to all nurses of cardiac ICU of Nemazee hospital , Shiraz , Iran. A886 Assessment of hemostasis in patients undergoing therapeutic plasma exchange by rotation thrombelastometry I. Zýková, B. Paldusová, P. Sedlák, D. Morman Regional Hospital Liberec, Department of Anesthesia and intensive Care, Liberec, Czech Republic Correspondence: I. Zýková – Regional Hospital Liberec, Department of Anesthesia and intensive Care, Liberec, Czech Republic Introduction: Therapeutic plasma exchange (TPE) is a therapeutic option in a variety of diseases. In most indications albumin is recommended as a replacement fluid. During TPE coagulation factors are removed and a depletion of coagulation factors results. We have found no recommendations for the assessment of hemostasis and management of a bleeding risk in patients undergoing TPE protocol. The increased bleeding risk is important in the management of these patients. The assesment of hemostasis by rotational thrombelastometry (ROTEM) is a fast bedside method, which can be used in assessment of these patients.The only published data about assessment of hemostasis by rotational thrombelastometry in patients after TPE are in patients using TPE before cadaveric donor kidney transplant (1). We found no data of ROTEM use in patients, where TPE protocols with series of TPE were used.

ch can be used in assessment of these patients.The only published data about assessment of hemostasis by rotational thrombelastometry in patients after TPE are in patients using TPE before cadaveric donor kidney transplant (1). We found no data of ROTEM use in patients, where TPE protocols with series of TPE were used. Objectives: To assess the usage of rotational thrombelastometry in patients undergoing series of TPE. Methods: We have performed rotational thrombelastometry (ROTEM) in 4 patients with Guillain Barré syndrome undergoing TPE therapy. All patients had a series of TPE, the interval between TPE was 24–48 hours. Albumin was used as a replacement fluid. We performed standard coagulation tests and rotational thrombelastometry before and after each TPE. Results: TPE reduced fibrinogen levels both in a standard laboratory test and according to FIBTEM (MCF). Clotting time (CT) and clot formation time (CFT) were prolonged both in INTEM and EXTEM. Standard coagulation tests were also prolonged (aPTT, INR). The trombelastometry results reflected a serious coagulopathy which corrected spontaneously between each TPE. We used thrombelastometry for timing of invasive procedures. We did not correct coagulopathy in patients without clinical bleeding. We corrected coagulopathy only in patients with bleeding or undergoing invasive procedures. Thrombelastometry was a fast and easy bedside test for the assessment of bleeding risk in these patients.Fig. 21 (abstract A886). ROTEM before TPE Fig. 22 (abstract A886). ROTEM after TPE

Results: TPE reduced fibrinogen levels both in a standard laboratory test and according to FIBTEM (MCF). Clotting time (CT) and clot formation time (CFT) were prolonged both in INTEM and EXTEM. Standard coagulation tests were also prolonged (aPTT, INR). The trombelastometry results reflected a serious coagulopathy which corrected spontaneously between each TPE. We used thrombelastometry for timing of invasive procedures. We did not correct coagulopathy in patients without clinical bleeding. We corrected coagulopathy only in patients with bleeding or undergoing invasive procedures. Thrombelastometry was a fast and easy bedside test for the assessment of bleeding risk in these patients.Fig. 21 (abstract A886). ROTEM before TPE Fig. 22 (abstract A886). ROTEM after TPE Conclusions: Hemostasis was severely infuenced by TPE, the resulting coagulopathy corrected spontaneously before next TPE. Assesment of hemostasis should always be done in patients on TPE using albumin as a replacement fluid. The bleeding risk should be monitored and measured. The timing of invasive procedures such as central vein cannulation and tracheostomy should be guided by coagulation studies. Rotational thrombelastometry can be used to asses hemostasis and bleeding risk. In patients with Guillain-Barre syndrome spinal tap is performed. Caution should be given to the timing of spinal tap and TPE. References 1. Tholking G, Mesters R, Dittrich R et al. Assessment of Hemostasis after Plasma Exchange Using Rotational Thrombelastometry (ROTEM).PLoS ONE 10(6):e0130402.doi:10.1371

Conclusions: Hemostasis was severely infuenced by TPE, the resulting coagulopathy corrected spontaneously before next TPE. Assesment of hemostasis should always be done in patients on TPE using albumin as a replacement fluid. The bleeding risk should be monitored and measured. The timing of invasive procedures such as central vein cannulation and tracheostomy should be guided by coagulation studies. Rotational thrombelastometry can be used to asses hemostasis and bleeding risk. In patients with Guillain-Barre syndrome spinal tap is performed. Caution should be given to the timing of spinal tap and TPE. References 1. Tholking G, Mesters R, Dittrich R et al. Assessment of Hemostasis after Plasma Exchange Using Rotational Thrombelastometry (ROTEM).PLoS ONE 10(6):e0130402.doi:10.1371 A887 Heated carrier fluids in decreasing propofol injection pain: a randomized controlled trial A.M. Youn Chungnam National University Hospital, Anesthesiology and Pain Medicine, Daejon, Republic of Korea Introduction: Propofol is a common intravenous drug used during anesthetic induction and sedation because of its rapid onset and short duration. Its downfall, however, is that patients experience injection pain so severe that they recall induction as the most painful part of the sedation process. Among numerous reports in efforts to decrease propofol injection pain, the most effective combination of drug and non-drug intervention evaluated through a quantitative systematic review revealed to be pretreatment with 0.5 mg/kg lidocaine in combination with a tourniquet for venous occlusion. The majority of these reports conclude that a single method is insufficient in eliminating propofol injection pain.

combination of drug and non-drug intervention evaluated through a quantitative systematic review revealed to be pretreatment with 0.5 mg/kg lidocaine in combination with a tourniquet for venous occlusion. The majority of these reports conclude that a single method is insufficient in eliminating propofol injection pain. Objectives: We evaluated the effect of heated carrier fluids (40 °C) in decreasing propofol injection pain. Methods: A randomized controlled clinical trial was conducted in 90 patients (ASA 1 or 2), ages 18 to 65. Patients were allocated into 3 groups (n = 30) each. Group W received 200 ml of heated carrier fluids for 20 minutes prior to 2 mg/kg propofol injection; group L received 200 ml of heated carrier fluids for 20 minutes prior to lidocaine pretreatment and 2 mg/kg propofol injection: and group C (control group) received 200 ml of room temperature fluids prior to 2 mg/kg propofol injection. Propofol injection pain was evaluated using the verbal pain score (VPS). Results: Group W and L showed significant reduction in the incidence and severity of injection pain compared to group C (P < 0.05). VPS was significantly lower in group W (p = 0.018) and L (p = 0.036) compared to group C. There was no statistical difference between group W and group L (p = 0.3). There was statistically significant difference in mean blood pressures measured after 2 mg/kg propofol injection among groups. Conclusions: Both heated carrier fluids and combination of lidocaine pretreatment effectively reduced propofol injection pain. References

Results: Group W and L showed significant reduction in the incidence and severity of injection pain compared to group C (P < 0.05). VPS was significantly lower in group W (p = 0.018) and L (p = 0.036) compared to group C. There was no statistical difference between group W and group L (p = 0.3). There was statistically significant difference in mean blood pressures measured after 2 mg/kg propofol injection among groups. Conclusions: Both heated carrier fluids and combination of lidocaine pretreatment effectively reduced propofol injection pain. References 1. Jalota L, Kalira V, George E, Shi YY, Hornuss C, Radke O, et al. Prevention of pain on injection of propofol: systematic review and meta-analysis. BMJ 2011; 342: d1110. 2. Picard P, Tramer MR. Prevention of pain on injection with propofol: a quantitative systematic review. Anesth Analg 2000; 90: 963–9.

Conclusions: Both heated carrier fluids and combination of lidocaine pretreatment effectively reduced propofol injection pain. References 1. Jalota L, Kalira V, George E, Shi YY, Hornuss C, Radke O, et al. Prevention of pain on injection of propofol: systematic review and meta-analysis. BMJ 2011; 342: d1110. 2. Picard P, Tramer MR. Prevention of pain on injection with propofol: a quantitative systematic review. Anesth Analg 2000; 90: 963–9. A888 In-hospital adverse events among surgical patients in japan: the JET study Y. Ohta1, M. Sakuma2, D. Bates3, T. Morimoto2 1Hyogo College of Medicine, General Internal Medicine, Nishinomiya, Japan; 2Hyogo College of Medicine, Clinical Epidemiology, Nishinomiya, Japan; 3Brigham and Women's Hospital, General Internal Medicine, Boston, United States Correspondence: Y. Ohta – Hyogo College of Medicine, General Internal Medicine, Nishinomiya, Japan Introduction: The epidemiology and the nature of adverse events (AEs) were mainly reported in Western countries and were limited in the outside Western countries. However, the local epidemiological data is important to extrapolate the effects of established preventive strategy of AEs in Western countries into other countries. Objectives: We conducted a prospective cohort study to clarify the epidemiology and the nature of AEs in surgical inpatients in Japan.

A888 In-hospital adverse events among surgical patients in japan: the JET study Y. Ohta1, M. Sakuma2, D. Bates3, T. Morimoto2 1Hyogo College of Medicine, General Internal Medicine, Nishinomiya, Japan; 2Hyogo College of Medicine, Clinical Epidemiology, Nishinomiya, Japan; 3Brigham and Women's Hospital, General Internal Medicine, Boston, United States Correspondence: Y. Ohta – Hyogo College of Medicine, General Internal Medicine, Nishinomiya, Japan Introduction: The epidemiology and the nature of adverse events (AEs) were mainly reported in Western countries and were limited in the outside Western countries. However, the local epidemiological data is important to extrapolate the effects of established preventive strategy of AEs in Western countries into other countries. Objectives: We conducted a prospective cohort study to clarify the epidemiology and the nature of AEs in surgical inpatients in Japan. Methods: The Japan adverse EvenT (JET) study was a prospective cohort study which had evaluated AEs and medical errors (MEs) at 2 tertiary care hospitals. The 38 medical and surgical wards were stratified according to hospital and whether they were medical or surgical wards, and study wards were randomly selected. Intensive care units (ICUs) were all included. We included all adult patients aged > = 15 years old who were admitted to any of the selected 23 study wards (10 medical, 11 surgical, and 2 ICUs) over a 2-month period. The primary outcome of this study was the epidemiology and the nature of AEs and MEs in the patients who had operation during the study period. Trained nurses placed at each participating hospital reviewed all charts daily on weekends, along with laboratories, incident reports, and prescription queries to collect any potential event. They also collected the characteristics of the patients in the cohort. Independent physician reviewers evaluated all potential events and classified to whether they were AEs or MEs, as well as to their classification, severity and preventability.

ncident reports, and prescription queries to collect any potential event. They also collected the characteristics of the patients in the cohort. Independent physician reviewers evaluated all potential events and classified to whether they were AEs or MEs, as well as to their classification, severity and preventability. Results: The JET study reported that 1,130 patients enrolled with 19,180 patient-days. We included 392 patients with 7046 patient-days. Among 392 patients, 227 patients (58 %) were male and the median of age at admission was 68 years old {Interquartile range (IQR): 55, 78}. Nine patients (2 %) died during hospital stay and the median hospital stay was 9 days (IQR: 5, 20). We found 489 AEs in 202 patients, for the incidence of 69.4 per 1000 patient-days and 124.7 per 100 admissions, and 208 MEs in 108 patients, for the incidence of 29.5 per 1000 patient-days and 53 per 100 admissions. We classified AEs into following type with duplicate count: surgical operations consisted of 35 % events, followed by drug 31 %, procedure 15 %, nursing 10 %, and others 8 %. In terms of severity of AEs, 1 % were fatal; 5 % were life-threatening; 31 % were serious; and 63 % were significant events. Among the 489 AEs, 119 (24 %) were associated with error and preventable. Conclusions: We found a significant high incidence of AEs and MEs among surgical inpatient in Japan. We also showed operation related AEs were the most frequent AEs.

We classified AEs into following type with duplicate count: surgical operations consisted of 35 % events, followed by drug 31 %, procedure 15 %, nursing 10 %, and others 8 %. In terms of severity of AEs, 1 % were fatal; 5 % were life-threatening; 31 % were serious; and 63 % were significant events. Among the 489 AEs, 119 (24 %) were associated with error and preventable. Conclusions: We found a significant high incidence of AEs and MEs among surgical inpatient in Japan. We also showed operation related AEs were the most frequent AEs. Oral Sessions. PEEP AND ALVEOLAR RECRUITMENT A889 Positive end-expiratory pressure selection in patients with acute respiratory distress syndrome: alveolar recruitment versus electrical impedance tomography based lung collapsibility P.-L. Su1, W.-Y. Chang1, W.-C. Lin1, C.-W. Chen1,2 1National Cheng Kung University Hospital, Department of Internal Medicine, Tainan, Taiwan, Province of China; 2National Cheng Kung University, Medical Device Innovation Center, Tainan, Taiwan, Province of China Correspondence: P.-L. Su – National Cheng Kung University Hospital, Department of Internal Medicine, Tainan, Taiwan, Province of China Introduction: Appropriate PEEP selection is crucial in the ventilator management of ARDS patients. Yet information of alveolar recruitment or degree of lung collapse in the selection of PEEP is unknown. Objectives: Selection of PEEP based on best dorsal(dependent) compliance through electrical impedance monitoring during decremental PEEP titration1 and its relationship to alveolar recruitment and lung collapsibility.

Oral Sessions. PEEP AND ALVEOLAR RECRUITMENT A889 Positive end-expiratory pressure selection in patients with acute respiratory distress syndrome: alveolar recruitment versus electrical impedance tomography based lung collapsibility P.-L. Su1, W.-Y. Chang1, W.-C. Lin1, C.-W. Chen1,2 1National Cheng Kung University Hospital, Department of Internal Medicine, Tainan, Taiwan, Province of China; 2National Cheng Kung University, Medical Device Innovation Center, Tainan, Taiwan, Province of China Correspondence: P.-L. Su – National Cheng Kung University Hospital, Department of Internal Medicine, Tainan, Taiwan, Province of China Introduction: Appropriate PEEP selection is crucial in the ventilator management of ARDS patients. Yet information of alveolar recruitment or degree of lung collapse in the selection of PEEP is unknown. Objectives: Selection of PEEP based on best dorsal(dependent) compliance through electrical impedance monitoring during decremental PEEP titration1 and its relationship to alveolar recruitment and lung collapsibility. Methods: Twenty-three patients with ARDS received decremental PEEP titration at 5 levels of PEEP (25, 20, 15, 11, 7 cmH2O) following a short cycle of recruitment maneuver. At PEEP level of 15, 11, 7 cmH2O, end-expiratory lung volume(EELV) was measured with nitrogen washout/washin technique and estimation of PEEP-induced alveolar recruitment was determined after subtraction of the minimum predicted increase in lung volume (△PEEP X Compliance rs at PEEP = 7 cmH2O). Estimation of lung collapsibility was obtained via regional lung compliance changes obtained from EIT at five PEEP levels according to the method of Costa et al2. The PEEP level with the best dorsal compliance is defined as the selected PEEP.

predicted increase in lung volume (△PEEP X Compliance rs at PEEP = 7 cmH2O). Estimation of lung collapsibility was obtained via regional lung compliance changes obtained from EIT at five PEEP levels according to the method of Costa et al2. The PEEP level with the best dorsal compliance is defined as the selected PEEP. Results: Data are presented as median [interquartile range] or mean ± SD. PEEP induced alveolar recruitment was 183.8 ml [22.1 - 340.3](PEEP level of 15 cmH2O to 7 cmH2O) and alveolar recruitment equal to or greater than the median (183.8 ml) as high recruiter (HR), otherwise it is low recruiter(LR). Degree of lung collapsibility between HR and LR at the PEEP level of 7 cmH2O was not significantly differed (25.2 ± 11.4 % in HR vs 21.4 ± 12.7 % in LR) and the final PEEP selected was not differed too. However, for patients with final selected PEEP level of 20 cmH2O/15 cmH2O/≤ 11cmH2O, the corresponding lung collapsibility at PEEP level of 7 cmH2O was 41.2 ± 2.7 %/23.4 ± 8.6 %/11.6 ± 4.5 % (P = 0.0004 between 3 groups). At selected PEEP level, lung collapsibility was 1.5 ± 1.8 % for HR and 1.3 ± 1.9 % for LR. Conclusion: The information of alveolar recruitment did not aid in our final determination of PEEP level in ARDS patients while information about lung collapsibility by EIT may be of help. EIT based lung collapsibility may be considered for the determination of PEEP level in ARDS patients. References

Results: Data are presented as median [interquartile range] or mean ± SD. PEEP induced alveolar recruitment was 183.8 ml [22.1 - 340.3](PEEP level of 15 cmH2O to 7 cmH2O) and alveolar recruitment equal to or greater than the median (183.8 ml) as high recruiter (HR), otherwise it is low recruiter(LR). Degree of lung collapsibility between HR and LR at the PEEP level of 7 cmH2O was not significantly differed (25.2 ± 11.4 % in HR vs 21.4 ± 12.7 % in LR) and the final PEEP selected was not differed too. However, for patients with final selected PEEP level of 20 cmH2O/15 cmH2O/≤ 11cmH2O, the corresponding lung collapsibility at PEEP level of 7 cmH2O was 41.2 ± 2.7 %/23.4 ± 8.6 %/11.6 ± 4.5 % (P = 0.0004 between 3 groups). At selected PEEP level, lung collapsibility was 1.5 ± 1.8 % for HR and 1.3 ± 1.9 % for LR. Conclusion: The information of alveolar recruitment did not aid in our final determination of PEEP level in ARDS patients while information about lung collapsibility by EIT may be of help. EIT based lung collapsibility may be considered for the determination of PEEP level in ARDS patients. References 1. Wolf GK, Gomez-Laberge C, Rettig JS, Vargas SO, Smallwood CD, Prabhu SP, Vitali SH, Zurakowski D, Arnold JH, (2013) Mechanical ventilation guided by electrical impedance tomography in experimental acute lung injury. Critical care medicine 41: 1296–1304 2. Costa EL, Borges JB, Melo A, Suarez-Sipmann F, Toufen C, Jr., Bohm SH, Amato MB, (2009) Bedside estimation of recruitable alveolar collapse and hyperdistension by electrical impedance tomography. Intensive care medicine 35: 1132–1137

1. Wolf GK, Gomez-Laberge C, Rettig JS, Vargas SO, Smallwood CD, Prabhu SP, Vitali SH, Zurakowski D, Arnold JH, (2013) Mechanical ventilation guided by electrical impedance tomography in experimental acute lung injury. Critical care medicine 41: 1296–1304 2. Costa EL, Borges JB, Melo A, Suarez-Sipmann F, Toufen C, Jr., Bohm SH, Amato MB, (2009) Bedside estimation of recruitable alveolar collapse and hyperdistension by electrical impedance tomography. Intensive care medicine 35: 1132–1137 Grant acknowledgement National Science Concil National Cheng Kung University Hospital

1. Wolf GK, Gomez-Laberge C, Rettig JS, Vargas SO, Smallwood CD, Prabhu SP, Vitali SH, Zurakowski D, Arnold JH, (2013) Mechanical ventilation guided by electrical impedance tomography in experimental acute lung injury. Critical care medicine 41: 1296–1304 2. Costa EL, Borges JB, Melo A, Suarez-Sipmann F, Toufen C, Jr., Bohm SH, Amato MB, (2009) Bedside estimation of recruitable alveolar collapse and hyperdistension by electrical impedance tomography. Intensive care medicine 35: 1132–1137 Grant acknowledgement National Science Concil National Cheng Kung University Hospital A890 ULTRAPEEP: lung ultrasound for the assessment of lung recruitment during esophageal pressure-guided PEEP in ARDS F. Facchin1, F. Zarantonello1, G. Panciera1, A. De Cassai1, A. Venrdramin1, A. Ballin1, T. Tonetti1, P. Persona2, C. Ori1, L. Del Sorbo3, S. Rossi2 1University of Padova, Department of Medicine (DIMED), Padova, Italy; 2Azienda Ospedaliera di Padova, Emergency Department, Padova, Italy; 3University of Toronto, Interdepartmental Division of Critical Care Medicine, Toronto, Canada Correspondence: F. Facchin – University of Padova, Department of Medicine (DIMED), Padova, Italy Introduction: Whereas the importance of low tidal volume to avoid ventilator-induced lung injury (VILI) in patients with ARDS is well known, several uncertainties still exist regarding how to set positive end-expiratory pressure (PEEP). Many approaches have been considered, but no one showed a clear effectiveness in terms of outcome. Recently a ventilator strategy using esophageal pressure to estimate the transpulmonary pressure has been proposed by Talmor and colleagues1. Although they found an improvement in arterial oxygenation, it was not explored whether the increase in oxygenation was due to lung recruitment.

ffectiveness in terms of outcome. Recently a ventilator strategy using esophageal pressure to estimate the transpulmonary pressure has been proposed by Talmor and colleagues1. Although they found an improvement in arterial oxygenation, it was not explored whether the increase in oxygenation was due to lung recruitment. Objectives: The aim of this study was to assess whether the PEEP set to maintain a positive end-expiratory transpulmonary pressure (PL) is associated with an increase in lung recruitment estimated by lung ultrasound score (LUS)2. Methods: 12 patients with moderate and severe ARDS were enrolled. For the first 2 hours, PEEP was set according to the Acute Respiratory Distress Syndrome Network standard-of-care recommendations (phase A). It was then adjusted according to measurements of esophageal pressure for the following 2 hours (phase B) to maintain a positive PL at the end of expiration. The primary end point was the improvement in lung recruitment assessed with lung ultrasound.

Distress Syndrome Network standard-of-care recommendations (phase A). It was then adjusted according to measurements of esophageal pressure for the following 2 hours (phase B) to maintain a positive PL at the end of expiration. The primary end point was the improvement in lung recruitment assessed with lung ultrasound. Results: Lower end-expiratory PL was found during phase A compared to phase B [median −3.5 cmH2O [−4.5 to −2] vs 3 cmH2O [1 to 4.25], p = 0.002]. During phase B PEEP has been increased from a median of 10 cmH2O to 17.5 cmH2O (p = 0.002). Arterial oxygenation improved in phase B compared to phase A, (median PaO2/FiO2: phase A 146.5 mmHg [117 to 166] vs phase B 164.5 [155 to 180], p = 0.037). The median LUS was 19.5 during phase A (IQR 11.75 to 28) and decreased to 17.5 (IQR 10.25 to 19.75) during phase B (p = 0.028). A decrease in LUS ≥4 (indicating lung recruitment) was found in 8 of 12 patients. Conclusions: In patients with moderate-severe ARDS, PEEP-induced LUS reduction and increase in oxygenation appears to indicate that setting PEEP according to the esophageal pressure-guided method results in a greater alveolar recruitment than setting PEEP according to ARDSNet strategy. Further investigations are needed to assess the potential impact of PEEP-induced alveolar overdistension. References 1. Talmor D, et al. Mechanical ventilation guided by esophageal pressure in acute lung injury. N Engl J Med. 2008 Nov 13;359(20):2095–104.

Conclusions: In patients with moderate-severe ARDS, PEEP-induced LUS reduction and increase in oxygenation appears to indicate that setting PEEP according to the esophageal pressure-guided method results in a greater alveolar recruitment than setting PEEP according to ARDSNet strategy. Further investigations are needed to assess the potential impact of PEEP-induced alveolar overdistension. References 1. Talmor D, et al. Mechanical ventilation guided by esophageal pressure in acute lung injury. N Engl J Med. 2008 Nov 13;359(20):2095–104. 2. Bouhemad B, et al. Ultrasound assessment of antibiotic-induced pulmonary reaeration in ventilator-associated pneumonia. Crit Care Med. 2010 Jan;38(1):84–92.

Conclusions: In patients with moderate-severe ARDS, PEEP-induced LUS reduction and increase in oxygenation appears to indicate that setting PEEP according to the esophageal pressure-guided method results in a greater alveolar recruitment than setting PEEP according to ARDSNet strategy. Further investigations are needed to assess the potential impact of PEEP-induced alveolar overdistension. References 1. Talmor D, et al. Mechanical ventilation guided by esophageal pressure in acute lung injury. N Engl J Med. 2008 Nov 13;359(20):2095–104. 2. Bouhemad B, et al. Ultrasound assessment of antibiotic-induced pulmonary reaeration in ventilator-associated pneumonia. Crit Care Med. 2010 Jan;38(1):84–92. A891 Inspiratory recruitment in acute respiratory distress syndrome G. Vergani1, M. Cressoni1, D. Chiumello2,3, C. Chiurazzi4, M. Brioni1, I. Algieri1, T. Tonetti5, M. Guanziroli1, A. Colombo2, I. Tomic6, A. Colombo1, F. Crimella1, E. Carlesso1, V. Gasparovic6, L. Gattinoni5 1Università degli Studi di Milano, Dipartimento di Fisiopatologia Medico Chirurgica e dei Trapianti, Milano, Italy; 2Policlinico Di Milano, Dipartimento di Anestesia, Rianimazione, Urgenza ed Emergenza, Milano, Italy; 3Plug Working Group, Milan, Italy; 4Università degli Studi di Milano, Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti, Milano, Italy; 5Georg-August-University Goettingen, Anesthesiology and Intensive Care Medicine, Goettingen, Germany; 6Univesity of Zagreb, Department of Intensive Care Medicine, Rebro, Croatia Correspondence: C. Chiurazzi – Università degli Studi di Milano, Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti, Milano, Italy Introduction: In ARDS lung protective strategy implies a fully open lung. Recruitment maneuvers are used in ARDS (Acute Respiratory Distress Syndrome) patients to reinflate the lung. Two studies described the distribution of airway opening pressures in ARDS patients. [1][2] No data are available on the relationship between opening pressures and disease severity.

gy implies a fully open lung. Recruitment maneuvers are used in ARDS (Acute Respiratory Distress Syndrome) patients to reinflate the lung. Two studies described the distribution of airway opening pressures in ARDS patients. [1][2] No data are available on the relationship between opening pressures and disease severity. Objectives: To describe lung recruitment as a function of the transpulmonary pressure in mild, moderate and severe ARDS. Methods: ARDS patients underwent a low-dose end-expiratory CT scan at PEEP 5 cmH2O and three end-inspiratory CT scans at the plateau pressures reached starting from PEEP 5 cmH2O, 30 cmH2O and 45 cmH2O. In each of the CT slices, lung profiles were manually delineated, excluding hilar structures. Thereafter, quantitative analysis of CT scan images was performed and the gas and tissue fractions were computed. We defined the recruitability as the difference of not inflated tissue between 45 and 5 cmH2O, that we arbitrarily assumed to be the “full recruitment”. [3] The grams of recruited tissue were computed across the pressure intervals at which the CT scan were performed, as the differences of not aerated tissue. Airway and esophageal pressures were continuously measured and transpulmonary pressure was computed as: Driving airway pressure (cmH2O) - (esophageal plateau pressure (cmH2O) - esophageal end-expiratory pressure at PEEP (cmH2O)[1].

intervals at which the CT scan were performed, as the differences of not aerated tissue. Airway and esophageal pressures were continuously measured and transpulmonary pressure was computed as: Driving airway pressure (cmH2O) - (esophageal plateau pressure (cmH2O) - esophageal end-expiratory pressure at PEEP (cmH2O)[1]. Results: Thirty-three patients were studied, 5 with mild, 10 with moderate and 18 with severe ARDS, according to the Berlin definition.[2] As reported in the Table 20 and Fig. 23, the amount of tissue which can be opened between 30 and 45 cmH2O was 8 %, 17 % and 37 % respectively in mild, moderate and severe ARDS). Mild ARDS patients nearly completed recruitment at approximately 15 cmH2O transpulmonary pressure while in moderate and severe ARDS recruitment continues up to 26 cmH2O transpulmonary pressure. Conclusions: At the clinically recommended plateau pressure of 30 cmH2O, in severe ARDS, up to 1/3 of the lung tissue recruitable at 45 cmH2O, stays always closed. Beyond contributing to the gas exchange impairment (depending on the perfusion), these “always” collapsed regions may also act as stress risers at their interface with aerated regions, though they are theoretically protected from the mechanical ventilation. References [1] Crotti S, Am J Respir Crit Care Med 2001. [2] Borges JB, Am J Respir Crit Care Med 2006. [3] Gattinoni L, N Engl J Med 2006. [4] Chiumello D,Am J Respir Crit Care Med 2008. [5] ARDS Definition Task Force, JAMA J Am Med Assoc 2012.Table 20 (abstract A891). Recruited tissue

Conclusions: At the clinically recommended plateau pressure of 30 cmH2O, in severe ARDS, up to 1/3 of the lung tissue recruitable at 45 cmH2O, stays always closed. Beyond contributing to the gas exchange impairment (depending on the perfusion), these “always” collapsed regions may also act as stress risers at their interface with aerated regions, though they are theoretically protected from the mechanical ventilation. References [1] Crotti S, Am J Respir Crit Care Med 2001. [2] Borges JB, Am J Respir Crit Care Med 2006. [3] Gattinoni L, N Engl J Med 2006. [4] Chiumello D,Am J Respir Crit Care Med 2008. [5] ARDS Definition Task Force, JAMA J Am Med Assoc 2012.Table 20 (abstract A891). Recruited tissue Experimental step 6 ml/kg IBW from PEEP 5 cmH2O 30 cmH2O set from PEEP 5 cmH2O 45 cmH2O set from PEEP 5 cmH2O Mild ARDS Recruited tissue, g (% of recruitability) 63 ± 26 (51 ± 23 %) 56 ± 50 (41 ± 33 %) 10 ± 29 (8 ± 21 %) Moderate ARDS Recruited tissue, g (% of recruitability) 92 ± 53 (42 ± 20) 105 ± 129* (41 ± 22 %) 54 ± 86* (17 ± 27 %) Severe ARDS Recruited tissue, g (% of recruitability) 123 ± 94 (26 ± 17 %) 209 ± 180* (36 ± 22 %) 173 ± 112*$ (37 ± 18 %) Fig. 23 (abstract A891). Recruitment as function of applied TP pressures

± 21 %) Moderate ARDS Recruited tissue, g (% of recruitability) 92 ± 53 (42 ± 20) 105 ± 129* (41 ± 22 %) 54 ± 86* (17 ± 27 %) Severe ARDS Recruited tissue, g (% of recruitability) 123 ± 94 (26 ± 17 %) 209 ± 180* (36 ± 22 %) 173 ± 112*$ (37 ± 18 %) Fig. 23 (abstract A891). Recruitment as function of applied TP pressures A892 Associations between ventilator settings during extracorporeal membrane oxygenation and outcome in patients with acute respiratory distress syndrome: a pooled individual patient data analysis A. Serpa Neto1,2, M. Schmidt3,4, T. Pham5,6,7, A. Combes4,8, M. Gama de Abreu9, P. Pelosi10, M.J. Schultz2, for the ReVA Research Network and the PROVE Network Investigators 1Hospital Israelita Albert Einstein, Critical Care Medicine, São Paulo, Brazil; 2Academic Medical Center, University of Amsterdam, Intensive Care, Amsterdam, Netherlands; 3Australian and New Zealand Intensive Care Research Centre; School of Public Health, Monash University, Department of Epidemiology and Preventive Medicine, Melbourne, Australia; 4Institute of Cardiometabolism and Nutrition (iCAN); Hôpital de la Pitié-Salpêtrière, Assistance Publique-Hôpitaux de Paris, Medical-Surgical Intensive Care Unit, Paris, France; 5Groupe Hospitalier Henri Mondor; Assistance Publique-Hôpitaux de Paris, Service de Réanimation Médicale, Créteil, France; 6Inserm, Sorbonne Paris Cité, ECSTRA Team; Université Paris Diderot, UMR 1153, Paris, France; 7Inserm; Université Paris Est Créteil, UMR 915, Créteil, France; 8Université Pierre et Marie Curie-Paris VI, Service de Réanimation Médicale, Paris, France; 9University Hospital Carl Gustav Carus, Technische Universität Dresden, 21Pulmonary Engineering Group, Department of Anesthesiology and Intensive Care Medicine, Dresden, Germany; 10IRCCS 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: Extracorporeal membrane oxygenation (ECMO) is a rescue therapy for patients with acute respiratory distress syndrome (ARDS) by providing additional oxygenation, and removing carbon dioxide thus permitting less injurious mechanical ventilation settings that have been shown to protect the lungs from additional injury.

n: Extracorporeal membrane oxygenation (ECMO) is a rescue therapy for patients with acute respiratory distress syndrome (ARDS) by providing additional oxygenation, and removing carbon dioxide thus permitting less injurious mechanical ventilation settings that have been shown to protect the lungs from additional injury. Objectives: To evaluate associations between distinct ventilator settings during ECMO, and outcome of ARDS patients. Methods: Individual patient data analysis of observational studies in adult ARDS patients receiving ECMO for refractory hypoxemia. Multilevel multivariable logistic regression models and Cox-proportional hazards models were used to determine which settings and parameters had an independent association with the primary endpoint all-cause mortality.

ata analysis of observational studies in adult ARDS patients receiving ECMO for refractory hypoxemia. Multilevel multivariable logistic regression models and Cox-proportional hazards models were used to determine which settings and parameters had an independent association with the primary endpoint all-cause mortality. Results: Nine studies with 545 patients were selected (Figure 24). Initiation of ECMO was accompanied by significant decreases in tidal volume, positive end-expiratory pressure (PEEP), plateau pressure (Pplat), and driving pressure (ΔP = Pplat - PEEP), respiratory rate and minute volume (Figure 25), and resulted in higher PaO2 to FiO2 ratios, higher arterial pH and lower PaCO2 (Figure 26). Higher age, lower body mass index, and higher lactate were associated with all-cause mortality after multivariable adjustment. ΔP, both before and during the first three days of ECMO, demonstrated an independent association with all-cause mortality (adjusted OR, 1.17 [95 % CI 1.06-1.29] for ΔP before ECMO and adjusted OR, 1.13 [1.02-1.27] for ΔP during the first three days of ECMO). Ventilator FiO2 during the first three days of ECMO was independently associated with all-cause mortality (adjusted OR, 1.04 [95 % CI 1.01-1.08]). Conclusions: In this series of ARDS patients receiving ECMO for refractory hypoxia, ΔP and FiO2 were the only ventilatory variables that had an independent association with outcome. These findings indicate the potential for improvement in the management of patients with ARDS undergoing ECMO. References Schmidt M, et al. Crit Care Med 2015;43:654–64.

Conclusions: In this series of ARDS patients receiving ECMO for refractory hypoxia, ΔP and FiO2 were the only ventilatory variables that had an independent association with outcome. These findings indicate the potential for improvement in the management of patients with ARDS undergoing ECMO. References Schmidt M, et al. Crit Care Med 2015;43:654–64. Marhong JD, et al. Intensive Care Med 2015;41:994–1003.Fig. 24 (abstract A892). PRISMA-IPD flow diagram Fig. 25 (abstract A892). Ventilatory parameters during ECMO Fig. 26 (abstract A892). Laboratory parameters during ECMO

Conclusions: In this series of ARDS patients receiving ECMO for refractory hypoxia, ΔP and FiO2 were the only ventilatory variables that had an independent association with outcome. These findings indicate the potential for improvement in the management of patients with ARDS undergoing ECMO. References Schmidt M, et al. Crit Care Med 2015;43:654–64. Marhong JD, et al. Intensive Care Med 2015;41:994–1003.Fig. 24 (abstract A892). PRISMA-IPD flow diagram Fig. 25 (abstract A892). Ventilatory parameters during ECMO Fig. 26 (abstract A892). Laboratory parameters during ECMO A893 Deflation injury following release of sustained PEEP B.H. Katira1,2,3, D. Engelberts1, R.E. Giesinger4, C. Ackerley5, T. Yoshida1,2,3, D. Zabini6, G. Otulakowski1, M. Post1, W.M. Kuebler6,7, P.J. McNamara4, B.P. Kavanagh1,2,3 1Hospital for Sick Children, Physiology and Experimental Medicine, Toronto, Canada; 2Hospital for Sick Children, University of Toronto, Critical Care Medicine and Anesthesiology, Toronto, Canada; 3University of Toronto, Interdepartmental Division of Critical Care Medicine, Toronto, Canada; 4Hospital for Sick Children, University of Toronto, Neonatal Medicine, Toronto, Canada; 5Hospital for Sick Children, Division of Electron Microscopy, Toronto, Canada; 6Keenan Research Centre, Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, Canada; 7University of Toronto, Surgery and Physiology, Toronto, Canada Correspondence: B.H. Katira – Hospital for Sick Children, Physiology and Experimental Medicine, Toronto, Canada Introduction: ARDS continues to be a significant problem with high mortality. Sustained inflation of lungs is routinely used to recruit the alveoli in ARDS but multiple RCTs of sustained inflation have been negative despite the physiological benefits. This, we think is because of a hidden form of injury offsetting the positive effects of lung inflation - injury induced by lung deflation.

tality. Sustained inflation of lungs is routinely used to recruit the alveoli in ARDS but multiple RCTs of sustained inflation have been negative despite the physiological benefits. This, we think is because of a hidden form of injury offsetting the positive effects of lung inflation - injury induced by lung deflation. Aim: To determine if sudden deflation of normal lungs after sustained inflation causes lung injury, its nature, timing, ultrastructural impact, and the possible hemodynamic mechanisms responsible for this injury. Methods: Healthy Sprague Dawley rats were anaesthetised, ventilated (tidal volume 6 mL/kg), and randomized to Intervention or Control. Intervention was incremental increases in PEEP (3 to 11 cmH2O (over 70 min), abrupt deflation to ZEEP and ventilation for 30 min; Control was ventilation for 100 min (PEEP 3cmH2O). Lungs were analysed for wet-to-dry ratio, BAL protein, static compliance, SpO2 and histology. To detect the timing of injury, rats received Evans Blue dye (EBD-30 mg/kg IV) at the initiation and were euthanized immediately before lung deflation or at 2, 5, 10 or 20 min afterwards (4/group). Terminal BAL analysed for EBD absorbance. Ultrastructural impact was studied by electron microscopy on lungs sampled from rats euthanized before deflation, and at 1 and 5 min after deflation. Hemodynamic data was obtained by ECHO performed at baseline (PEEP 3 cmH2O), immediately before and after deflation, and at 30mins after deflation. RV pressure was measured with a Millar catheter.

s studied by electron microscopy on lungs sampled from rats euthanized before deflation, and at 1 and 5 min after deflation. Hemodynamic data was obtained by ECHO performed at baseline (PEEP 3 cmH2O), immediately before and after deflation, and at 30mins after deflation. RV pressure was measured with a Millar catheter. Results: Wet-to-dry ratio (6.1 ± 0.6 vs 4.6 ± 0.4; P = 0.00) and BAL protein (3.9 ± 4.0 vs 1.5 ± 0.7; P = 0.18) was higher; and static compliance (0.48 ± 0.97 vs 0.82 ± 0.2; P = 0.00) and SpO2 (67 ± 23.5 vs 91 ± 4.4; P = 0.04) were lower in Intervention vs Control. Histology revealed collapse, hemorrhage and neutrophil accumulation in the intervention group. BAL Evans Blue demonstrated that microvascular leak was absent before deflation and was maximal by 10 min of deflation. Ultrastructural analysis showed that sustained inflation caused minimal swelling of epithelium and endothelium before deflation; deflation resulted in major cellular and interstitial edema, and endothelial injury. Hemodynamic data showed that RV and LV were under-filled during inflation. Upon deflation, RV output, pulmonary vascular resistance, RV systolic transmural and diastolic pressures increased precipitously. RV/LV ratio increased progressively. Conclusion: Sudden deflation after sustained inflation with PEEP causes protein leak, inflammation, hypoxemia, reduced compliance, endothelial injury and RV failure. The mechanism appears to be endothelial injury resulting in microvascular leakage, pulmonary hypertension and RV failure.

Ultrastructural analysis showed that sustained inflation caused minimal swelling of epithelium and endothelium before deflation; deflation resulted in major cellular and interstitial edema, and endothelial injury. Hemodynamic data showed that RV and LV were under-filled during inflation. Upon deflation, RV output, pulmonary vascular resistance, RV systolic transmural and diastolic pressures increased precipitously. RV/LV ratio increased progressively. Conclusion: Sudden deflation after sustained inflation with PEEP causes protein leak, inflammation, hypoxemia, reduced compliance, endothelial injury and RV failure. The mechanism appears to be endothelial injury resulting in microvascular leakage, pulmonary hypertension and RV failure. Significance: Deflation injury may be an important entity to prevent when using sustained inflation manoeuvres and may explain -in part- why several important RCTs in ARDS have been negative.Fig. 27 (abstract A893). Airway Pressures in Intervention Group Fig. 28 (abstract A893). Microvascular leak Fig. 29 (abstract A893). Development of right ventricular dysfunction

Significance: Deflation injury may be an important entity to prevent when using sustained inflation manoeuvres and may explain -in part- why several important RCTs in ARDS have been negative.Fig. 27 (abstract A893). Airway Pressures in Intervention Group Fig. 28 (abstract A893). Microvascular leak Fig. 29 (abstract A893). Development of right ventricular dysfunction SEPSIS THERAPEUTICS A894 Low dose steroids reduce short term mortality in septic shock patients: results of an individual patient data meta-analysis R. Pirracchio1,2, M. Resche Rigon3, M. Carone4, S. Chevret3, D. Annane5 1Hopital Europeen Georges Pompidou, Anesthesiology and Intensive Care Medicine, Paris, France; 2University of California Berkeley, Berkeley, United States; 3Université Paris Diderot, Paris, France; 4University of Washington, Seattle, United States; 5Université de Versailles Saint-Quentin-en-Yvelines, Garches, France Correspondence: R. Pirracchio – Hopital Europeen Georges Pompidou, Anesthesiology and Intensive Care Medicine, Paris, France Introduction: Previous research has suggested that the use of low dose steroids may be beneficial during septic shock. However subsequent inconsistent results explain the lack of consensus amongst doctors around the world about whether treatment with low dose steroids does improve the overall recovery and survival in patients with septic shock. We hypothetize that the lack of consistent evidence on the effect of low-dose steroids on short term mortality may be related to underpower. treated for septic shock.

gst doctors around the world about whether treatment with low dose steroids does improve the overall recovery and survival in patients with septic shock. We hypothetize that the lack of consistent evidence on the effect of low-dose steroids on short term mortality may be related to underpower. treated for septic shock. Objectives: The primary objective of the present study was to estimate the effect of three different therapeutic regimens (hydrocortisone alone, hydrocortisone plus fludrocortisone, neither hydrocortisone nor fludrocortisone) on 28-day mortality in patients treated for septic shock using an individual patient data meta-analysis. Methods: Individual patient data meta-analysis including the 3 major recent randomized controlled trials comparing early low-dose short course hydrocortisone and fludrocortisone to placebo (GER-inf (1)), hydrocortisone alone to placebo (CORTICUS (2)) or hydrocortisone to hydrocortisone and fludrocortisone (COIITSS (3)) in septic shock patients. The primary outcome measure was all cause 28-day mortality. Secondary outcomes measures were 90-day mortality, resolution of organ dysfunction (as measured by the time to reach a Sequential Organ Failure Assessement score < 8), time to vasopressor and mechanical ventilation discontinuation, intensive care unit and hospital lengths of stay as well as the rate of superinfection. Treatment effect on the primary outcome was quantified using relative risk and estimated using targeted maximum likelihood estimation.

n Failure Assessement score < 8), time to vasopressor and mechanical ventilation discontinuation, intensive care unit and hospital lengths of stay as well as the rate of superinfection. Treatment effect on the primary outcome was quantified using relative risk and estimated using targeted maximum likelihood estimation. Results: A total of 1,307 patients were enrolled in the 3 trials. When compared to the placebo, hydrocortisone + fludrocortisone significantly reduced 28-day mortality (RR = 0.73, 95%CI = 0.67-0.79, p < 0.001). Hydrocortisone + fludrocortisone was also superior when compared to the placebo and hydrocortisone pooled together (RR = 0.86, 95%CI = 0.79- 0.94, p = 0.002). Hydrocortisone + fludrocortisone significantly decreased 28-day mortality (RR = 0.75, 95%CI = 0.67-0.85, p < 0.001) in the nonresponders, while it was associated with an increase in 28-day mortality in the responders (RR = 1.17, 95%CI = 1.06-1.29, p = 0.002) (Figure 30). Hydrocortisone + fludrocortisone was also superior when considering secondary outcomes such as vasopressor discontinuation or lengths of stay. Conclusions: In this individual patient data meta-analysis including the 3 major randomized controlled trials on the subject, we found that an early short course of low-dose hydrocortisone and fludrocortisone decreases 28-day mortality and improves recovery from organ failure in septic shock patients non responding to a corticotropin stimulation test. References (1) JAMA 2002 21;288(7):862–71; (2) NEJM 2008 10;358(2):111–24;

Conclusions: In this individual patient data meta-analysis including the 3 major randomized controlled trials on the subject, we found that an early short course of low-dose hydrocortisone and fludrocortisone decreases 28-day mortality and improves recovery from organ failure in septic shock patients non responding to a corticotropin stimulation test. References (1) JAMA 2002 21;288(7):862–71; (2) NEJM 2008 10;358(2):111–24; (3) JAMA 2010 27;303(4):341–8Fig. 30 (abstract A894). Effect of hydro + fludro vs placebo or hydro alone A895 The clinical outcome of simvastatin plus standard therapy versus standard therapy alone in critically ill septic patients: randomized controlled clinical trial S. Eladawy1, M. El-Hamamsy1, N. Bazan2, M. Elgendy3 1Faculty of Pharmacy-Ain Shams University, Cairo, Egypt; 2Critical Care Medicine Department Cairo University Hospitals, Cairo, Egypt; 3Faculty of Medicine - Ain Shams University, Cairo, Egypt Correspondence: S. Eladawy – Faculty of Pharmacy-Ain Shams University, Cairo, Egypt Introduction: Statin therapy during intensive care unit (ICU) stay has been associated with a reduction in all-cause hospital mortality in some studies. This association was especially noted in septic patients. However, potential benefit needs to be validated in randomized, controlled trials. Objectives: The purpose of this study was to compare the effect of simvastatin plus standard therapy on mortality and total ICU length of stay (LOS) to that of standard therapy alone in critically ill septic patients.

A895 The clinical outcome of simvastatin plus standard therapy versus standard therapy alone in critically ill septic patients: randomized controlled clinical trial S. Eladawy1, M. El-Hamamsy1, N. Bazan2, M. Elgendy3 1Faculty of Pharmacy-Ain Shams University, Cairo, Egypt; 2Critical Care Medicine Department Cairo University Hospitals, Cairo, Egypt; 3Faculty of Medicine - Ain Shams University, Cairo, Egypt Correspondence: S. Eladawy – Faculty of Pharmacy-Ain Shams University, Cairo, Egypt Introduction: Statin therapy during intensive care unit (ICU) stay has been associated with a reduction in all-cause hospital mortality in some studies. This association was especially noted in septic patients. However, potential benefit needs to be validated in randomized, controlled trials. Objectives: The purpose of this study was to compare the effect of simvastatin plus standard therapy on mortality and total ICU length of stay (LOS) to that of standard therapy alone in critically ill septic patients. Methods: A prospective randomized, open label, controlled pilot clinical trial was conducted on patients diagnosed with sepsis/severe sepsis as defined by the American College of Chest Physicians (ACCP). Hundred patients met the study criteria and were randomized into two groups; a standard group who received standard treatment and simvastatin group who received the standard treatment plus 40 mg simvastatin. Primary outcomes were 28 days ICU mortality and total ICU LOS. Plasma C-reactive protein (CRP), total creatine kinase (CK) and liver enzymes [alanine aminotransferase (ALT) and aspartate aminotransferase (AST)] were measured as secondary outcome measures.

roup who received the standard treatment plus 40 mg simvastatin. Primary outcomes were 28 days ICU mortality and total ICU LOS. Plasma C-reactive protein (CRP), total creatine kinase (CK) and liver enzymes [alanine aminotransferase (ALT) and aspartate aminotransferase (AST)] were measured as secondary outcome measures. Results: A total of 72 patients completed the study. Simvastatin was well tolerated, with no increase in adverse events between the two groups. Total ICU LOS was significantly lower in the simvastatin group. However, the number of patients with 28 days ICU mortality in the simvastatin group was lower compared to standard group; but survival failed to reach statistical significance. Similarly, plasma C-reactive protein failed to reach statistical significance between the two groups Conclusions: Treatment with simvastatin 40 mg in patients with sepsis/severe sepsis is safe and associated with an improvement in number of deaths and ICU LOS but without subsequent improvement in survival.

Results: A total of 72 patients completed the study. Simvastatin was well tolerated, with no increase in adverse events between the two groups. Total ICU LOS was significantly lower in the simvastatin group. However, the number of patients with 28 days ICU mortality in the simvastatin group was lower compared to standard group; but survival failed to reach statistical significance. Similarly, plasma C-reactive protein failed to reach statistical significance between the two groups Conclusions: Treatment with simvastatin 40 mg in patients with sepsis/severe sepsis is safe and associated with an improvement in number of deaths and ICU LOS but without subsequent improvement in survival. A896 The use of anapnoguard 100 system in intubated critically ill patients a randomized controlled study G. De Pascale, M.S. Vallecoccia, S.L. Cutuli, V. Di Gravio, M.A. Pennisi, G. Conti, M. Antonelli Sacro Cuore Catholic University, A. Gemelli Hospital, Department of Anesthesiology and Intensive Care, Rome, Italy Correspondence: G. De Pascale – Sacro Cuore Catholic University, A. Gemelli Hospital, Department of Anesthesiology and Intensive Care, Rome, Italy Introduction: The AnapnoGuard 100 system (AG) (Hospitech Respiration LTD., Petach-Tikva, Israel) is an innovative respiratory guard system that continuously monitors and controls the cuff pressure by measurements of CO2 levels above the cuff, and allowing simultaneous rinsing and aspiration of subglottic secretions. Objectives: To determine the safety and clinical efficacy of AG 100 system compared with usual care in critically ill patients.

A896 The use of anapnoguard 100 system in intubated critically ill patients a randomized controlled study G. De Pascale, M.S. Vallecoccia, S.L. Cutuli, V. Di Gravio, M.A. Pennisi, G. Conti, M. Antonelli Sacro Cuore Catholic University, A. Gemelli Hospital, Department of Anesthesiology and Intensive Care, Rome, Italy Correspondence: G. De Pascale – Sacro Cuore Catholic University, A. Gemelli Hospital, Department of Anesthesiology and Intensive Care, Rome, Italy Introduction: The AnapnoGuard 100 system (AG) (Hospitech Respiration LTD., Petach-Tikva, Israel) is an innovative respiratory guard system that continuously monitors and controls the cuff pressure by measurements of CO2 levels above the cuff, and allowing simultaneous rinsing and aspiration of subglottic secretions. Objectives: To determine the safety and clinical efficacy of AG 100 system compared with usual care in critically ill patients. Methods: Prospective, single centre, open-label, randomized, controlled feasibility and safety trial. Sixty patients, without pneumonia, were randomized to be intubated with the AG tube and connected to the system (n = 30) or with a conventional tube (n = 30) combined with subglottic secretion drainage and manually control of tracheal cuff pressure (Pcuff). Primary outcome was the rate of adverse events. Other outcomes included the rate of mechanical complications, the level of ICU staff satisfaction, the incidence of ventilator-associated pneumonia (VAP), the quality of Pcuff control, and the amount of SS drained.

e and manually control of tracheal cuff pressure (Pcuff). Primary outcome was the rate of adverse events. Other outcomes included the rate of mechanical complications, the level of ICU staff satisfaction, the incidence of ventilator-associated pneumonia (VAP), the quality of Pcuff control, and the amount of SS drained. Results: Out of 60 patients enrolled in the study, 56 were included in the analysis (28 per each group). Both groups were similar at randomization in demographic characteristics, ICU admission diagnosis, main comorbidities and severity of illness. No device-related adverse events occurred in any of the two groups. No differences were detected using AG system vs conventional tubes in terms of post-extubation throat pain level (0[0–2] vs. 0[0–3]; p = 0.7), hoarseness (42.9 % vs. 75 %; p = 0.55) and tracheal mucosa oedema (16.7 % vs. 10 %; p = 0.65). On the basis of a predefined questionnaire (0–5 grading scale), ICU staff satisfaction level was high. The use of AG system was associated with a significantly higher percentage of Pcuff determinations in the safety range (97.3 % vs. 71.2 %; p < 0.001) and with a trend to a better daily subglottic secretions drainage (67.8 [20–88.7] ml vs. 50 [18.7-62] ml; p = 0.11). Patients enrolled in the AG group showed a lower percentage of ventilator-associated pneumonia (14.8 % vs. 40 %; p = 0.06) which were more frequently monomicrobial (25 % vs. 70 %; p = 0.2). No statistically significant between-group differences were observed in durations of mechanical ventilation, ICU stay, and mortality.

ients enrolled in the AG group showed a lower percentage of ventilator-associated pneumonia (14.8 % vs. 40 %; p = 0.06) which were more frequently monomicrobial (25 % vs. 70 %; p = 0.2). No statistically significant between-group differences were observed in durations of mechanical ventilation, ICU stay, and mortality. Conclusions: The use AG 100 system in critically ill intubated patients is safe and effective in Pcuff control ad SS drainage. The observed promising role as a tool to prevent VAP needs to be confirmed in a larger, adequately powered, randomized trial. Clinical trial registered with www.clinicaltrials.gov (NCT01550978). References 1. Vallecoccia MS, De Pascale G, Cutuli SL, Di Gravio V, Pennisi MA, Antonelli M. Endotracheal tubes cuff pressure control: does the CO2 matter? Minerva Anestesiol. 2015 Mar;81(3):352–3.

Conclusions: The use AG 100 system in critically ill intubated patients is safe and effective in Pcuff control ad SS drainage. The observed promising role as a tool to prevent VAP needs to be confirmed in a larger, adequately powered, randomized trial. Clinical trial registered with www.clinicaltrials.gov (NCT01550978). References 1. Vallecoccia MS, De Pascale G, Cutuli SL, Di Gravio V, Pennisi MA, Antonelli M. Endotracheal tubes cuff pressure control: does the CO2 matter? Minerva Anestesiol. 2015 Mar;81(3):352–3. A897 Survival impact of β-blockade in a long-term model of fluid-resuscitated sepsis depends on prognostic risk D.T. Andreis1,2, W. Khaliq1, M. Singer1 1Bloomsbury Institute of Intensive Care Medicine, University College London, London, United Kingdom; 2Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti, Milan, Italy Correspondence: D.T. Andreis – Bloomsbury Institute of Intensive Care Medicine, University College London, London, United Kingdom Introduction: During sepsis, intrinsic stress responses may become maladaptive and contribute to poor outcomes. Targeted intervention with β-blockade to 'de-stress' such patients may be beneficial. We developed a 72-h rodent model of fluid-resuscitated faecal peritonitis in which mortality (occurring between 18 and 42 h) can be predicted at 6 h by a low stroke volume (AUROC 0.87), and where survivors are clinically improving by study end.[1] Objectives: To investigate the impact of β-blockade on outcomes in predicted survivors and nonsurvivors of faecal peritonitis.

A897 Survival impact of β-blockade in a long-term model of fluid-resuscitated sepsis depends on prognostic risk D.T. Andreis1,2, W. Khaliq1, M. Singer1 1Bloomsbury Institute of Intensive Care Medicine, University College London, London, United Kingdom; 2Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti, Milan, Italy Correspondence: D.T. Andreis – Bloomsbury Institute of Intensive Care Medicine, University College London, London, United Kingdom Introduction: During sepsis, intrinsic stress responses may become maladaptive and contribute to poor outcomes. Targeted intervention with β-blockade to 'de-stress' such patients may be beneficial. We developed a 72-h rodent model of fluid-resuscitated faecal peritonitis in which mortality (occurring between 18 and 42 h) can be predicted at 6 h by a low stroke volume (AUROC 0.87), and where survivors are clinically improving by study end.[1] Objectives: To investigate the impact of β-blockade on outcomes in predicted survivors and nonsurvivors of faecal peritonitis. Methods: Instrumented, fluid resuscitated, male Wistar rats (300–400 g) had sepsis induced by intraperitoneal injection of faecal slurry (8.5 ml/kg). At 6 h, under brief isoflurane sedation, echocardiography was performed to differentiate predicted survivors from nonsurvivors based on a stroke volume cut-off of 0.20 ml. Rats in each prognostic group were then randomised to receive either esmolol (500 μg/kg over 1 min followed by 75 μg/kg/min infusion) or matching placebo (0.9 % NaCl) until 24 h. Animals were observed for up to 72 h, and time of death was recorded. The study was powered to detect a mortality reduction in predicted nonsurvivors from 90 % to 45 % with esmolol, with a power of 0.80 and type-1 error of 0.05.

over 1 min followed by 75 μg/kg/min infusion) or matching placebo (0.9 % NaCl) until 24 h. Animals were observed for up to 72 h, and time of death was recorded. The study was powered to detect a mortality reduction in predicted nonsurvivors from 90 % to 45 % with esmolol, with a power of 0.80 and type-1 error of 0.05. Results: 64 rats were randomised after prognostication to receive either esmolol or placebo. At 6 h, predicted survivors and nonsurvivors were clinically indistinguishable (both groups appeared only mildly unwell), though predicted nonsurvivors (stroke volume < 0.20 ml) had lower cardiac output (86 ± 9 vs. 130 ± 25 ml/min), higher heart rate (487 ± 36 vs. 459 ± 24 bpm) and blood pressure (131 ± 12 vs. 125 ± 10 mmHg) and more haemoconcentration (haemoglobin 16.9 ± 1.6 vs. 15.0 ± 1.3 g/dl) (all p < 0.05). Survival was significantly improved by esmolol in predicted nonsurvivors (p = 0.004), but worsened in predicted survivors (p = 0.06). Conclusions: Mortality was approximately halved in predicted nonsurvivors by esmolol, but doubled in predicted survivors. Early prognostication appears key in identifying the subset(s) of animals (and, potentially, patients) who might benefit from additional treatment, while avoiding iatrogenic harm in those that would naturally survive. Mechanisms by which esmolol impact upon mortality are under investigation. References [1] Rudiger A et al. Clin Sci 2013; 124: 391–401. Grant acknowledgement

Conclusions: Mortality was approximately halved in predicted nonsurvivors by esmolol, but doubled in predicted survivors. Early prognostication appears key in identifying the subset(s) of animals (and, potentially, patients) who might benefit from additional treatment, while avoiding iatrogenic harm in those that would naturally survive. Mechanisms by which esmolol impact upon mortality are under investigation. References [1] Rudiger A et al. Clin Sci 2013; 124: 391–401. Grant acknowledgement ESICM Basic Science Award, UK Intensive Care Society Young Investigator Award, NIHR.Fig. 31 (abstract A897). Effect of esmolol (blue) vs. placebo (orange) on mortality in the overall population (left, n = 32 per group), predicted survivors (centre, n = 16 per group) and predicted nonsurvivors (right, n = 16 per group). P values refer to log rank test A898 Polymyxin coated cytokine adsorbent for supportive treatment of sepsis J. Hartmann, S. Harm Danube University, Department for Health Sciences and Biomedicine, Krems, Austria Correspondence: S. Harm – Danube University, Department for Health Sciences and Biomedicine, Krems, Austria Introduction: Endotoxins (lipopolysaccharides, LPS) have become interesting targets in extracorporeal therapies. LPS is a major constituent of the outer cell wall of gram-negative bacteria and strongly triggers inflammatory responses in humans at concentrations as low as 1 ng/kg body weight. Although the elimination of LPS is promising for the supportive therapy of sepsis and liver failure, endotoxin neutralization using endotoxin adsorbents is controversial.

the outer cell wall of gram-negative bacteria and strongly triggers inflammatory responses in humans at concentrations as low as 1 ng/kg body weight. Although the elimination of LPS is promising for the supportive therapy of sepsis and liver failure, endotoxin neutralization using endotoxin adsorbents is controversial. Objectives: We could recently show that endotoxin inactivation by low-dose Polymyxin B (PMB; 200 ng/ml) could be applied for endotoxin inactivation in blood [1]. Aim of this study was to establish an adsorbent-based system which combines constant PMB release for endotoxin inactivation and effective cytokine adsorption during extracorporeal treatment. Methods: We established an adsorbent-based PMB release system which ensures a constant PMB level in plasma during extracorporeal therapies. A polystyrene-divinylbenzene based cytokine adsorbent (CG161c) with nanostructured pores was coated with a defined amount of PMB by hydrophobic interactions. The endotoxin inactivation and cytokine adsorption was tested in an in vitro model using fresh donated blood which was stimulated with 1 ng/ml lipopolysaccaride from E. coli.

e-divinylbenzene based cytokine adsorbent (CG161c) with nanostructured pores was coated with a defined amount of PMB by hydrophobic interactions. The endotoxin inactivation and cytokine adsorption was tested in an in vitro model using fresh donated blood which was stimulated with 1 ng/ml lipopolysaccaride from E. coli. Results: In plasma or blood an equilibration between the free and bound form of PMB will lead to a constant PMB level in plasma. The PMB release experiments in plasma clearly show that the adsorption and desorption is a function of the ratio PMB concentration: adsorbent surface. Furthermore the PMB release depends on the protein concentration of the plasma. It makes a big difference whether the PMB coated adsorbent is used in plasma or in fractionated plasma where the hydrophobicity is much lower. The experiments suggest that the PMB coating of the CG161c adsorbent doesn´t influence the cytokine removal which can take place in parallel. The ability of LPS inactivation by the PMB coated CG161c adsorbents was similar to PMB which was infused directly into the plasma. Conclusions: Our in vitro model shows that the combination of cytokine removal and controlled PMB release by the same adsorbent results in a strong suppression of inflammatory effects in blood. References [1] Harm S, Hartmann J: Low-Dose Polymyxin: an Option for Therapy of Gram-Negative Sepsis. Innate Immunity, 2016, accepted for publication. Grant acknowledgement This work was supported by the government of Lower Austria within the project ID WST3-T-91/036-2014).

Conclusions: Our in vitro model shows that the combination of cytokine removal and controlled PMB release by the same adsorbent results in a strong suppression of inflammatory effects in blood. References [1] Harm S, Hartmann J: Low-Dose Polymyxin: an Option for Therapy of Gram-Negative Sepsis. Innate Immunity, 2016, accepted for publication. Grant acknowledgement This work was supported by the government of Lower Austria within the project ID WST3-T-91/036-2014). Cardivascular Challenges In The ICU A899 Local intraarterial thrombolysis for the management of patients with hemodynamically stable pulmonary embolism and right ventricular dysfunction S. Alcantara Carmona1, P. Matia Almudevar1, A. Naharro Abellán1, J. Veganzones Ramos1, L. Pérez Pérez, B. Lobo Valbuena, N. Martínez Sanz, I. Fernández Simón Hospital Universtario Puerta de Hierro Majadahonda, Intensive Care Unit, Madrid, Spain Correspondence: S. Alcantara Carmona –Hospital Universtario Puerta de Hierro Majadahonda, Intensive Care Unit, Madrid, Spain Objective: Management of hemodynamically stable pulmonary embolism (PE) with right ventricular (RV) dysfunction is still controversial. The objective of our study is to evaluate the effectiveness of local intraarterial thrombolysis (LIT) in this group of patients and analyze its complications.

Care Unit, Madrid, Spain Objective: Management of hemodynamically stable pulmonary embolism (PE) with right ventricular (RV) dysfunction is still controversial. The objective of our study is to evaluate the effectiveness of local intraarterial thrombolysis (LIT) in this group of patients and analyze its complications. Patients and methods: Prospective study (January 2008-December 2105). Patients included had been diagnosed of PE by computed tomography (CT), were hemodynamically stable [systolic arterial pressure (SAP) > 90 mmHg] and had a clinical suspicion of RV dysfunction (biventricular quotient in CT > 1 or elevated levels of troponin I), that was confirmed afterwards by the presence of at least one of the following findings in the echocardiogram: subjective alteration of RV contractility, RV basal diameter (four chamber view) > 40 mm, tricuspid annular plane systolic excursion (TAPSE) < 15 mm or estimated systolic pulmonary artery pressure (SPAP) > 30 mmHg. LIT was done with a urokinase infusion (bolus dose of 200.000 UI followed by a perfusion of 100.000 UI/h) administered thru a pulmonary artery catheter, placed with radiological guidance, using an antecubital puncture. Patients received simultaneous systemic anticoagulation with unfractionated heparin. After 48–72 h of treatment, and before ending the urokinase infusion, a radiological control was done using angiography or CT. Within the seven days after LIT, patients underwent a follow-up echocardiogram. Statistical analysis was performed with Student´s T test for parametric paired data, Wilcoxon´s Test for non parametric and Stuart-Maxwell for qualitative values.

urokinase infusion, a radiological control was done using angiography or CT. Within the seven days after LIT, patients underwent a follow-up echocardiogram. Statistical analysis was performed with Student´s T test for parametric paired data, Wilcoxon´s Test for non parametric and Stuart-Maxwell for qualitative values. Results: Eighty-seven patients were included and their general data are detailed in Fig. 32. Mean treatment time was 56,3 ± 15,5 h. Ninety percent of patients experienced a radiological improvement (50.5 % a complete/almost complete resolution and, 37.8 % a significant improvement). Only 5,7 % didn´t improve radiologically. The evolutions of the different RV parameters studied are shown in Fig. 33. Minimum fibrinogen and platelet values where 251,9 ± 91,8 mg/dl and 134 x 103 ± 44.6 x 103 cells/mm3. Eighteen patients (20,7 %) suffered form hemorrhagic complications that, in 13 cases, where puncture site hematomas and, in six occasions (6,9 %) required an early interruption of the treatment. Three patients (3,4 %) received a blood cell transfusion of ≤ 2 blood units. Mean ICU and hospital stays where 4 ± 1,5 and 14 ± 9,2 days. All patients survived. Conclusion: In our group of patients, LIT rapidly improved the function and decreased the hemodynamic strain of the RV, while being associated with a low incidence of major complications.Fig. 32 (abstract A899). ᅟ Fig. 33 (abstract A899). ᅟ

Results: Eighty-seven patients were included and their general data are detailed in Fig. 32. Mean treatment time was 56,3 ± 15,5 h. Ninety percent of patients experienced a radiological improvement (50.5 % a complete/almost complete resolution and, 37.8 % a significant improvement). Only 5,7 % didn´t improve radiologically. The evolutions of the different RV parameters studied are shown in Fig. 33. Minimum fibrinogen and platelet values where 251,9 ± 91,8 mg/dl and 134 x 103 ± 44.6 x 103 cells/mm3. Eighteen patients (20,7 %) suffered form hemorrhagic complications that, in 13 cases, where puncture site hematomas and, in six occasions (6,9 %) required an early interruption of the treatment. Three patients (3,4 %) received a blood cell transfusion of ≤ 2 blood units. Mean ICU and hospital stays where 4 ± 1,5 and 14 ± 9,2 days. All patients survived. Conclusion: In our group of patients, LIT rapidly improved the function and decreased the hemodynamic strain of the RV, while being associated with a low incidence of major complications.Fig. 32 (abstract A899). ᅟ Fig. 33 (abstract A899). ᅟ A900 New-onset atrial fibrillation has detrimental effects on in-hospital but not on long-term survival in critically ill patients: a report from the FROG-ICU study M. Arrigo1, E. Feliot1, N. Deye1, A. Cariou2, B. Guidet3, S. Jaber4, M. Leone5, M. Resche-Rigon6, A. Vieillard Baron7, M. Legrand1, E. Gayat1, A. Mebazaa1, from the FROG ICU Investigators 1AP-HP, Saint Louis and Lariboisière University Hospitals, Department of Anesthesiology and Critical Care Medicine, Paris, France; 2AP-HP, Hôpital Cochin, Medical Intensive Care Unit, Paris, France; 3AP-HP, Hôpital Saint-Antoine, Medical Intensive Care Unit, Paris, France; 4Saint-Eloi University Hospital, Department of Anesthesiology and Critical Care Medicine, Montpellier, France; 5AP-HM, Hôpital Nord, Department of Anesthesiology and Critical Care Medicine, Marseille, France; 6AP-HP, Saint Louis University Hospital, Biostatistics and Clinical Epidemiology Research Unit Department, Paris, France; 7AP-HP, Hôpital Ambroise Paré, Department of Anesthesiology and Critical Care Medicine, Paris, France Correspondence: M. Arrigo – AP-HP, Saint Louis and Lariboisière University Hospitals, Department of Anesthesiology and Critical Care Medicine, Paris, France Introduction: Atrial fibrillation (AFib) is associated with higher short-term mortality in critical illness, but it is still uncertain whether AFib independently contributes to unfavorable outcome.

Saint Louis and Lariboisière University Hospitals, Department of Anesthesiology and Critical Care Medicine, Paris, France Introduction: Atrial fibrillation (AFib) is associated with higher short-term mortality in critical illness, but it is still uncertain whether AFib independently contributes to unfavorable outcome. Objectives: The aim of this study was to test the hypothesis that AFib during critical illness is independently associated with increased in-hospital and long-term risk of death. Methods: The FROG-ICU study was a prospective, observational, multi-center cohort study designed to investigate outcome of critically ill patients. Heart rhythm was assessed at inclusion and during ICU stay with digital ECG recordings. Among patients who had any AFib during ICU stay, new-onset and recurrent AFib were diagnosed in patients without and with previous history of AFib, respectively. Primary endpoints were in-hospital and 1-year mortality. Covariate adjusted logistic regression models and Cox proportional hazards models were used to evaluate the association between AFib and in-hospital mortality or 1-year mortality, respectively. In-hospital mortality was adjusted for 7 independent covariates (age, gender, Simplified Acute Physiology Score (SAPS II), treatment with inotropes or vasopressors, serum lactate level, high-sensitive troponin I, B-type natriuretic peptide), 1-year mortality was adjusted for 8 covariates (age, gender, SAPS II, history of congestive heart failure, treatment with inotropes or vasopressors, serum lactate level, C-reactive protein and serum creatinine).

with inotropes or vasopressors, serum lactate level, high-sensitive troponin I, B-type natriuretic peptide), 1-year mortality was adjusted for 8 covariates (age, gender, SAPS II, history of congestive heart failure, treatment with inotropes or vasopressors, serum lactate level, C-reactive protein and serum creatinine). Results: The study included 2087 critically ill patients. The study population consisted of 1841 patients for whom data about heart rhythm during ICU stay was available. AFib occurred in 343 patients (19 %). New-onset AFib (n = 212) had higher in-hospital mortality (47 %) compared to no AFib (23 %, P < 0.001) or recurrent AFib (34 %, P = 0.032). New-onset AFib showed increased in-hospital risk of death after multivariable adjustment compared to no AFib (OR 1.6, 95 % CI 1.2-2.2, P = 0.003) or recurrent AFib (OR 1.8, 95 % CI 1.1-2.9, P = 0.02). Among the 1464 ICU-survivors, new-onset AFib during ICU stay showed higher 1-year mortality compared to no AFib (log-rank P < 0.001), but similar to recurrent AFib. After multivariable adjustment, new-onset AFib showed similar 1-year risk of death compared to no AFib (HR 1.3, 95 % CI 0.9-1.8, P = 0.17) or recurrent AFib (HR 1.1, 95 % CI 0.7-1.7, P = 0.81). Conclusions: New-onset AFib independently increases in-hospital but not long-term risk of death of critically ill patients. Grant acknowledgement The FROG-ICU study was supported by the national PHRC and the SFAR

Results: The study included 2087 critically ill patients. The study population consisted of 1841 patients for whom data about heart rhythm during ICU stay was available. AFib occurred in 343 patients (19 %). New-onset AFib (n = 212) had higher in-hospital mortality (47 %) compared to no AFib (23 %, P < 0.001) or recurrent AFib (34 %, P = 0.032). New-onset AFib showed increased in-hospital risk of death after multivariable adjustment compared to no AFib (OR 1.6, 95 % CI 1.2-2.2, P = 0.003) or recurrent AFib (OR 1.8, 95 % CI 1.1-2.9, P = 0.02). Among the 1464 ICU-survivors, new-onset AFib during ICU stay showed higher 1-year mortality compared to no AFib (log-rank P < 0.001), but similar to recurrent AFib. After multivariable adjustment, new-onset AFib showed similar 1-year risk of death compared to no AFib (HR 1.3, 95 % CI 0.9-1.8, P = 0.17) or recurrent AFib (HR 1.1, 95 % CI 0.7-1.7, P = 0.81). Conclusions: New-onset AFib independently increases in-hospital but not long-term risk of death of critically ill patients. Grant acknowledgement The FROG-ICU study was supported by the national PHRC and the SFAR A901 Antiarrhythmic therapy for supraventricular arrhythmias in septic shock M. Balik1, I. Kolnikova1, M. Maly1, P. Waldauf2, G. Tavazzi3, J. Kristof1 11st Medical Faculty, Charles University, Prague, Czech Republic; 23rd Medical Faculty, Charles University, Prague, Czech Republic; 3University of Pavia, Pavia, Italy Correspondence: M. Balik – 1st Medical Faculty, Charles University, Prague, Czech Republic Introduction: The incidence of the supraventricular arrhythmias is increased in septic shock patient, and it is associated with worse short and long term prognosis.

sity, Prague, Czech Republic; 3University of Pavia, Pavia, Italy Correspondence: M. Balik – 1st Medical Faculty, Charles University, Prague, Czech Republic Introduction: The incidence of the supraventricular arrhythmias is increased in septic shock patient, and it is associated with worse short and long term prognosis. Objective: To test that propafenon could be a feasible antiarrhythmic in the absence of contraindications. Methods: Patients with septic shock who received antiarrhythmic drugs for supraventricular arrhythmias were included over 24 months. The patients were divided into the three groups according to antiarrhythmic agent: amiodarone (Group1), propafenon (Group2) and metoprolol (Group3). In the first 24 h the type of arrhythmia, dosages, cardioversion rates, demographic, haemodynamic, laboratory parameters were recorded. Mortality was compared between the groups and between the cardioverted vs those remaining in acute and chronic arrhythmias. Results: 234 (17.3 % of all ICU) patients with rates of IPPV 99.1 % were included, 14.5 % had chronic atrial fibrillation (AF). Prevailing arrhythmia was acute onset AF(69.7 %). The rates of additional electric cardioversion to secure sinus rhythm were not different (23.7 % Group1, 35.5 % Group2, ns). Except for the dosage of noradrenaline the mortality predictors were not different between the three groups.

ad chronic atrial fibrillation (AF). Prevailing arrhythmia was acute onset AF(69.7 %). The rates of additional electric cardioversion to secure sinus rhythm were not different (23.7 % Group1, 35.5 % Group2, ns). Except for the dosage of noradrenaline the mortality predictors were not different between the three groups. ICU (30.9 %) and 28-day (38.8 %) mortalities of cardioverted patients were not significantly different from mortality of chronic AF patients (ICU 38.2 %, OR 1.38, 28-day 41.2 % OR 1.11) and both were lower than the ICU (54.3 %, OR 2.64 vs sinus, OR 1.92 vs chronic AF) and 28-day mortality (65 %, OR 2.98 vs sinus, OR 2.69 vs chronic AF) of those remaining in an acute arrhythmia. Conclusions: A chance to cardiovert a ventilated septic shock patient seems to be higher under propafenon than in amiodarone with a promising impact upon ICU and 28-day mortality. The patients remaining in acute onset arrhythmia demonstrated significantly higher ICU and 28-day mortality compared to those succesfully cardioverted or to those having chronic atrial fibrillation.Table 21 (abstract A901). ᅟ

her under propafenon than in amiodarone with a promising impact upon ICU and 28-day mortality. The patients remaining in acute onset arrhythmia demonstrated significantly higher ICU and 28-day mortality compared to those succesfully cardioverted or to those having chronic atrial fibrillation.Table 21 (abstract A901). ᅟ Logistic regression: ICU and 28-day mortality Group1: Amiodarone (n = 142) Group2: Propafenon (n = 78) Group3: Metoprolol (n = 14) APACHE II: OR 1.74 and 1.9 (p < 0.001) 26 (16–33) 23.5 (17.5–33.5) 20 (6–28) SOFA: OR 3.79 and 2.88 (p < 0.001) 10 (8–14) 11 (7.5–14.5) 9 (4–11) EF_LV [%] 45 (30–60) 50 (40–60) (ns) 48(35–58) (ns) NAD [ug/kg.min]: OR 2.02 and 1.7 (p < 0.001) 0.35 (0.14–0.80) 0.25 (0.08–0.50) p < 0.01 vs Group1 0.08(0.05–0.23) p < 0.05 vs Group2, p < 0.01 vs Group1 CRRT: OR 2.57 and 3.28 (p < 0.001) 27.2 % 32.4 % (ns) 15.4 % (ns) PCT [ng/ml]:OR 1.01,ns 3.20 (1.18–10.42) 1.2 (0.43–8.17) (ns) 7.2 (1.22–7.49) (ns) Cardioversion rate 73.5 % 86.1 % (ns) 92.3 % (ns) ICU mortality 40 % (OR 1.58 vs Group2, ns) 30.3 % (ns) 23.1 % (ns) 28-day mortality 49 % (OR 1.72 vs Group2, ns) 39.7 % (ns) 23.1 % (ns)

nd 3.28 (p < 0.001) 27.2 % 32.4 % (ns) 15.4 % (ns) PCT [ng/ml]:OR 1.01,ns 3.20 (1.18–10.42) 1.2 (0.43–8.17) (ns) 7.2 (1.22–7.49) (ns) Cardioversion rate 73.5 % 86.1 % (ns) 92.3 % (ns) ICU mortality 40 % (OR 1.58 vs Group2, ns) 30.3 % (ns) 23.1 % (ns) 28-day mortality 49 % (OR 1.72 vs Group2, ns) 39.7 % (ns) 23.1 % (ns) A902 Comparison of heart rate control with intravenous ivabradine or esmolol perfusion in a large animal model of septic shock A. Herpain1,2, F. Su2, E. Post2, F. Taccone1,2, J.-L. Vincent1,2, J. Creteur1,2 1Erasme University Hospital, Université Libre de Bruxelles, Intensive Care Unit, Bruxelles, Belgium; 2Experimental Laboratory of Intensive Care, Université Libre de Bruxelles, Bruxelles, Belgium Correspondence: A. Herpain – Erasme University Hospital, Université Libre de Bruxelles, Intensive Care Unit, Bruxelles, Belgium Introduction: In septic shock, tachycardia and hyperdynamic hemodynamic status have been associated with a worse outcome in clinical studies; presumably due to an impairment of myocardial oxygenation and ventricular filling. A randomised control trial of heart rate (HR) control in septic shock showed an increase of survival for the patients receiving esmolol1. An animal study observed a similar improvement of survival and an increase in left ventricular (LV) contractility when esmolol was associated with norepinephrine (NE)2. However beta-blockers therapy in sepsis is still debated considering its negative inotropic side effect. Ivabradine, a pure bradycardic agent, blocking selectively the If channels in the sinus node, could represent a safer option for HR control.

(LV) contractility when esmolol was associated with norepinephrine (NE)2. However beta-blockers therapy in sepsis is still debated considering its negative inotropic side effect. Ivabradine, a pure bradycardic agent, blocking selectively the If channels in the sinus node, could represent a safer option for HR control. Objectives: Compare the hemodynamic tolerance of HR control either with intravenous (IV) ivabradine or esmolol perfusion, in a large animal model of septic shock. Methods: We used a closed chest swine model of fecal peritonitis. Analgesia and sedation were provided by sufentanil and sevoflurane. Hemodynamic monitoring included arterial blood pressure (ABP); continuous cardiac output (CCO); LV maximum rate of pressure (dP/dTmax) and LV elastance (E-LV); mixed venous oxygen saturation (SVO2) and arterial lactate (Lac). After the development of septic shock, fluid resuscitation was started and animals were randomised in 3 groups of 6 pigs: ivabradine (IVB), esmolol (ESM) or control. Ivabradine was administered with an IV bolus of 0,1 mg/kg that could be repeated at 0,3 mg/kg, aiming an HR between 80 and 90 beats per minute (BPM). Continuous IV perfusion of esmolol was started at 1 mg/kg/h and adapted to reach the same HR range. After 5 hours of HR control, a fixed dose of 0,3 mcg/kg/min NE was introduced in all groups.

tered with an IV bolus of 0,1 mg/kg that could be repeated at 0,3 mg/kg, aiming an HR between 80 and 90 beats per minute (BPM). Continuous IV perfusion of esmolol was started at 1 mg/kg/h and adapted to reach the same HR range. After 5 hours of HR control, a fixed dose of 0,3 mcg/kg/min NE was introduced in all groups. Results: All animals developed an hyperdynamic distributive shock, including tachycardia above 110 bpm. HR control between 80 and 90 bpm was successful in both IVB and ESM groups. IVB administration didn't affect ABP, CCO, dP/dTmax, E-LV, SVO2 or Lac. ESM perfusion tended to decrease ABP, CCO and SVO2; E-LV and Lac were unaffected but dP/dTmax decreased markedly. Under NE perfusion, E-LV was similar in all groups but dP/dTmax was lower in ESM group. Conclusions: In septic shock, HR control with an IV administration of ivabradine doesn´t alter global organs perfusion and cardiac function. Esmolol perfusion, in order to achieve the same goal, reduces LV dP/dTmax and didn´t enhance LV contractility in association with NE. References: 1. Morelli A. Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock: a randomized clinical trial. JAMA. 2013; 310(16):1683–91. 2. Kimmoun A. Beta1-Adrenergic Inhibition Improves Cardiac and Vascular Function in Experimental Septic Shock. Crit Care Med. 2015; 43(9):e332-40. Grant acknowledgment Fonds Erasme pour la Recherche Scientifique. Fonds pour la Chirurgie Cardiaque.

1. Morelli A. Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock: a randomized clinical trial. JAMA. 2013; 310(16):1683–91. 2. Kimmoun A. Beta1-Adrenergic Inhibition Improves Cardiac and Vascular Function in Experimental Septic Shock. Crit Care Med. 2015; 43(9):e332-40. Grant acknowledgment Fonds Erasme pour la Recherche Scientifique. Fonds pour la Chirurgie Cardiaque. A903 Hypotension probability algorithm accuracy on MIMIC II ICU patients C. Lee1,2, F. Hatib1, Z. Jian1, S. Buddi1, M. Cannesson3 1Edwards Lifesciences, Critical Care, Irvine, United States; 2University of California Irvine, Biomedical Engineering, Irvine, United States; 3University of California Los Angeles, Anesthesiology, Los Angeles, United States Correspondence: C. Lee – Edwards Lifesciences, Critical Care, Irvine, United States Introduction: Patients in critical care settings are often at risk of developing hypotension, which can lead to poor outcomes such as increased morbidity and mortality. Current hemodynamic parameters for monitoring such hypotension often exhibit pronounced changes only when the hypotensive event is already occurring or when it is too late. We have developed a hypotension probability indicator (HPI™) to predict hypotensive episodes based on machine learning techniques. The HPI™ model was trained on ~3000 ICU and OR patients. The objective of this study is two-fold: 1) To test the accuracy of HPI™ to predict events on a completely independent test data set of ICU patients, not used in the development of the algorithm; and 2) To compare timing of interventions in response to an event to the timing of detection of an event by HPI™.

patients. The objective of this study is two-fold: 1) To test the accuracy of HPI™ to predict events on a completely independent test data set of ICU patients, not used in the development of the algorithm; and 2) To compare timing of interventions in response to an event to the timing of detection of an event by HPI™. Methods: Data used in this study came from the MIMIC II MIT Research Database. Arterial pressure waveforms of 326 patients were analyzed for HPI™ and then tested for event detection and prediction accuracy. All features of the HPI™ as well as other hemodynamic parameters for comparison were calculated using FloTrac (Edwards Lifesciences, Irvine, CA). A hypotensive event was defined as any time period where MAP < 65 mmHg for at least 1 minute. An ROC analysis was performed to assess AUC, sensitivity, and specificity of the HPI™ to identify an event during the event, and 5, 10, and 15 minutes prior to the start of event. Next, clinical records of the 326 patients were reviewed for any drug or fluid interventions during start of event to 5 minutes after an event and the elapsed time from start of event to intervention time was calculated. A drug or fluid intervention was defined as any bolus or IV infusion start. In addition, the time at which HPI™ probability of event > 0.85 prior to the start of an event was also calculated for comparison. Data are presented in median [25- 75th percentiles].

ime from start of event to intervention time was calculated. A drug or fluid intervention was defined as any bolus or IV infusion start. In addition, the time at which HPI™ probability of event > 0.85 prior to the start of an event was also calculated for comparison. Data are presented in median [25- 75th percentiles]. Results: Patient demographics are presented in Fig. 34. Overall, hypotensive events had a duration of 2.0 [0.7-6.3] minutes. HPI™ can accurately detect an event up to 15 minutes prior to the start of an event with high sensitivity and specificity (>0.85) (Fig. 35). In addition, out of 26,021 total number of events, 15.4 % of events had a drug or fluid intervention start sometime between the start of an event and 5 minutes after the end of an event. Of these interventions, the median start time of intervention was 4.7 [2.3-10.2] minutes after the start of an event. In comparison, HPI™ detected 99.99 % of events and detected at 17.3 [4.7-53.7] minutes before the start of an event.

start sometime between the start of an event and 5 minutes after the end of an event. Of these interventions, the median start time of intervention was 4.7 [2.3-10.2] minutes after the start of an event. In comparison, HPI™ detected 99.99 % of events and detected at 17.3 [4.7-53.7] minutes before the start of an event. Conclusion: In conclusion, HPI™ can accurately detect an event up to 15 minutes prior. HPI™ may serve as a useful addition in the care of critically ill patients by potentially facilitating earlier intervention either in response to an event or serve as a decision support and direct a physician's attention to potential oncoming events when HPI™ is high.Fig. 34 (abstract A903). Patient demographics (n = 236). Data represented as #(% of total patients) OR mean ± standard deviation. SAPSI and SOFA scores were calculated by the MIMIC II MIT Database. CCU = Coronary Care Unit; CSRU = Cardiac Surgery Recovery Unit; FICU = Finard Medical Surgical ICU; SICU = Surgical ICU; MICU = Medical ICU. It should also be noted that 60 patients had empty ICD9 records. Fig. 35 (abstract A903). AUC, Sensitivity, and Specificty Results for the hypotension probability indicator (HPITM) on MIMIC II ICU Patients. Results were calculated for the HPI‘sTM ability to correctly detect a hypotensive event 15 minutes prior to the start of a defined event, 10 minutes, 5 minutes, and 0 minutes

Conclusion: In conclusion, HPI™ can accurately detect an event up to 15 minutes prior. HPI™ may serve as a useful addition in the care of critically ill patients by potentially facilitating earlier intervention either in response to an event or serve as a decision support and direct a physician's attention to potential oncoming events when HPI™ is high.Fig. 34 (abstract A903). Patient demographics (n = 236). Data represented as #(% of total patients) OR mean ± standard deviation. SAPSI and SOFA scores were calculated by the MIMIC II MIT Database. CCU = Coronary Care Unit; CSRU = Cardiac Surgery Recovery Unit; FICU = Finard Medical Surgical ICU; SICU = Surgical ICU; MICU = Medical ICU. It should also be noted that 60 patients had empty ICD9 records. Fig. 35 (abstract A903). AUC, Sensitivity, and Specificty Results for the hypotension probability indicator (HPITM) on MIMIC II ICU Patients. Results were calculated for the HPI‘sTM ability to correctly detect a hypotensive event 15 minutes prior to the start of a defined event, 10 minutes, 5 minutes, and 0 minutes Post-Cardiac Arrest Care A904 An estimation of the number of successful resuscitations after in hospital cardiac arrest in one single country and possibilities for improvement F. Hessulf1, P. Lundgren2, J. Herlitz2,3, J. Engdahl2,4 1Halland Hospital Halmstad, Department of Intensive Care Medicine, Halmstad, Sweden; 2Sahlgrenska University Hospital, Inst of Internal Medicine, Department of Molecular and Clinical Medicine, Gothenburg, Sweden; 3University of Borås, The Center for Prehospital Care in Western Sweden, Borås, Sweden; 4Halland Hospital Halmstad, Department of Internal Medicine, Halmstad, Sweden Correspondence: F. Hessulf – Halland Hospital Halmstad, Department of Intensive Care Medicine, Halmstad, Sweden Introduction: Despite the development of the chain of survival, the majority of patients suffering a cardiac arrest (CA) do not survive. This statement is valid for both in-hospital as well as out-of-hospital cardiac arrest. Regardless of the location of the cardiac arrest, there are at least four factors that appear to be of major importance for survival. The first is the time from collapse to delivery of treatment; the second is the quality of cardiopulmonary resuscitation (CPR); the third is the patient's co-morbidity and the fourth is the aetiology of the CA and the presenting rhythm. The present study will focus on the first three parts of the chain of survival, time from collapse to call/CPR/defibrillation.

very of treatment; the second is the quality of cardiopulmonary resuscitation (CPR); the third is the patient's co-morbidity and the fourth is the aetiology of the CA and the presenting rhythm. The present study will focus on the first three parts of the chain of survival, time from collapse to call/CPR/defibrillation. Objectives: To describe the number of survivors following in-hospital cardiac arrest (IHCA) in Sweden during one year and, based on estimations and assumptions, calculate the potential number of additional lives saved following improvements in the chain of survival. Methods: A retrospective register study based on the Swedish Register of Cardiopulmonary resuscitation 2005–2014. The inclusion criterions were: a confirmed in-hospital cardiac arrest where resuscitation was initiated among patiens aged >18 years. Based on the chance of survival in relation to the delay from collapse to call for the rescue team and start of treatment, we estimated the potential number of lives that could be saved if the delay to call and treatment was reduced to the treatment recommendations by the Swedish Resuscitation Council.

aged >18 years. Based on the chance of survival in relation to the delay from collapse to call for the rescue team and start of treatment, we estimated the potential number of lives that could be saved if the delay to call and treatment was reduced to the treatment recommendations by the Swedish Resuscitation Council. Results: We estimate that in 2014 resuscitation was initiated on 2850 adult patients (>18 years) suffering IHCA. This corresponds to 14 IHCA cases per 100 hospital beds and year. 30-day survival was 31 % i.e. 900 survivors (four survivors per 100 hospital beds). There was a strong inverse relation between delay to call for the rescue team and delay to treatment and survival. If delay from collapse to a/call and, b/start of CPR were reduced to < 1 minute in patients with a longer delay than that and if c/time from collapse to defibrillation was reduced to < 3 minutes among those with a longer delay than that: a/57; b/32; and c/35 further lives could potentially be saved. We speculate that about 100 additional lives (one per 200 hospital beds each year) could theoretically be saved by improved adherence to guidelines regarding the first three components in the chain of survival in Swedish hospitals yearly.

r delay than that: a/57; b/32; and c/35 further lives could potentially be saved. We speculate that about 100 additional lives (one per 200 hospital beds each year) could theoretically be saved by improved adherence to guidelines regarding the first three components in the chain of survival in Swedish hospitals yearly. Conclusions: In 2014, approximately 900 patients (four per 100 hospital beds) were successfully resuscitated following IHCA in Sweden. There was a strong negative relation between collapse and call for rescue team/CPR/defibrillation and 30-day survival. With reduced delay times a further 100 lives (one per 200 hospital beds) could theoretically be saved each year in Sweden. Grant acknowledgment The study was supported by grants from the Laerdal Foundation of Acute Medicine in Norway (JH) and the Scientific Council of Halland (FH).

Conclusions: In 2014, approximately 900 patients (four per 100 hospital beds) were successfully resuscitated following IHCA in Sweden. There was a strong negative relation between collapse and call for rescue team/CPR/defibrillation and 30-day survival. With reduced delay times a further 100 lives (one per 200 hospital beds) could theoretically be saved each year in Sweden. Grant acknowledgment The study was supported by grants from the Laerdal Foundation of Acute Medicine in Norway (JH) and the Scientific Council of Halland (FH). A905 Prophylactic versus clinically-driven antibiotics in comatose survivors of out-of-hospital cardiac arrest - a pilot study S. Fileković1, M. Turel2, R. Knafelj3, V. Gorjup3, R. Stanić1, P. Gradišek1, O. Cerović1, T. Mirković1, M. Noč3 1University Medical Centre Ljubljana, Department of Anaesthesiology and Surgical Intensive Therapy, Ljubljana, Slovenia; 2University Medical Center Ljubljana, Department of Pulmonary Diseases and Allergy, Ljubljana, Slovenia; 3University Medical Center Ljubljana, Center for Intensive Internal Medicin, Ljubljana, Slovenia Correspondence: S. Fileković – University Medical Centre Ljubljana, Department of Anaesthesiology and Surgical Intensive Therapy, Ljubljana, Slovenia Introduction: Benefit of prophylactic antibiotics, which may suppress development of postresuscitation infection and especially early onset pneumonia and thereby decrease the severity of postresuscitation systemic inflammatory response, is controversial.

aesthesiology and Surgical Intensive Therapy, Ljubljana, Slovenia Introduction: Benefit of prophylactic antibiotics, which may suppress development of postresuscitation infection and especially early onset pneumonia and thereby decrease the severity of postresuscitation systemic inflammatory response, is controversial. Objectives: We investigated potential benefits of prophylactic antibiotic treatment in comatose survivors of out-of-hospital cardiac arrest (OHCA). Methods: Among 83 patients undergoing admission bronchoscopy from September 2013 to February 2015, 60 patients without tracheobronchial aspiration were randomized to prophylactic (Amoxicillin-Clavulanic acid 1.2 g every 8 h) or clinically-driven antibiotics administered if signs of infection developed.

Objectives: We investigated potential benefits of prophylactic antibiotic treatment in comatose survivors of out-of-hospital cardiac arrest (OHCA). Methods: Among 83 patients undergoing admission bronchoscopy from September 2013 to February 2015, 60 patients without tracheobronchial aspiration were randomized to prophylactic (Amoxicillin-Clavulanic acid 1.2 g every 8 h) or clinically-driven antibiotics administered if signs of infection developed. Results: Proportion of patients on antibiotics was significantly greater from day 1 to 5 in prophylactic group while there was no difference on days 6 to 7. Peak C-reactive protein in prophylactic group was significantly smaller (186 ± 61 vs. 229 ± 60 mg/L; p = 0.04). There was no difference in peak white blood cell count (14.6 ± 6.6 vs. 16.6 ± 6.2; p = 0.24), procalcitonin (4.02 ± 10.12 vs. 4.84 ± 8.5 microg/L; p = 0.80) and CD 64. Except for positive mini BAL on day 3 (7 % vs. 42 %; p < 0.01), there was no significant impact on other microbiological samples and X-ray signs of pneumonia (50 % in each group). Use vasopressors/inotropes (93 % in each groups), duration of mechanical ventilation (5.4 ± 3.7 vs. 5.2 ± 3.1 days), tracheal intubation (6.5 ± 4.6 vs. 5.9 ± 4.3 days), ICU stay (7.7 ± 5.2 vs. 6.9 ± 4.5 days), survival (73 % vs. 73 %) and survival with good neurological outcome (50 % vs. 40 %) were also comparable.

p). Use vasopressors/inotropes (93 % in each groups), duration of mechanical ventilation (5.4 ± 3.7 vs. 5.2 ± 3.1 days), tracheal intubation (6.5 ± 4.6 vs. 5.9 ± 4.3 days), ICU stay (7.7 ± 5.2 vs. 6.9 ± 4.5 days), survival (73 % vs. 73 %) and survival with good neurological outcome (50 % vs. 40 %) were also comparable. Conclusions: Tracheobronchial aspiration was documented in more than a quarter of comatose survivors of OHCA using bronchoscopy on admission. In the absence of aspiration, prophylactic antibiotics reduced peak CRP and the incidence of positive mini-BAL on day 3 and had no significant impact on other microbiological samples, incidence of pneumonia, ICU treatment and outcome. References 1. Tomte O, Andersen GO, Jacobsen D, Draegni T et al. (2011) Strong and weak aspects of an established post-resuscitation treatment protocol - A five year observational study. Resuscitation 82:1186–93 2. Stub D, Hengel C, Chan W, Jackson D, Sanders K et al. (2011) Usefulness of cooling and coronary catheterization to improve survival in out-of-hospital cardiac arrest. Am J Cardiol 107:522–7 3. Nielsen N, Wetterslev J, Cronberg T et al. (2013) Targeted temperature management at 33 C versus 36 C after cardiac arrest. N Engl J Med 369: 2197–206 4. Kocjancic ST, Jazbec A, Noc M (2014) Impact of intensified postresuscitation treatment on outcome of comatose survivors of out-of-hospital cardiac arrest according to initial rhythm. Resuscitation 85:1364–9 Grant acknowledgment

3. Nielsen N, Wetterslev J, Cronberg T et al. (2013) Targeted temperature management at 33 C versus 36 C after cardiac arrest. N Engl J Med 369: 2197–206 4. Kocjancic ST, Jazbec A, Noc M (2014) Impact of intensified postresuscitation treatment on outcome of comatose survivors of out-of-hospital cardiac arrest according to initial rhythm. Resuscitation 85:1364–9 Grant acknowledgment Suada Fileković received educational grant "Innovative scheme to co-finance doctoral studies" number 291–499 by Ministry of higher education, science and technology, Republic of Slovenia. A906 Aetiology of cardiac arrests in hospital general wards J. Tirkkonen1,2, H. Hellevuo3, K.T. Olkkola4, S. Hoppu2 1Seinäjoki Central Hospital, Department of Anaesthesiology and Intensive Care Medicine, Seinäjoki, Finland; 2Tampere University Hospital, Department of Intensive Care Medicine, Tampere, Finland; 3Tampere University Hospital, Department of Emergency Medicine, Tampere, Finland; 4Helsinki University Hospital, Department of Anaesthesiology, Intensive Care, Emergency Care and Pain Medicine, Helsinki, Finland Correspondence: J. Tirkkonen – Seinäjoki Central Hospital, Department of Anaesthesiology and Intensive Care Medicine, Seinäjoki, Finland Introduction: Survival to discharge after in-hospital cardiac arrest (IHCA) is poor (10 − 20 %) and has not improved despite developments in modern medicine.1 Data on the aetiology of in-hospital cardiac arrests is very limited, and conducted studies include IHCA patients resuscitated in emergency departments, intensive care units and high dependency units.

er in-hospital cardiac arrest (IHCA) is poor (10 − 20 %) and has not improved despite developments in modern medicine.1 Data on the aetiology of in-hospital cardiac arrests is very limited, and conducted studies include IHCA patients resuscitated in emergency departments, intensive care units and high dependency units. Objectives: To determine the underlying causes of IHCAs occurring on general wards and investigate, whether the aetiology is independently associated with six months survival. Methods: A prospective observational study between 2009–2011 in a Finnish university hospital. We included all adult IHCA patients on general wards who were attended by ICU´s medical emergency team. Definite aetiology was determined from the autopsy records and medical records. No autopsies were conducted solely for study purposes. The local Ethics Committee approved the study protocol (Approval no: R08116).

al. We included all adult IHCA patients on general wards who were attended by ICU´s medical emergency team. Definite aetiology was determined from the autopsy records and medical records. No autopsies were conducted solely for study purposes. The local Ethics Committee approved the study protocol (Approval no: R08116). Results: The cohort consisted of 279 patients, of which 185 (66 %) were male. Median age of the patients was 72 (64, 80) years. Altogether 178 (64 %) IHCAs were monitored/witnessed, first rhythm was shockable in 42 (15 %) cases and 53 (19 %) patients survived six months. Autopsy was conducted in 153 (55 %) cases. Aetiology was determined as cardiac in 141 events, 73 of which were due to acute myocardial infarction and 26 due to acute myocardial ischaemia without infarction. Congestive heart failure was the third most prevalent reason in cardiac sub cohort (16). Altogether 138 IHCAs were considered non-cardiac; most common causes were pneumonia (39), exsanguination (16), pulmonary embolism (12) and peritonitis (11). Cardiac IHCAs were more commonly preceded by subjective symptoms (e.g. chest pain, respiratory distress) than non-cardiac IHCAs (47 % vs. 32 %, p = 0.022), while objective vital dysfunctions preceded IHCAs as often in both sub cohorts (40 % vs. 44 %, p = 0.448). In a multivariate logistic regression model monitored/witnessed event, shockable primary rhythm and low age-adjusted Charlson comorbidity index score were factors independently associated with 180-day survival, but the aetiology (cardiac vs. non-cardiac) was not.

as often in both sub cohorts (40 % vs. 44 %, p = 0.448). In a multivariate logistic regression model monitored/witnessed event, shockable primary rhythm and low age-adjusted Charlson comorbidity index score were factors independently associated with 180-day survival, but the aetiology (cardiac vs. non-cardiac) was not. Conclusions: Aetiology of IHCAs on general wards is cardiac in 50 % of the events. Ischaemic reasons for IHCAs were twice as common as shockable primary rhythms in this study. Subjective symptoms and objective vital dysfunctions often precede general ward IHCAs. However, neither the aetiology nor the presence of antecedents, but low comorbidity, observed arrest and shockable primary rhythm are factors associated with a favorable outcome. References 1. Nolan J, Soar J, Smith GB, et al. Incidence and outcome of in-hospital cardiac arrest in the United Kingdom National Cardiac Arrest Audit. Resuscitation 2014;85:987–92.

Conclusions: Aetiology of IHCAs on general wards is cardiac in 50 % of the events. Ischaemic reasons for IHCAs were twice as common as shockable primary rhythms in this study. Subjective symptoms and objective vital dysfunctions often precede general ward IHCAs. However, neither the aetiology nor the presence of antecedents, but low comorbidity, observed arrest and shockable primary rhythm are factors associated with a favorable outcome. References 1. Nolan J, Soar J, Smith GB, et al. Incidence and outcome of in-hospital cardiac arrest in the United Kingdom National Cardiac Arrest Audit. Resuscitation 2014;85:987–92. A907 Reducing in-hospital cardiac arrest by implementation of innovative early warning information system in a tertiary medical center K.-C. Lin1, W.-T. Hung1, C.-C. Chiang2, W.-C. Huang1, W.-C. Juan3, S.-C. Lin1, C.-C. Cheng4, P.-H. Lin3, K.-Y. Fong3, D.-S. Hou3, P.-L. Kang4, S.-R. Wann3, Y.-S. Chen5, G.-Y. Mar4, C.-P. Liu2 1Kaohsiung Veterans General Hospital, Department of Critical Care Medicine, Kaohsiung City, Taiwan, Province of China; 2Kaohsiung Veterans General Hospital, Cardiovascular division, Kaohsiung City, Taiwan, Province of China; 3Kaohsiung Veterans General Hospital, Department of Emergency, Kaohsiung City, Taiwan, Province of China; 4Kaohsiung Veterans General Hospital, Cardiovascular Division, Kaohsiung City, Taiwan, Province of China; 5Kaohsiung Veterans General Hospital, Department of Internal Medicine, Kaohsiung City, Taiwan, Province of China Correspondence: W.-T. Hung – Kaohsiung Veterans General Hospital, Department of Critical Care Medicine, Kaohsiung City, Taiwan, Province of China Introduction: In-hospital cardiac arrest (IHCA) is a common and high-risk issue with less than 20 % surviving to hospital discharge. Most patients show signs of clinical deterioration in the hours before IHCA. As a result, the development of vital sign-based early warning system was designed to detect early signs of clinical deterioration before IHCA attack in order to trigger early intensive care.

ith less than 20 % surviving to hospital discharge. Most patients show signs of clinical deterioration in the hours before IHCA. As a result, the development of vital sign-based early warning system was designed to detect early signs of clinical deterioration before IHCA attack in order to trigger early intensive care. Objectives: In this study, we investigate the impact of the implementation of an innovative early warning information system on the rate of IHCA and survival rate in IHCA patients. Methods: A multidisciplinary team among intensivists, cardiologists, emergency physicians, and nursing staffs in a tertiary medical center was organized since May 2015. The key interventions include automatic national early warning score (NEWS) calculating information system, nurses and physicians computer-based reminding alarm if NEWS ≥ 7 or more than highest scores among previous 3 measurements, real time early warning screen saver and electric board, in service education and early warning monitor team. All patients admitted between January 2013 and January 2016 were enrolled. Total 143,450 patients were divided into three groups: pre-interventional group from Jan 2013 to April 2015 (n = 107,437), Interventional group from May to June 2015 (n = 7,923) and post-interventional group from July 2015 to Jan 2016 (n = 28,090). The definition of In-hospital cardiac arrest is the number of In-hospital cardiac arrest per thousand admitted patients. We compared the rates of IHCA, 48 hours survival rate and discharge survival rate in IHCA patients among these 3 groups.

7,923) and post-interventional group from July 2015 to Jan 2016 (n = 28,090). The definition of In-hospital cardiac arrest is the number of In-hospital cardiac arrest per thousand admitted patients. We compared the rates of IHCA, 48 hours survival rate and discharge survival rate in IHCA patients among these 3 groups. Results: The rate of In-hospital cardiac arrest improved from 2.53‰ in pre-interventional group, to 2.15‰ in interventional group and to 1.32‰ in post-interventional group (p < 0.05). The 48 hours survival rate in IHCA patients increased from 34.4 % in pre-interventional group, to 41.2 % in interventional group and to 45.4 % in post-interventional group (p < 0.05). The discharge survival rate in IHCA patients also increased from 15.6 % in pre-interventional group, to 29.4 % in interventional group and to 36.4 % in post-interventional group (p < 0.05). Conclusions: The study demonstrated that implementation of early warning information system and innovative strategies could attenuate the rate of IHCA, 48 hours survival rate and discharge survival rate in IHCA patients. Grant acknowledgment None

Results: The rate of In-hospital cardiac arrest improved from 2.53‰ in pre-interventional group, to 2.15‰ in interventional group and to 1.32‰ in post-interventional group (p < 0.05). The 48 hours survival rate in IHCA patients increased from 34.4 % in pre-interventional group, to 41.2 % in interventional group and to 45.4 % in post-interventional group (p < 0.05). The discharge survival rate in IHCA patients also increased from 15.6 % in pre-interventional group, to 29.4 % in interventional group and to 36.4 % in post-interventional group (p < 0.05). Conclusions: The study demonstrated that implementation of early warning information system and innovative strategies could attenuate the rate of IHCA, 48 hours survival rate and discharge survival rate in IHCA patients. Grant acknowledgment None A908 Delayed awakening after cardiac arrest: prevalence and risk factors in the Parisian registry M. Paul1, W. Bougouin1, G. Geri1, F. Dumas2, B. Champigneulle1, S. Legriel3, J. Charpentier1, J.-P. Mira1, C. Sandroni4, A. Cariou1 1ICU Cochin Hospital, Paris, France; 2Emergency Department, Cochin Hospital, Paris, France; 3ICU Mignot Hospital, Versailles, France; 4Anesthesiology and Intensive Care, Catholic University School of Medicine, Rome, Italy Correspondence: M. Paul – ICU Cochin Hospital, Paris, France Introduction: Although prolonged unconsciousness after cardiac arrest (CA) is a sign of poor neurological outcome, limited evidence shows that a late recovery may occur in a minority of patients.

nsive Care, Catholic University School of Medicine, Rome, Italy Correspondence: M. Paul – ICU Cochin Hospital, Paris, France Introduction: Although prolonged unconsciousness after cardiac arrest (CA) is a sign of poor neurological outcome, limited evidence shows that a late recovery may occur in a minority of patients. Objectives: We investigated the prevalence and the predictive factors of delayed awakening in comatose CA survivors treated with targeted temperature management (TTM). Methods: Retrospective analysis of the Parisian Region Out-of-Hospital CA Registry (2008–2013). In adult comatose CA survivors treated with TTM, sedated with midazolam and fentanyl, time to awakening was measured starting from discontinuation of sedation at the end of rewarming. Awakening was defined as delayed when it occurred after more than 48 h. Results: A total of 326 patients (71 % male, mean age 59 ± 16 years) were included, among whom 194 awoke. Delayed awakening occurred in 56/194 (29 %) patients, at a median time of 93 h (IQR 70–117) from discontinuation of sedation. In 5/56 (9 %) late awakeners, pupillary reflex and motor response were both absent 48 h after sedation discontinuation. In multivariate analysis, age over 59 years (OR 2.1, 95 % CI 1.0-4.3), post-resuscitation shock (OR 2.6 [1.3-5.2]), and renal insufficiency at admission (OR 3.1 [1.4- 6.8]) were associated with significantly higher rates of delayed awakening.

flex and motor response were both absent 48 h after sedation discontinuation. In multivariate analysis, age over 59 years (OR 2.1, 95 % CI 1.0-4.3), post-resuscitation shock (OR 2.6 [1.3-5.2]), and renal insufficiency at admission (OR 3.1 [1.4- 6.8]) were associated with significantly higher rates of delayed awakening. Conclusions: Delayed awakening is common among patients recovering from coma after CA. Renal insufficiency, older age, and post-resuscitation shock were independent predictors of delayed awakening. Presence of unfavorable neurological signs at 48 h after rewarming from TTM and discontinuation of sedation did not rule out recovery of consciousness in late awakeners. Grant acknowledgment None 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. Paul M, Bougoun W, Geri G, Dumas F, Chapigneulle B, Legriet S, Charpentier J, Mira JP, Sandroni C, Cariou A (2016) Delayed awakening after cardiac arrest: prevalence and risk factors in the Parisian registry. Intensive Care Med 42(7) 1128–1136.

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. Paul M, Bougoun W, Geri G, Dumas F, Chapigneulle B, Legriet S, Charpentier J, Mira JP, Sandroni C, Cariou A (2016) Delayed awakening after cardiac arrest: prevalence and risk factors in the Parisian registry. Intensive Care Med 42(7) 1128–1136. Host Response In Sepsis A909 Discovery and validation of a host immune response gene expression signature for viral systemic inflammation J. Zimmerman1, E. Sullivan1, M. Noursadeghi2, B. Fox3, D. Sampson3, L. McHugh3, T. Yager3, S. Cermelli3, T. Seldon3, S. Bhide3, R.A. Brandon3, R.B. Brandon3 1University of Washington School of Medicine, Seattle Children's Hospital, Pediatric Critical Care Medicine, Seattle, United States; 2University College London, Faculty of Medical Sciences, Division of Infection & Immunity, London, United Kingdom; 3Immunexpress, Seattle, United States Correspondence: J. Zimmerman – University of Washington School of Medicine, Seattle Children's Hospital, Pediatric Critical Care Medicine, Seattle, United States Introduction: Viral infections play a key role in preventable deaths of children globally, and can be antecedents to bacterial pneumonia and sepsis. Diagnosis of viral infection is often problematic due to non-specific clinical presentation. We developed a host immune response gene expression signature to distinguish systemic inflammation due to viral infection vs. bacterial or non-infectious causes.

obally, and can be antecedents to bacterial pneumonia and sepsis. Diagnosis of viral infection is often problematic due to non-specific clinical presentation. We developed a host immune response gene expression signature to distinguish systemic inflammation due to viral infection vs. bacterial or non-infectious causes. Objectives: To define and validate the host immune response gene expression signature against multiple independent datasets. Methods: Four public GEO datasets describing transcriptomic responses to viral infection were used to identify biomarkers, ranked by AUC, which could separate affected from unaffected subjects. Biomarkers that also responded (AUC > 0.80) to non-viral causes of systemic inflammation were removed. Remaining biomarkers were then ranked for performance in 10 other GEO transcriptomic datasets for viral infection; those with mean AUC >0.75 were retained. Next, a greedy search was applied to the merged (4 + 10) viral GEO datasets to identify the best combinations of biomarkers for discrimination of viral infection. The signature was then validated using independent datasets.

other GEO transcriptomic datasets for viral infection; those with mean AUC >0.75 were retained. Next, a greedy search was applied to the merged (4 + 10) viral GEO datasets to identify the best combinations of biomarkers for discrimination of viral infection. The signature was then validated using independent datasets. Results: A 4-gene signature (comprised of ISG15, IL16, OASL, ADGRE5) had AUC 0.94 across the merged (4 + 10) viral GEO datasets. This signature was validated in 11 additional GEO datasets covering a wide variety of viral pathogens including a time-course study of respiratory syncytial virus (RSV) in children (Fig 36), and in two independent datasets of our own: adults from the emergency department (Fig 37) and children from intensive care (Fig 38). Comparative performance: AUC = 0.85 (95 % CI 0.72-0.99) for adults (Fig 37, 22 control vs. 14 viral), and AUC = 0.91 (95 % CI 0.80-1.00) for children (Fig 38, 29 control vs. 5 viral), when comparing patients retrospectively diagnosed with either a viral infection or no infection. In summary, the signature performs well in GEO datasets for HIV-1, fluA, RSV, measles, herpes viruses, hepatitis C and E, and Marburg virus; across five different mammalian species; and in human pediatric and adult populations with systemic inflammation.

ectively diagnosed with either a viral infection or no infection. In summary, the signature performs well in GEO datasets for HIV-1, fluA, RSV, measles, herpes viruses, hepatitis C and E, and Marburg virus; across five different mammalian species; and in human pediatric and adult populations with systemic inflammation. Conclusions: We have discovered and validated a 4-gene host-immune response signature for differentiating systemic inflammation driven by viral vs. other causes. Following further validation, the signature may have clinical utility in combination with other host response signatures for diagnosing and managing patients with systemic inflammation. Grant acknowledgment Funded by Immunexpress, Seattle Children's Research Institute, and the UK National Institute for Health Research.Fig. 36 (abstract A909). Performance of viral signature in Geo GSE69609 (RSV infection) Fig. 37 (abstract A909). Performance of viral signature in adults from the emergency department Fig. 38 (abstract A909). Performance of viral signature in children from intensive care

Funded by Immunexpress, Seattle Children's Research Institute, and the UK National Institute for Health Research.Fig. 36 (abstract A909). Performance of viral signature in Geo GSE69609 (RSV infection) Fig. 37 (abstract A909). Performance of viral signature in adults from the emergency department Fig. 38 (abstract A909). Performance of viral signature in children from intensive care A910 Remote ischemic preconditioning does not influence the innate immune response during human endotoxemia J. Zwaag, R. Beunders, P. Pickkers, M. Kox Radboud University Medical Center, Intensive Care Research, Nijmegen, Netherlands Correspondence: J. Zwaag – Radboud University Medical Center, Intensive Care Research, Nijmegen, Netherlands Introduction: Using a tourniquet to temporary cut off blood supply to the arm (Remote Ischemic Preconditioning - RIPC) has been shown to result in myocardial protection and reduced incidence of AKI in patients undergoing cardiac surgery. However, a recently performed large multi-center trial in CABG patients showed no beneficial effects on clinically relevant endpoints [1]. Animal studies have shown an `early window of protection' in the 1–2 hours after RIPC as well as a `late window of protection` 12–24 hours after RIPC. Several mechanisms have been suggested to mediate the protective effects of RIPC, of which attenuation of the immune response is an important candidate, although this has hitherto also only been shown in animal studies [2].

on' in the 1–2 hours after RIPC as well as a `late window of protection` 12–24 hours after RIPC. Several mechanisms have been suggested to mediate the protective effects of RIPC, of which attenuation of the immune response is an important candidate, although this has hitherto also only been shown in animal studies [2]. Objectives: To determine the effect of single and repeated RIPC, thereby investigating both the early and late windows of protection, on the inflammatory response during endotoxemia, a standardized, controlled model of systemic inflammation in humans in vivo. Methods: We performed a randomized controlled study in 30 healthy non-smoking male volunteers. Subjects were assigned to either the single-dose RIPC group, multiple-dose RIPC group, or the control group (n = 10 per group). The single-dose RIPC group received 1 dose of RIPC, consisting of 4 cycles of 5-minute ischemia of the arm followed by 5 minutes of reperfusion just before administration of 2 ng/kg lipopolysaccharide (LPS). The multiple-dose RIPC group received one dose of RIPC per day on the 6 days before the endotoxemia experiment day, and 1 dose just before LPS administration. Results: LPS administration resulted in a typical increase in body temperature, flu-like symptoms, and hemodynamic changes, with no differences between groups. Administration of LPS resulted in a sharp increase in plasma levels of the pro-inflammatory cytokines TNF-α, IL-6, and IL-8 as well as the anti-inflammatory cytokine IL-10. No differences in plasma levels of these cytokines were observed between the different groups (Figure 39).

ges, with no differences between groups. Administration of LPS resulted in a sharp increase in plasma levels of the pro-inflammatory cytokines TNF-α, IL-6, and IL-8 as well as the anti-inflammatory cytokine IL-10. No differences in plasma levels of these cytokines were observed between the different groups (Figure 39). Conclusions: In the present study, we demonstrate that RIPC does not affect the in vivo inflammatory response induced by administration of endotoxin in humans. These results implicate that RIPC does not exert direct anti-inflammatory effects and that the previously observed protective effects are mediated through other mechanisms. Furthermore, the absence of immunomodulatory effects of RIPC in the present study tempers expectations of using RIPC as an immunomodulatory treatment strategy in patients. References 1. Meybohm, P., et al., A Multicenter Trial of Remote Ischemic Preconditioning for Heart Surgery. N Engl J Med, 2015. 373(15): p. 1397–407. 2. Cai, Z.P., et al., Remote ischemic preconditioning confers late protection against myocardial ischemia-reperfusion injury in mice by upregulating interleukin-10. Basic Res Cardiol, 2012. 107(4): p. 277.Fig. 39 (abstract A910). Plasma concentrations of (A) tumor necrosis factor (TNF)α, (B) interleukin (IL)-6, (C) IL-8 and (D) IL-10 during LPS-induced systemic inflammation. Data are expressed as medians with IQR. Differences between groups were evaluated using 2-way ANOVAs on log-transformed data. No significant differences were found. RIPC: Remote Ischemic Preconditioning, SD: Single Dose, MD: Multiple Dose

erleukin (IL)-6, (C) IL-8 and (D) IL-10 during LPS-induced systemic inflammation. Data are expressed as medians with IQR. Differences between groups were evaluated using 2-way ANOVAs on log-transformed data. No significant differences were found. RIPC: Remote Ischemic Preconditioning, SD: Single Dose, MD: Multiple Dose A911 Does stem cell applications in sepsis can promote the regulatory t cell mediated immune response? F. Gul1, M.K. Arslantas1, D. Genc2, N. Zibandah2, L. Topcu1, T. Akkoc2, I. Cinel1 1Marmara University, Anaesthesiology and Reanimation, Istanbul, Turkey; 2Marmara University, Pediatric Immunology, Istanbul, Turkey Correspondence: F. Gul – Marmara University, Anaesthesiology and Reanimation, Istanbul, Turkey Introduction: Sepsis-induced immune alterations are associated with secondary infections and increased risk of death (1). Mesenchymal Stem Cells (MSCs) have been described as a novel therapeutic strategy for the treatment of diseases related to inflammation and tissue injury with their potent modulatory effects on immune system (2). Objectives: In this study, we evaluated the immune-modulatory effects of Human Dental Follicle Mesenchymal Stem Cells (HD-MSCs) on lymphocytes which are isolated from peripheral blood samples of sepsis and septic shock patients.

A911 Does stem cell applications in sepsis can promote the regulatory t cell mediated immune response? F. Gul1, M.K. Arslantas1, D. Genc2, N. Zibandah2, L. Topcu1, T. Akkoc2, I. Cinel1 1Marmara University, Anaesthesiology and Reanimation, Istanbul, Turkey; 2Marmara University, Pediatric Immunology, Istanbul, Turkey Correspondence: F. Gul – Marmara University, Anaesthesiology and Reanimation, Istanbul, Turkey Introduction: Sepsis-induced immune alterations are associated with secondary infections and increased risk of death (1). Mesenchymal Stem Cells (MSCs) have been described as a novel therapeutic strategy for the treatment of diseases related to inflammation and tissue injury with their potent modulatory effects on immune system (2). Objectives: In this study, we evaluated the immune-modulatory effects of Human Dental Follicle Mesenchymal Stem Cells (HD-MSCs) on lymphocytes which are isolated from peripheral blood samples of sepsis and septic shock patients. Methods: According to the International Sepsis Definitions Conference(3), patients divided into two groups as sepsis (Group I, n = 10) and septic shock (Group II, n = 10). Peripheral Blood Mononuclear Cells (PBMCs) were isolated from venous blood samples of Group I, Group II and healthy subjects named as Group III, n = 10. Anti-CD3/CD28 PBMCs were co-cultured with DF-MSCs, IFN-g stimulated DF-MSCs and with no MSCs about 72 hour. CD4 + CD25 + FoxP3+ T cells levels (Treg), lymphocyte proliferation and apoptosis were evaluated with the flow cytometry.

venous blood samples of Group I, Group II and healthy subjects named as Group III, n = 10. Anti-CD3/CD28 PBMCs were co-cultured with DF-MSCs, IFN-g stimulated DF-MSCs and with no MSCs about 72 hour. CD4 + CD25 + FoxP3+ T cells levels (Treg), lymphocyte proliferation and apoptosis were evaluated with the flow cytometry. Results: DF-MSCs and IFN-g induced DF-MSCs cultures significantly supressed proliferation in sepsis group when compare to septic shock group(p < 0,005). Conclusions: MSCs demonstrate their effects on immune system by increasing the number and activity of regulatory T cells (Treg) (4) .In our study, MSCs suppressed lymphocyte proliferation and apoptosis but increased the rate of Treg cells in sepsis co-cultures. This effect was more obvious with IFN -g stimulation. These responses were not seen in septic shock patients´ blood samples and might be explained with anergy. Our findings revealed that DF-MSCs application has immunoregulatory effects in sepsis. This approach opened a new area to work how will MSCs be used to reduce organ dysfunctions and mortality in the clinical practice. References 1) Jesús FBM, David A, Raquel A,Eduardo T. Defining immunological dysfunction in sepsis: A requisite tool for precision medicine. Journal of Infection; 2016.doi:10.1016/j.jinf.2016.01.010 2) 2) Shirley HJM, Jack JH, Claudia CD, et all. Mesenchymal Stem Cells Reduce Inflammation while Enhancing Bacterial Clearance and Improving Survival in Sepsis. American Journal of Respiratory and Critical Care Medicine, 2010; 8:1047–1057.

1) Jesús FBM, David A, Raquel A,Eduardo T. Defining immunological dysfunction in sepsis: A requisite tool for precision medicine. Journal of Infection; 2016.doi:10.1016/j.jinf.2016.01.010 2) 2) Shirley HJM, Jack JH, Claudia CD, et all. Mesenchymal Stem Cells Reduce Inflammation while Enhancing Bacterial Clearance and Improving Survival in Sepsis. American Journal of Respiratory and Critical Care Medicine, 2010; 8:1047–1057. 3) Levy MM, Fink MP,Marshall CJ et al., 2001SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Intensive Care Medicine, 2003; 29:530–538 4) Wang Y ,Zhang A ,Ye Z,et all Bone Marrow Derived Mesenchymal Stem Cells Inhibit Acute Rejection of Rat Liver Allografts in Association With Regulatory T-Cell Expansion. Transplantation, 2009; 41:10, 4352–4356Fig. 40 (abstract A911). DF-MSCs and IFN-g induced DF-MSCs cultures significantly increased Treg ratio in sepsis but this effect was not observed in septic shock group Fig. 41 (abstract A911). Reduction of lymphocyte apoptosis was observed in both sepsis and septic shock groups Fig. 42 (abstract A911). DFSCs and IFN-pre-stimulated DFSCs were significantly reduced CD4+ lymphocyte apoptosis in Sepsis and Septic Shock

4) Wang Y ,Zhang A ,Ye Z,et all Bone Marrow Derived Mesenchymal Stem Cells Inhibit Acute Rejection of Rat Liver Allografts in Association With Regulatory T-Cell Expansion. Transplantation, 2009; 41:10, 4352–4356Fig. 40 (abstract A911). DF-MSCs and IFN-g induced DF-MSCs cultures significantly increased Treg ratio in sepsis but this effect was not observed in septic shock group Fig. 41 (abstract A911). Reduction of lymphocyte apoptosis was observed in both sepsis and septic shock groups Fig. 42 (abstract A911). DFSCs and IFN-pre-stimulated DFSCs were significantly reduced CD4+ lymphocyte apoptosis in Sepsis and Septic Shock A912 Inhibition of mitochondrial complex i by metformin recapitulates the blunted response to exogenous mitochondrial uncoupling seen in sepsis E. Greco, M.P. Lauretta, D.T. Andreis, M. Singer Bloomsbury Institute of Intensive Care Medicine, University College London, London, United Kingdom Correspondence: E. Greco – Bloomsbury Institute of Intensive Care Medicine, University College London, London, United Kingdom Introduction: Inhibition of mitochondrial Complex I is described in human and animal sepsis.1,2 This may be responsible, at least in part, for the decrease in mitochondrial functionality seen in sepsis. We have recently demonstrated that the mitochondrial uncoupling agent, dinitrophenol (DNP) failed to increase body temperature and oxygen consumption (VO2) in septic rats, as was seen in healthy controls. This suggests that uncoupling is active in sepsis and can contribute to fever. We further postulated that the blunted effects of DNP in sepsis may be related in part to upstream mitochondrial inhibition.

P) failed to increase body temperature and oxygen consumption (VO2) in septic rats, as was seen in healthy controls. This suggests that uncoupling is active in sepsis and can contribute to fever. We further postulated that the blunted effects of DNP in sepsis may be related in part to upstream mitochondrial inhibition. Objectives: To determine if complex I inhibition by metformin in healthy rats can prevent the increment in temperature and oxygen consumption (VO2) by DNP, and thus mimic the pattern seen in sepsis. Methods: VO2 was measured in awake, cannulated male Wistar rats (approx 300 g body weight) in metabolic cages (Oxymax, Columbus Instruments). Sepsis was induced with an intraperitoneal injection of faecal slurry at time 0. Sham control animals received no slurry. Fluid resuscitation (10 ml/kg/h crystalloid) was started at 2 hours and continued throughout the whole experiment. Half the septic and sham animals were treated with an IV infusion of metformin (186 mg/kg) between hours 2–6. At 6 and 24 hours, all animals received iv DNP (30 mg/kg). Arterial blood gases, echocardiography and core temperature were measured at times 0, 6 and 8, and 24 and 26 hours (i.e. before and after the two doses of DNP). Mean arterial pressure was recorded continuously. Wilcoxon Rank Sum test was used to compare groups and two-way ANOVA to compare changes in continuous variables from baseline between groups. p values < 0.05 were considered statistically significant.

d 8, and 24 and 26 hours (i.e. before and after the two doses of DNP). Mean arterial pressure was recorded continuously. Wilcoxon Rank Sum test was used to compare groups and two-way ANOVA to compare changes in continuous variables from baseline between groups. p values < 0.05 were considered statistically significant. Results: Pretreatment with metformin completely prevented the increase in temperature and VO2 induced by DNP in sham animals at 6 hours and reflected that seen in non-metformin treated septic rats (Figure 43). The reduction in myocardial contractility (stroke volume and Vmax) seen in the septic animals treated with DNP was prevented by Complex I inhibition at 24 h. Metformin was metabolically well tolerated, with no increase in blood lactate. Conclusions: Inhibiting complex I with metformin prevents the uncoupling effect of DNP in sham animals. This mimics the pattern seen in septic animals and confirms that both Complex I inhibition and pre-existing mitochondrial uncoupling could be active in septic rats. References 1. Brealey D, et al. Lancet. 2002;360:219–23, 2Brealey D, et al. Am J Physiol Regul Integr Comp Physiol. 2004;286:R491-7Fig. 43 (abstract A912). Met = Metformin, VO2 = oxygen consumption

Conclusions: Inhibiting complex I with metformin prevents the uncoupling effect of DNP in sham animals. This mimics the pattern seen in septic animals and confirms that both Complex I inhibition and pre-existing mitochondrial uncoupling could be active in septic rats. References 1. Brealey D, et al. Lancet. 2002;360:219–23, 2Brealey D, et al. Am J Physiol Regul Integr Comp Physiol. 2004;286:R491-7Fig. 43 (abstract A912). Met = Metformin, VO2 = oxygen consumption A913 Behavior and prognostic value of nlrp3 inflammasome in sepsis and septic shock I. Palacios Garcia, M. Cordero, A. Diaz Martin, T. Aldabó Pallás, J. Garnacho Montero, J. Revuelto Rey, L. Roman Malo Hospital Virgen del Rocio, Sevilla, Spain Correspondence: I. Palacios Garcia – Hospital Virgen del Rocio, Sevilla, Spain Objectives: The inflammasome is a multiprotein complex that stimulates cytokines release such as interleukin-1β (IL-1β) and IL-18, involved in the inflammatory response. Our aim is to quantify the state of activation of the inflammasome complex in septic patients, as well as to study possible differences in the cytokines levels in sepsis and septic shock, its temporary evolution, and its prognostic value.

such as interleukin-1β (IL-1β) and IL-18, involved in the inflammatory response. Our aim is to quantify the state of activation of the inflammasome complex in septic patients, as well as to study possible differences in the cytokines levels in sepsis and septic shock, its temporary evolution, and its prognostic value. Methods: Prospective study including patients admitted to the ICU with sepsis or septic shock during 15 months. On days 1, 3 and 7, IL 1-β serum levels and real-time expression of NLRP3 inflammasome (nucleotide- binding oligomerization domain, leucine rich repeat domain containing protein and Pyrin) were determined by Elisa and Real time-PCR respectively. Demographic variables, severity scores on ICU admission (APACHE II and SOFA), sepsis focus and mortality were collected. Statistical analysis: T-Student, Kruskal-Wallis and U-Mann-Whitney test as appropriate. Results: There were included 31 patients (severe sepsis 14 and septic shock 18). Overall mortality was 29 % (9 patients). The levels of IL-1β on day 1 (16.5 ± 2.6 vs 14.1 ± 2.4 pg/mL; p < 0.05) and NLRP3-inflammasome (8.5 ± 1.3 vs 7.5 ± 1 mRNA arbitrary units; p < 0.05) were significantly higher in septic shock patients than in sepsis, with no differences in the following days set (3 and 7). The IL-1β and NLRP3 inflammasome levels decreased significantly on days 3 and 7 compared to first day (p < 0.001), without differences between survivors and deceased patients.

ary units; p < 0.05) were significantly higher in septic shock patients than in sepsis, with no differences in the following days set (3 and 7). The IL-1β and NLRP3 inflammasome levels decreased significantly on days 3 and 7 compared to first day (p < 0.001), without differences between survivors and deceased patients. Conclusions: In septic patients, inflammasome activation complex occurs, with higher levels detected in septic shock. Decreased levels of IL-1β and NLRP3 inflammasome in septic process have been observed during evolution, actually without relation with mortality. Weaning And Noninvasive Ventilation A914 Diaphragm thickening fraction multiplied by rapid shallow breathing index could be the best parameter for extubation A.A. Tanaka Montoya1, A.D.C. Amador Martinez2, L.Y. Delgado Ayala2, E. Monares Zepeda2, J. Franco Granillo2, J. Aguirre Sanchez2, G. Camarena Alejo2, A. Rugerio Cabrera2, A. Pedraza Montenegro2 1ABC Medical Center, Critical Care Unit, Mexico, Mexico; 2ABC Medical Center, Mexico, Mexico Correspondence: A.A. Tanaka Montoya – ABC Medical Center, Critical Care Unit, Mexico, Mexico Introduction: Measuring diaphragm thickening fraction (DTF) multiplied by Rapid shallow breathing index (RSBI), it can be used as a measure to predict the success or failure extubation. Objectives: the usefulness of a new formula DTF*RSBI as a parameter to withdrawal mechanical ventilation.

Weaning And Noninvasive Ventilation A914 Diaphragm thickening fraction multiplied by rapid shallow breathing index could be the best parameter for extubation A.A. Tanaka Montoya1, A.D.C. Amador Martinez2, L.Y. Delgado Ayala2, E. Monares Zepeda2, J. Franco Granillo2, J. Aguirre Sanchez2, G. Camarena Alejo2, A. Rugerio Cabrera2, A. Pedraza Montenegro2 1ABC Medical Center, Critical Care Unit, Mexico, Mexico; 2ABC Medical Center, Mexico, Mexico Correspondence: A.A. Tanaka Montoya – ABC Medical Center, Critical Care Unit, Mexico, Mexico Introduction: Measuring diaphragm thickening fraction (DTF) multiplied by Rapid shallow breathing index (RSBI), it can be used as a measure to predict the success or failure extubation. Objectives: the usefulness of a new formula DTF*RSBI as a parameter to withdrawal mechanical ventilation. Methods: 65 patients with invasive mechanical ventilation were recruited prospectively at the Department of Critical Care Medicine Medical Center ABC in a period of 9 months on August 2015 April 2016. The DTF was measured in the area of apposition of the diaphragm to the chest, using an ultrasound transducer 4 MHz and was performed by physician radiologist. It proved difficult to visualize left hemidiaphragm, the reason RSBI*DTF were not performed on left side. Enrolled patients underwent with following weaning criteria: patient awake without continuous infusion of sedatives, SpO2 ≥ 90 % with FiO2 ≤ 50 % and PEEP ≤ 5 cmH2O, and no need for vasopressors. The percentage change in DTF (the end of inspiration-end expiration between the end of expiration) in pressure support (PS) ventilation mode with ventilatory progression purposes extubation was calculated. PS around 7 cm H2O without PEEP. A successful extubation was defined as spontaneous breathing > 48 hrs without ventilatory support.

ange in DTF (the end of inspiration-end expiration between the end of expiration) in pressure support (PS) ventilation mode with ventilatory progression purposes extubation was calculated. PS around 7 cm H2O without PEEP. A successful extubation was defined as spontaneous breathing > 48 hrs without ventilatory support. Results: We included data obtained from 65 patients, 23(35.4 %) women and 42 (64.6 %) men, mean Body Mass Index (BMI) of 25.83 (SD ± 4.19), The frequency of extubation failure was 21.5 %, and mortality 24.6 %. RSBI with a mean of 57.0 ± 12.3 points. USG measurements were: Right thickness at end expiration (RTEEx) 0.28 ± .05 cm, Right Thickness at end inspiration (RTEIs) 0.21 ± .05 cm, Right Diaphragm Thickening fraction (RDTF) 23.1 ± 10.7 %. Mechanical ventilation time in median of 4 days (IQR 3–6) vs 8.5 (IQR 7 to 11), p < 0.001 and RSBI 68.2 +/− 9.6 vs 53.9 +/− 11.1, p < 0.001. Variables able to discriminate failures and their predictive yield were as follows: RSBI with ROC = 0.82 (CI 95 % 0.70-0.94) with a cutoff point of > =68, sensibility 0.64, specificity 0.84 Positive Likelihood ratio (LR+) =4.09 Negative Likelihood Ratio (LR-) =0.4; DTF with ROC = 0.85 (CI 95 % 0.71-0.99) cutoff of > =31, sensibility 0.71, specificity 0.92 LR+ =9.1, LR- =0.31; DTF*RSBI with ROC = 0.90 (CI95% 0.79 - 0.99) cutoff point of > =19.5, sensibility 0.85, specificity 0.92, LR+ =10.92, LR- =0.15. All comparisons with a p < 0.05.

) =4.09 Negative Likelihood Ratio (LR-) =0.4; DTF with ROC = 0.85 (CI 95 % 0.71-0.99) cutoff of > =31, sensibility 0.71, specificity 0.92 LR+ =9.1, LR- =0.31; DTF*RSBI with ROC = 0.90 (CI95% 0.79 - 0.99) cutoff point of > =19.5, sensibility 0.85, specificity 0.92, LR+ =10.92, LR- =0.15. All comparisons with a p < 0.05. Conclusions: There is no relationship between COPD patients, BMI, age and extubation failure. This new formula combine 2 parameters, DTF*RSBI, as a good parameter for extubation. References 1. Giovanni Ferrari et al.Diaphragm ultrasound as a new index of discontinuation from mechanical ventilation critical ultrasound journal 2014 2. Gayan-Ramirez G.Ventilator-induced diaphragm dysfunction: time for (contr)action! Eur Respir J. 2013;42:12–5 3. Schepens T. et al.The course of diaphragm atrophy in ventilated patients assessed with ultrasound: a longitudinal cohort study Critical Care (2015) 19:422

1. Giovanni Ferrari et al.Diaphragm ultrasound as a new index of discontinuation from mechanical ventilation critical ultrasound journal 2014 2. Gayan-Ramirez G.Ventilator-induced diaphragm dysfunction: time for (contr)action! Eur Respir J. 2013;42:12–5 3. Schepens T. et al.The course of diaphragm atrophy in ventilated patients assessed with ultrasound: a longitudinal cohort study Critical Care (2015) 19:422 A915 Patients extubated without any spontaneous breathing trial. A sub-analysis of the wind study T. Pham1,2,3, G. Beduneau4,5, F. Schortgen6, L. Piquilloud7, E. Zogheib8,9, M. Jonas10, F. Grelon11, I. Runge12, N. Terzi13,14, S. Grangé4, G. Barberet15, P.-G. Guitard16, J.-P. Frat17,18, A. Constan19, J.-M. Chrétien20, J. Mancebo21, A. Mercat22, J.-C.M. Richard23, L. Brochard3, The WIND study group 1Hôpital Tenon, APHP, Medical and Surgical ICU, Paris, France; 2Université Paris Diderot, Sorbonne Paris Cité, UMR 1153, Paris, France; 3University of Toronto Saint Michael's Hospital and Keenan Research Centre, Interdepartmental Division of Critical Care, Toronto, Canada; 4Rouen University Hospital, Medical Intensive Care, Rouen, France; 5Rouen University Hospital, UPRES EA3830 IRIB, Rouen, France; 6CHU Henri Mondor, Medical ICU, Créteil, France; 7University Hospital of Lausanne, Intensive Care and Burn Unit, Lausanne, Switzerland; 8CHU d'Amiens, Cardiothoracic and Vascular ICU, Amiens, France; 9Université Jules Verne, Picardie, INSERM U1088 - CURS, Amiens, France; 10CHU Hotel Dieu, Medical ICU, Nantes, France; 11Hospital Le Mans, Intensive Care Unit, Le Mans, France; 12CHR d'Orleans, Medical ICU, Orleans, France; 13CHU Grenoble Alpes, Medical ICU, Grenoble, France; 14Université Grenoble-Alpes, Inserm U1042, Grenoble, France, 15CHR de Mulhouse, Medical ICU, Mulhouse, France; 16Rouen University Hospital, Surgical Intensive Care, Rouen, France; 17CHU de Poitiers, Medical ICU, Poitiers, France; 18Université de Poitiers, INSERM, CIC- 1402, Équipe 5 ALIVE, Poitiers, France; 19CHU Henri Mondor and ReVA Network, Medical ICU, Créteil, France; 20CHU d'Angers, Clinical Research Institute, Angers, France; 21Hospital de Sant Pau, Servei de Medicina Intensiva, Barcelona, Spain; 22CHU d'Angers, Medical ICU, Angers, France; 23Annecy Genevois General Hospital, Annecy, France Correspondence: G. Beduneau – Rouen University Hospital, Medical Intensive Care, Rouen, France Introduction: According the consensus conference on weaning from mechanical ventilation, intubated patients should pass a spontaneous breathing trial (SBT) to assess their readiness to be extubated.

l Hospital, Annecy, France Correspondence: G. Beduneau – Rouen University Hospital, Medical Intensive Care, Rouen, France Introduction: According the consensus conference on weaning from mechanical ventilation, intubated patients should pass a spontaneous breathing trial (SBT) to assess their readiness to be extubated. Objectives: To characterize patients who are extubated without any SBT and to compare them to patients who had at least 1 SBT during their weaning period. Methods: The prospective multicentre observational WIND (Weaning accordIng New Definition) study was performed from April to August 2013. Ventilation and weaning modalities were daily assessed until discharge in all intubated patients admitted to the participating ICUs. We defined 1) weaning attempt (WA) as a spontaneous breathing trial (SBT) or an extubation (with or without SBT), 2) successful weaning as an extubation without death or invasive mechanical ventilation within 7 days. Variables are presented as mean ± standard deviation, median [interquartile range] or number (percentage). Comparisons were made using Chi2 test, exact Fisher tests, Student t-test or Wilcoxon rank sum test as appropriate. All statistical tests were two-sided and P value ≤ 0.05 were considered significant.

thin 7 days. Variables are presented as mean ± standard deviation, median [interquartile range] or number (percentage). Comparisons were made using Chi2 test, exact Fisher tests, Student t-test or Wilcoxon rank sum test as appropriate. All statistical tests were two-sided and P value ≤ 0.05 were considered significant. Results: Among the 2729 patients included, 2051 patients had at least 1 WA comprising 1669 patients whose first WA was a SBT and 382 who had another type of first WA. These 382 patients with no SBT had a total of 454 WA: 252 (55.6 %) planned extubation without SBT, 124 self-extubations (27.3 %), 30 WA while tracheostomized (6.6 %) and 48 SBT after their first WA (10.6 %). The majority of patients with self-extubation had a successful weaning not requiring reintubation (75.8 %). Almost a quarter (n = 95) of the patients who were extubated without any SBT had a decision of withholding or withdrawing invasive mechanical ventilation, representing 89.7 % (N = 78) of the 87 deceased patients. We then excluded patients with a decision of limitation and patients with a self extubation to compare patients who had a planned extubation with or without SBT as first WA (Table 22). Patients with no SBT were younger, less severe and were more often admitted for unplanned surgery: they had an easier weaning with a lower (but non significative) rate of reintubation, a shorter duration of invasive mechanical ventilation and a shorter length of stay in the ICU.

We then excluded patients with a decision of limitation and patients with a self extubation to compare patients who had a planned extubation with or without SBT as first WA (Table 22). Patients with no SBT were younger, less severe and were more often admitted for unplanned surgery: they had an easier weaning with a lower (but non significative) rate of reintubation, a shorter duration of invasive mechanical ventilation and a shorter length of stay in the ICU. Conclusion: Patients who are extubated without SBT seem to belong to three different groups: self-extubation, terminal extubation and patients in whom physicians anticipate an uneventful weaning and extubation. Among the patients with a planned extubation and without any limitation decision, clinical judgment regarding weanability appears to be effective as this group of patients had a good outcome with a low reintubation rate. Grant acknowledgment This study benefited of a grant of the non-profit Association Départementale des Insuffisants Respiratoires (ADIR) of the Haute Normandie, FranceTable 22 (abstract A915). Mean ± SD, median[IQR]

Conclusion: Patients who are extubated without SBT seem to belong to three different groups: self-extubation, terminal extubation and patients in whom physicians anticipate an uneventful weaning and extubation. Among the patients with a planned extubation and without any limitation decision, clinical judgment regarding weanability appears to be effective as this group of patients had a good outcome with a low reintubation rate. Grant acknowledgment This study benefited of a grant of the non-profit Association Départementale des Insuffisants Respiratoires (ADIR) of the Haute Normandie, FranceTable 22 (abstract A915). Mean ± SD, median[IQR] Patient with no limitation and no self extubation N = 1666 Patients with planned extubation without SBT N = 177 Patients with planned extubation after a SBT N = 1489 p-value Age, y 60 ± 16 53 ± 16 61 ± 16 <0.001 SAPS II at admission, points 45 ± 17 41 ± 16 46 ± 17 <0.001 SOFA at admission, points 6.7 ± 3.5 5.9 ± 3.2 6.7 ± 3.5 <0.001 Admission: Medical/Planned/Unplanned surgery, n (%) 1163 (69.8)/247 (14.8)/256 (15.4) 107 (60.5)/27 (15.3)/43 (24.3) 1056 (70.9)/220 (14.8)/213 (9.9) 0.004/0.86/<0.001 Reintubations, n(%) 159 (9.5) 11 (6.3) 148 (11.7) 0.11 Total number of days of invasive MV, days 4 [2;8] 2 [1;6] 4 [2;9] <0.001 Delay from intubation to 1st WA, days 3 [2;6] 2 [1;5] 3 [2;6] <0.001 Length of stay in the ICU, days 7 [4;13] 4 [3;12] 7 [4;13] <0.001 Death, n (%) 46 (2.8) 5 (2.8) 41 (2.8) 0.85

4/0.86/<0.001 Reintubations, n(%) 159 (9.5) 11 (6.3) 148 (11.7) 0.11 Total number of days of invasive MV, days 4 [2;8] 2 [1;6] 4 [2;9] <0.001 Delay from intubation to 1st WA, days 3 [2;6] 2 [1;5] 3 [2;6] <0.001 Length of stay in the ICU, days 7 [4;13] 4 [3;12] 7 [4;13] <0.001 Death, n (%) 46 (2.8) 5 (2.8) 41 (2.8) 0.85 A916 Maximal relaxation rate (MRR) of the diaphragm: could diaphragmatic sonograpy be a reliable alternative to transdiaphragmatic pressure (PDI) measurements? E. Soilemezi, E. Koco, S. Savvidou, C. Nouris, D. Matamis, Plug Working Group 'Papageorgiou' General Hospital, ICU, Thessaloniki, Greece Correspondence: E. Soilemezi – 'Papageorgiou' General Hospital, ICU, Thessaloniki, Greece Introduction: Decrease in diaphragmatic maximal relaxation rate (MRR) occurs early in the process of diaphragmatic fatigue and well before the diaphragm fails as a force generator; its measurement would, therefore, be especially valuable in ICU patients during a weaning trial. However, the use of oesophageal pressure catheters for that purpose impedes wide clinical use. On the contrary, M-mode sonography, allows non-invasive, real-time measurement of the speed of the diaphragmatic motion. Objective: Purpose of our study was to investigate a possible correlation between diaphragmatic MRR traditionally acquired with transdiaphragmatic pressure (Pdi) catheters (MRR-Pdi) and an ECHO equivalent MRR (MRR-ECHO) acquired during different breathing conditions.

A916 Maximal relaxation rate (MRR) of the diaphragm: could diaphragmatic sonograpy be a reliable alternative to transdiaphragmatic pressure (PDI) measurements? E. Soilemezi, E. Koco, S. Savvidou, C. Nouris, D. Matamis, Plug Working Group 'Papageorgiou' General Hospital, ICU, Thessaloniki, Greece Correspondence: E. Soilemezi – 'Papageorgiou' General Hospital, ICU, Thessaloniki, Greece Introduction: Decrease in diaphragmatic maximal relaxation rate (MRR) occurs early in the process of diaphragmatic fatigue and well before the diaphragm fails as a force generator; its measurement would, therefore, be especially valuable in ICU patients during a weaning trial. However, the use of oesophageal pressure catheters for that purpose impedes wide clinical use. On the contrary, M-mode sonography, allows non-invasive, real-time measurement of the speed of the diaphragmatic motion. Objective: Purpose of our study was to investigate a possible correlation between diaphragmatic MRR traditionally acquired with transdiaphragmatic pressure (Pdi) catheters (MRR-Pdi) and an ECHO equivalent MRR (MRR-ECHO) acquired during different breathing conditions. Methods: The slope of MRR was measured from the initial steepest part of the descending Pdi curve simultaneously with the slope of the initial steepest descending part of diaphragmatic excursion with M-mode sonography. The protocol entrained four consecutive stages: i) breathing spontaneously during T-piece trial, ii) breathing spontaneously with performance of sniff-like maneuvers, iii) breathing with resistances of 40cmH2O/L, and iv) breathing with resistances of 40cmH2O/L with performance of sniff-like maneuvers. Statistical comparisons between slope recordings from the two methods were performed with Pearson correlation, while Bland and Altman plots were obtained in order to demonstrate reliable agreement between methods at each different breathing condition.

h resistances of 40cmH2O/L with performance of sniff-like maneuvers. Statistical comparisons between slope recordings from the two methods were performed with Pearson correlation, while Bland and Altman plots were obtained in order to demonstrate reliable agreement between methods at each different breathing condition. Results: A total of 512 separate breaths during the four previously reported breathing conditions from six ICU patients were recorded. Table 23 summarizes the slopes measured from MRR-Pdi and MRR-ECHO as means ± standard deviations (SD), and their linear correlations with p values. Statistical significant correlations were observed in all four stages; i) Pearson correlation coefficient r = 0.474, p < 0.001, R2 = 0.2247, ii) r = 0.834, p < 0.001, R2 = 0.6950, iii) r = 0.653, p < 0.001, R2 = 0.4269, and iv) r = 0.794, p < 0.001, R2 = 0.6302. Bland and Altman plots demonstrating differences of measurements against means, as well as confidence intervals (means of differences ± 2SD) were obtained for each breathing condition. Graph 1 represents the Bland and Altman plot for spontaneous breathing with sniff-like maneuvers without resistances. High R2 indexes, indicating high agreement between the two methods were noted: i) 0.9915, ii) 0.9986, iii) 0.9941, and iv) 0.9990.

Statistical significant correlations were observed in all four stages; i) Pearson correlation coefficient r = 0.474, p < 0.001, R2 = 0.2247, ii) r = 0.834, p < 0.001, R2 = 0.6950, iii) r = 0.653, p < 0.001, R2 = 0.4269, and iv) r = 0.794, p < 0.001, R2 = 0.6302. Bland and Altman plots demonstrating differences of measurements against means, as well as confidence intervals (means of differences ± 2SD) were obtained for each breathing condition. Graph 1 represents the Bland and Altman plot for spontaneous breathing with sniff-like maneuvers without resistances. High R2 indexes, indicating high agreement between the two methods were noted: i) 0.9915, ii) 0.9986, iii) 0.9941, and iv) 0.9990. Conclusions: The results of our study suggest a statistical significant correlation and reliability between diaphragmatic MRR measured from Pdi tracings and the assumed diaphragmatic relaxation rate calculated from simultaneous M-mode sonographic recordings. Clinical studies are required to confirm the potential of this non-invasive index of diaphragmatic MRR to be used as a predictor for weaning success. Grant acknolwedgement None declared.Table 23 (abstract A916). ᅟ

Conclusions: The results of our study suggest a statistical significant correlation and reliability between diaphragmatic MRR measured from Pdi tracings and the assumed diaphragmatic relaxation rate calculated from simultaneous M-mode sonographic recordings. Clinical studies are required to confirm the potential of this non-invasive index of diaphragmatic MRR to be used as a predictor for weaning success. Grant acknolwedgement None declared.Table 23 (abstract A916). ᅟ Slope MRR-Pdi Slope MRR-ECHO Linear correlations p values (Y axis, in cmH2O/sec) (X axis, in cm/sec) i)Spontaneous breathing 77.06 ± 15.83 3.03 ± 0.97 Y = 45 + 9X <0.0001 ii)Sniff-like maneuver 140.48 ± 48.15 3.77 ± 1.58 Y = 45 + 25X <0.0001 iii)Resistances 167.49 ± 41.49 4.89 ± 2.21 Y = 110 + 12X <0.0001 iv)Sniff with resistances 221.64 ± 60.29 5.67 ± 1.62 Y = 53 + 30X <0.0001 Fig. 44 (abstract A916). Spontaneous breathing with sniff-like maneuvers A917 Effects of high flow nasal cannula oxygen on diaphragmatic electrical activity in the post extubation period R. Di Mussi1, S. Spadaro2, C.A. Volta2, M. Mariani1, A. Colaprico1, C. Antonio1, F. Bruno1, S. Grasso1 1Dipartimento dell'Emergenza e dei Trapianti d'Organo, Università degli Studi “Aldo Moro”, Bari, Italy; 2Ferrara University, Ferrara, Italy Correspondence: R. Di Mussi – Dipartimento dell'Emergenza e dei Trapianti d'Organo, Università degli Studi “Aldo Moro”, Bari, Italy Introduction: High flow nasal cannula oxygen therapy (HF OXY) has been recently shown to decrease re-intubation rate, as compared with low flow oxygen therapy (LF OXY).[1], [2]

rara, Italy Correspondence: R. Di Mussi – Dipartimento dell'Emergenza e dei Trapianti d'Organo, Università degli Studi “Aldo Moro”, Bari, Italy Introduction: High flow nasal cannula oxygen therapy (HF OXY) has been recently shown to decrease re-intubation rate, as compared with low flow oxygen therapy (LF OXY).[1], [2] Objectives: To assess the effects of HF OXY as compared with LF OXY on diaphragmatic electrical activity (Eadi), respiratory rate (RR), tidal volume (VT) and gas exchange in the post extubation period. Our hypothesis was that HF OXY, as compared with LF OXY, would improve gas exchange and decrease EAdi. Methods: 10 patients underwent a crossover study immediately after extubation. Each patient was submitted to three consecutive steps of 1 hour each, according to an ON-OFF design: 1) HF OXY; 2) LF OXY; 3) HF OXY. Oxygen fraction was maintained stable throughout the study. The Eadi was continuously monitored through Eadi cathether (Maquet, Solna Sweden). The heated and humidified HF OXY was delivered through nasal cannula at flow rates of 50–70 L/min, (F&P, Auckland New Zealand). Results: RR remained similar throughout the study, VT was significantly higher during the LF OXY step as compared with the HF OXY steps. Oxygenation significantly improved during the HF period, whereas PaCO2 remained unchanged throughout the study (Table 24). EAdi was significantly higher during LF OXY (Figure 45)

Methods: 10 patients underwent a crossover study immediately after extubation. Each patient was submitted to three consecutive steps of 1 hour each, according to an ON-OFF design: 1) HF OXY; 2) LF OXY; 3) HF OXY. Oxygen fraction was maintained stable throughout the study. The Eadi was continuously monitored through Eadi cathether (Maquet, Solna Sweden). The heated and humidified HF OXY was delivered through nasal cannula at flow rates of 50–70 L/min, (F&P, Auckland New Zealand). Results: RR remained similar throughout the study, VT was significantly higher during the LF OXY step as compared with the HF OXY steps. Oxygenation significantly improved during the HF period, whereas PaCO2 remained unchanged throughout the study (Table 24). EAdi was significantly higher during LF OXY (Figure 45) Conclusions: Since the EAdi is correlated to work of breathing, our physiological data suggest that HF OXY significantly reduces WOB while improving oxygenation in the post extubation period. Further studies are required to define if diaphragm unloading may explain the favourable results of HF OXY in clinical trials. References 1. Hernández G, Vaquero C, González P, Subira C, Frutos-Vivar F, Rialp G, Laborda C, Colinas L, Cuena R,Fernández R. - Effect of Postextubation High-Flow Nasal Cannula vs Conventional Oxygen Therapyon Reintubation in Low-Risk Patients: A Randomized Clinical Trial.- JAMA 2016 Apr 5;315(13):1354–61.

Conclusions: Since the EAdi is correlated to work of breathing, our physiological data suggest that HF OXY significantly reduces WOB while improving oxygenation in the post extubation period. Further studies are required to define if diaphragm unloading may explain the favourable results of HF OXY in clinical trials. References 1. Hernández G, Vaquero C, González P, Subira C, Frutos-Vivar F, Rialp G, Laborda C, Colinas L, Cuena R,Fernández R. - Effect of Postextubation High-Flow Nasal Cannula vs Conventional Oxygen Therapyon Reintubation in Low-Risk Patients: A Randomized Clinical Trial.- JAMA 2016 Apr 5;315(13):1354–61. 2. Maggiore SM, Idone FA, Vaschetto R, Festa R, Cataldo A, Antonicelli F, Montini L, De Gaetano A, Navalesi P, Antonelli M. - Nasal high-flow versus Venturi mask oxygen therapy after extubation. Effects on oxygenation, comfort, and clinical outcome - Am J Respir Crit Care Med. 2014 Aug 1;190(3):282–8Table 24 (abstract A917). Breathing pattern parameters LF OXY HF OXY P value RR (breaths/min.) 27.76 ± 2.32 26.34 ± 1.84 0.1 VT (L) 0.47 ± 0.12 0.36 ± 0.09 0.04 PaCO2 (mmHg) 49.24 ± 2.82 47.84 ± 1.64 0.09 Data are expressed as mean ± standard deviation (SD) Fig. 45 (abstract A917). Diaphragmatic activity: LF OXY vs HF oxy

2. Maggiore SM, Idone FA, Vaschetto R, Festa R, Cataldo A, Antonicelli F, Montini L, De Gaetano A, Navalesi P, Antonelli M. - Nasal high-flow versus Venturi mask oxygen therapy after extubation. Effects on oxygenation, comfort, and clinical outcome - Am J Respir Crit Care Med. 2014 Aug 1;190(3):282–8Table 24 (abstract A917). Breathing pattern parameters LF OXY HF OXY P value RR (breaths/min.) 27.76 ± 2.32 26.34 ± 1.84 0.1 VT (L) 0.47 ± 0.12 0.36 ± 0.09 0.04 PaCO2 (mmHg) 49.24 ± 2.82 47.84 ± 1.64 0.09 Data are expressed as mean ± standard deviation (SD) Fig. 45 (abstract A917). Diaphragmatic activity: LF OXY vs HF oxy A918 Non-invasive mechanical ventilation in acute respiratory failure in critically ill patients with confirmed influenza infection: a chaid decision-tree analysis A. Rodriguez1,2, I. Martín-Loeches3, E. Díaz4, J.R. Masclans5, F. Gordo6, J. Solé-Violán7, M. Bodí1,2, F.X. Avilés-Jurado2,8, S. Trefler1,2, M. Magret1,2, L.F. Reyes9, J. Marín-Corral5, J.C. Yebenes10, A. Esteban11, A. Anzueto9, S. Aliberti12, M.I. Restrepo9, GETGAG/SEMICYUC 1Hospital Universitari Joan XXIII, Critical Care Medicine, Tarragona, Spain; 2Universitat Rovira i Virgili/IISPV, Tarragona, Spain; 3St James's University Hospital. Trinity Center for Health Sciences, Anaesthesia and Critical Care, Dublin, Ireland; 4Hospital Parc Tauli, Critical Care Medicine, Sabadell, Spain; 5Hospital del Mar/CIBERES/UPF, Critical Care Medicine, Barcelona, Spain; 6Hospital del Henares, Critical Care Medicine, Madrid, Spain; 7Hospital Dr. Negrín, Critical Care Medicine, Las Palmas de Gran Canaria, Spain; 8Hospital Universitari Joan XXIII, ORL, Tarragona, Spain; 9UT Heatlh Science Center at San Antonio and South Texas Veterans Health Care System, Pulmonary Disease and Critical Care, San Antonio, United States; 10Hospital de Mataró, Critical Care Medicine, Mataró, Spain; 11Hospital de Getafe, Critical Care Medicine, Madrid, Spain; 12University of Milan - Biocca San Genaro Hospital, School of Medicine and Surgery, Monza, Italy Correspondence: A. Rodriguez – Hospital Universitari Joan XXIII, Critical Care Medicine, Tarragona, Spain Introduction: I Non-invasive mechanical ventilation (NIV) has been seen to play a major role in decreasing intubation rates in patients with severe exacerbation of chronic obstructive pulmonary disease and congestive heart failure. Unsuccessful NIV has been found to be independently associated with increased mortality in patients with ARF. The NIV failure and their impact on mortality in patients with inlfuenza infection is unknown.

n rates in patients with severe exacerbation of chronic obstructive pulmonary disease and congestive heart failure. Unsuccessful NIV has been found to be independently associated with increased mortality in patients with ARF. The NIV failure and their impact on mortality in patients with inlfuenza infection is unknown. Objectives:to describe non-invasive ventilation failure (NIVf) rate, to identify risk factor for NIVf using CHAID (Chi-square Automatic Interaction Detection) and to determine if NIVf is associated with ICU-mortality. Methods: Secondary analysis in 1,898 patients with influenza requiring mechanical ventilation(MV). Three groups were considered:patients with NIV who failed (Group A); patients with NIV who succeeded (Group B); and patients with invasive MV (Group C). Cox analysis was used to assess survival. Risk factors for NIVf were obtained using CHAID.

Methods: Secondary analysis in 1,898 patients with influenza requiring mechanical ventilation(MV). Three groups were considered:patients with NIV who failed (Group A); patients with NIV who succeeded (Group B); and patients with invasive MV (Group C). Cox analysis was used to assess survival. Risk factors for NIVf were obtained using CHAID. Results: From a total study population of 1898, 806 patients underwent NIV and 1092 received IMV (Group C). Among NIV patients, 56.8 % (Group A) failed and had higher (p < 0.01) APACHE II (17 vs. 14), SOFA (7 vs. 4), infiltrates in chest x-ray (CXR = 3 vs. 2) and ICU-mortality (38.4 % vs. 6.3 %) than Group B. SOFA(S) was the variable most associated with NIVf by CHAID and 2 cut-offs were determined. Patients with S > 5 had more risk of NIVf (OR = 3.3,95%CI:2.4-4.5). Among patients with S < 5, 40.3 % failed and chronic obstructive pulmonary disease (COPD) was associated with low risk of NIVf (OR = 0.59,95%CI:0.41-0.83). NIVf was independently associated with ICU-mortality (OR = 11.4,95%CI:6.5-20.1). ICU-mortality was higher for Group A (38.4 %) than Group C (31.3 %, p < 0.01). A trend of high adjusted ICU-mortality was observed for Group A patients versus Group C (HR = 1.19;95%CI:0.99-1.44;p = 0.07). Conclusions: NIV failure is frequent and independently associated with ICU-mortality in patients with influenza. CHAID analysis might be a promising tool to assist in clinical decision-making. References 1. Masclans JR, et al. Early non-invasive ventilation treatment for severe influenza pneumonia. Clin Microbiol Infect 2013;19: 249–2562.

Conclusions: NIV failure is frequent and independently associated with ICU-mortality in patients with influenza. CHAID analysis might be a promising tool to assist in clinical decision-making. References 1. Masclans JR, et al. Early non-invasive ventilation treatment for severe influenza pneumonia. Clin Microbiol Infect 2013;19: 249–2562. 2. Nicolini A, et al. Effectiveness and predictors of success of noninvasive ventilation during H1M1 pandemics: a multicenter study. Minerva Anestesiol 2012; 78: 1333–40 3. Nin N, et al. Clinical characteristics and outcome of patients with 2009 influenza A(H1N1) virus infection with respiratory failure requiring mechanical ventilation. J Crit Care 2011; 26: 186–192 Grant acknowledgment GETGAG/SEMICYUC

2. Nicolini A, et al. Effectiveness and predictors of success of noninvasive ventilation during H1M1 pandemics: a multicenter study. Minerva Anestesiol 2012; 78: 1333–40 3. Nin N, et al. Clinical characteristics and outcome of patients with 2009 influenza A(H1N1) virus infection with respiratory failure requiring mechanical ventilation. J Crit Care 2011; 26: 186–192 Grant acknowledgment GETGAG/SEMICYUC AKI: OUTCOMES AND IMPROVING MANAGEMENT A919 Renal perfusion, function and oxygenation in the early postoperative period after liver transplantation J. Skytte Larsson1, B. Redfors2, S.-E. Ricksten1 1University of Gothenburg, Department of Anesthesia and Intensive Care, Sahlgrenska University Hospital, Göteborg, Sweden; 2University of Gothenburg, Department of Cardiothoracic Anesthesia and Intensive Care, Sahlgrenska University Hospital, Göteborg, Sweden Correspondence: J. Skytte Larsson – University of Gothenburg, Department of Anesthesia and Intensive Care, Sahlgrenska University Hospital, Göteborg, Sweden Introduction: Acute kidney injury (AKI) after liver transplantation is a common complication with an incidence of approximately 50 % [1], resulting in high morbidity and mortality. To increase the possibilities to prevent or treat AKI after liver transplantation, it is essential to increase the knowledge on changes in renal physiology after liver transplantation.

liver transplantation is a common complication with an incidence of approximately 50 % [1], resulting in high morbidity and mortality. To increase the possibilities to prevent or treat AKI after liver transplantation, it is essential to increase the knowledge on changes in renal physiology after liver transplantation. Objectives: The aim of this study was to gain insights into renal perfusion, filtration and oxygenation in the immediate postoperative period in patients undergoing liver transplantation and to compare these data to those obtained from a group of patients undergoing major surgery with no postoperative renal impairment.

Objectives: The aim of this study was to gain insights into renal perfusion, filtration and oxygenation in the immediate postoperative period in patients undergoing liver transplantation and to compare these data to those obtained from a group of patients undergoing major surgery with no postoperative renal impairment. Methods: Informed consent was obtained preoperatively from twelve patients with normal renal function accepted for liver transplantation. Glomerular filtration rate (GFR) was measured preoperatively by plasma clearance of Cr-EDTA. The patients were studied after liver transplantation in the ICU in the immediate postoperative period, sedated and mechanically ventilated. Systemic haemodynamics and renal variables where obtained during two 30-min periods. Renal blood flow (RBF) and GFR were measured by the renal vein retrograde thermodilution technique and by renal extraction of Cr-EDTA (=filtration fraction, FF), respectively. Arterial (a) and renal vein (rv) blood samples were taken for measurements of arterial (CaO2) and renal vein (CrvO2) oxygen contents. Renal oxygen consumption [RVO2 = RBF x (CaO2-CrvO2)], renal oxygen delivery (RDO2 = RBF x CaO2) and renal oxygen extraction [RO2Ex = (CaO2-CrvO2)/CaO2)] were calculated. Sixty-three patients undergoing uneventful cardiac surgery with no postoperative renal impairment served as controls.

al vein (CrvO2) oxygen contents. Renal oxygen consumption [RVO2 = RBF x (CaO2-CrvO2)], renal oxygen delivery (RDO2 = RBF x CaO2) and renal oxygen extraction [RO2Ex = (CaO2-CrvO2)/CaO2)] were calculated. Sixty-three patients undergoing uneventful cardiac surgery with no postoperative renal impairment served as controls. Results: Cardiac index (65 %) and systemic oxygen delivery index (62 %) were higher and systemic vascular resistance index was lower (−38 %) in the liver transplant group compared to controls (p < 0.001). RBF was 17 % higher and renal vascular resistance was 16 % lower compared to controls (p < 0.05). In the liver transplanted group, GFR was 35 % lower compared to the preoperative value (p = 0.016), accompanied by a 41 % increase in serum creatinine (p < 0.05). After surgery, when compared to controls, GFR and FF was 23 % and 40 % lower, respectively (p < 0.05, p < 0.01), and RVO2 and RO2Ex were 42 % and 24 % higher, respectively, in the liver transplanted patients (p < 0.01, p < 0.05). Conclusions: Despite the hyperdynamic systemic circulation, GFR is considerably reduced immediately after liver transplantation, most likely caused by a post-glomerular renal vasodilation decreasing upstream glomerular filtration pressure. Renal oxygenation is impaired after liver transplantation due to the high RVO2, which was not met by a proportional increase in RDO2. References 1. Hilmi IA et al. Acute kidney injury following orthotopic liver transplantation: incidence, risk factors, and effects on patient and graft outcomes. Brit J Anaesth 2015, 114:919.

Conclusions: Despite the hyperdynamic systemic circulation, GFR is considerably reduced immediately after liver transplantation, most likely caused by a post-glomerular renal vasodilation decreasing upstream glomerular filtration pressure. Renal oxygenation is impaired after liver transplantation due to the high RVO2, which was not met by a proportional increase in RDO2. References 1. Hilmi IA et al. Acute kidney injury following orthotopic liver transplantation: incidence, risk factors, and effects on patient and graft outcomes. Brit J Anaesth 2015, 114:919. A920 Association between aki stages and long-term mortality R. Haines, J. Powell-Tuck, H. Leonard, M. Ostermann Guy’s & St Thomas Hospital, London, United Kingdom Correspondence: M. Ostermann – Guy's & St Thomas Hospital, London, United Kingdom Introduction: Acute kidney injury is common in critically ill patients and associated with increased short and long-term mortality. Most published studies have focussed on patients with severe AKI. Little is known about the long-term outcome of patients with less severe AKI. Objectives: Our objective was to determine the outcome of patients with different stages of AKI at 5 and 7 years after admission to the Intensive Care Unit (ICU).

A920 Association between aki stages and long-term mortality R. Haines, J. Powell-Tuck, H. Leonard, M. Ostermann Guy’s & St Thomas Hospital, London, United Kingdom Correspondence: M. Ostermann – Guy's & St Thomas Hospital, London, United Kingdom Introduction: Acute kidney injury is common in critically ill patients and associated with increased short and long-term mortality. Most published studies have focussed on patients with severe AKI. Little is known about the long-term outcome of patients with less severe AKI. Objectives: Our objective was to determine the outcome of patients with different stages of AKI at 5 and 7 years after admission to the Intensive Care Unit (ICU). Methods: We retrospectively analysed the data of all adult patients admitted to a multi-disciplinary ICU in a teaching hospital in the UK between March 2004 - May 2009. Patients with chronic dialysis dependent renal failure were excluded. Patients were categorised according to their maximum stage of AKI during stay in ICU as defined by the serum creatinine criteria of the KDIGO classification. APACHE II and SOFA scores were used to describe severity of illness on admission to ICU. In patients with >1 admission to ICU, we only included the first admission in the analysis. Results: Data of 3094 adult patients were analysed of whom 53 % had AKI during their stay in ICU. Patients with any degree of AKI had a higher mortality at 5 and 7 years but they were also sicker on admission to ICU.

Methods: We retrospectively analysed the data of all adult patients admitted to a multi-disciplinary ICU in a teaching hospital in the UK between March 2004 - May 2009. Patients with chronic dialysis dependent renal failure were excluded. Patients were categorised according to their maximum stage of AKI during stay in ICU as defined by the serum creatinine criteria of the KDIGO classification. APACHE II and SOFA scores were used to describe severity of illness on admission to ICU. In patients with >1 admission to ICU, we only included the first admission in the analysis. Results: Data of 3094 adult patients were analysed of whom 53 % had AKI during their stay in ICU. Patients with any degree of AKI had a higher mortality at 5 and 7 years but they were also sicker on admission to ICU. Conclusions: Any stage of AKI during critical illness is associated with an increased risk of mortality at 5 and 7 years. Mortality is highest in patients with AKI II and III. More work is necessary to explore the relationship between AKI and long-term outcome and to identify independent risk factors for mortality.Table 25 (abstract A920). AKI stages and outcome

l illness is associated with an increased risk of mortality at 5 and 7 years. Mortality is highest in patients with AKI II and III. More work is necessary to explore the relationship between AKI and long-term outcome and to identify independent risk factors for mortality.Table 25 (abstract A920). AKI stages and outcome No AKI AKI I AKI II AKI III Number of patients 1457 (47 %) 776 (25 %) 166 (5.4 %) 695 (22.5 %) Age, mean (SD) 53 (18) 62 (17) 60 (17) 62 (16) Male gender 59 % 68 % 61 % 64 % APACHE II on admission to ICU, mean (SD) 12.9 (5.0) 16 (5.4) 16.6 (5.6) 19.8 (5.9) SOFA score on admission to ICU, mean (SD) 3.5 (2.4) 5.3 (2.7) 6.1 (2.8) 7.5 (3.1) 1-year mortality 13.2 % 18.9 % 27.7 % 22.3 % 5-year mortality 27.3 % 34.3 % 42.2 % 40.1 % 7-year mortality 30.8 % 38.1 % 45.2 % 46.2 %

59 % 68 % 61 % 64 % APACHE II on admission to ICU, mean (SD) 12.9 (5.0) 16 (5.4) 16.6 (5.6) 19.8 (5.9) SOFA score on admission to ICU, mean (SD) 3.5 (2.4) 5.3 (2.7) 6.1 (2.8) 7.5 (3.1) 1-year mortality 13.2 % 18.9 % 27.7 % 22.3 % 5-year mortality 27.3 % 34.3 % 42.2 % 40.1 % 7-year mortality 30.8 % 38.1 % 45.2 % 46.2 % A921 Forced fluid removal vs. usual care in intensive care patients with high-risk acute kidney injury and severe fluid overload (FFAKI) - a randomized clinical pilot trial R.E. Berthelsen1, T.S. Itenov1, A. Perner2, J.U. Jensen3, M. Ibsen1, A.E.K. Jensen4, M.H. Bestle1 1Nordsjællands Hospital, Dept. of Anaesthesiology and Intensive Care, Hillerød, Denmark; 2Rigshospitalet, Copenhagen University Hospital, Dept. of intensive Care, Copenhagen, Denmark; 3Rigshospitalet, Copenhagen University Hospital, CHIP & PERSIMUNE, Dept. of Infectious Diseases, Copenhagen, Denmark; 4University of Copenhagen, Dept. of Biostatistics, Copenhagen, Denmark Correspondence: R.E. Berthelsen – Nordsjællands Hospital, Dept. of Anaesthesiology and Intensive Care, Hillerød, Denmark Introduction and objective: Observational studies of intensive care unit (ICU) patients with acute kidney injury have shown a negative correlation between accumulation of fluids and survival [1]. It is unknown whether rapid removal of accumulated fluids is feasible and beneficial. Therefore we wish to perform a pilot trial of forced fluid removal vs. standard care in critically ill patients with high-risk acute kidney injury and severe fluid overload.

correlation between accumulation of fluids and survival [1]. It is unknown whether rapid removal of accumulated fluids is feasible and beneficial. Therefore we wish to perform a pilot trial of forced fluid removal vs. standard care in critically ill patients with high-risk acute kidney injury and severe fluid overload. Methods: The FFAKI-trial is a pilot, multicenter, randomized clinical trial recruiting adult intensive care patients with high-risk acute kidney injury and fluid overload defined as > 10 % of ideal bodyweight. To reduce the signal-to-noise ratio we only wish to include patients with a high baseline risk of persistent renal failure. Baseline risk will be calculated using a newly developed model, the renal recovery score (RRS), to predict the chance of recovering renal function within 28 days. In- and exclusion criteria are shown in Tables 26 and 27.

noise ratio we only wish to include patients with a high baseline risk of persistent renal failure. Baseline risk will be calculated using a newly developed model, the renal recovery score (RRS), to predict the chance of recovering renal function within 28 days. In- and exclusion criteria are shown in Tables 26 and 27. Patients are randomized to either forced fluid removal or standard care for the entire ICU stay. Forced fluid removal is done by infusion of furosemide and/or fluid removal with continuous renal replacement therapy. The fluid removal rate is adjusted 3 times daily to achieve a therapeutic goal of net negative fluid balance ≥ 1 ml/kg/h. Physiologic tolerance to fluid removal is continually evaluated according to predefined criteria of hypoperfusion: Lactate ≥ 4 mmol/l, mean arterial pressure < 50 mmHg or mottling beyond the edge of the kneecaps. In case of hypoperfusion, fluid removal is suspended until all criteria have been resolved for a minimum of 1 hour. The flow chart for the experimental FFAKI-treatment is seen in Figure 46, 47, 48. The primary outcome is cumulative fluid balance 5 days after randomization. By inclusion of 50 patients we are able to detect a difference of 5.9 L between groups (α = 0.05 and β = 0.80). Secondary outcomes include mean daily fluid balances, fluid balance at discharge from ICU, time to neutral fluid balance, number of serious adverse reactions and number of protocol violations. All patients are followed for 90 days.

ents we are able to detect a difference of 5.9 L between groups (α = 0.05 and β = 0.80). Secondary outcomes include mean daily fluid balances, fluid balance at discharge from ICU, time to neutral fluid balance, number of serious adverse reactions and number of protocol violations. All patients are followed for 90 days. Conclusions: The FFAKI trial started in October 2015 and will, when completed, provide data to evaluate if a larger trial of forced fluid removal in ICU patients is feasible. Our primary outcome will show if the experimental intervention leads to a clinically relevant difference in fluid balance which would warrant a definitive trial powered for survival at 90 days. Grant acknowledgment The FFAKI-trial is funded by the Research Department and the Dept. of Anaesthesiology at Nordsjællands Hospital, Denmark Trial registration. EudraCT: 2015-001701-13, ClinicalTrials.gov: NCT02458157 References [1] Zhang L, Chen Z et al. Associations of fluid overload with mortality and kidney recovery in patients with acute kidney injury: a systematic review and meta-analysis. J Crit Care 2015;30:860.e7-860.e13Table 26 (abstract A921). Inclusion criteria Age ≥ 18 years Acute Kidney Injury defined according to the KDIGO criteria. Renal Recovery Score ≤ 60 %. Fluid overload defined as a positive fluid balance ≥ 10 % of ideal body weight. Able to undergo randomization within 12 hours of fulfilling the other inclusion criteria Table 27 (abstract A921). Exclusion criteria

[1] Zhang L, Chen Z et al. Associations of fluid overload with mortality and kidney recovery in patients with acute kidney injury: a systematic review and meta-analysis. J Crit Care 2015;30:860.e7-860.e13Table 26 (abstract A921). Inclusion criteria Age ≥ 18 years Acute Kidney Injury defined according to the KDIGO criteria. Renal Recovery Score ≤ 60 %. Fluid overload defined as a positive fluid balance ≥ 10 % of ideal body weight. Able to undergo randomization within 12 hours of fulfilling the other inclusion criteria Table 27 (abstract A921). Exclusion criteria Known allergy to furosemide or sulphonamides Known pre-hospitalization advanced chronic kidney disease. (eGFR < 30 mL/minute/1.73 m2 or chronic RRT) Severe hypoxic respiratory failure (use of invasive ventilation and FiO2 > 80 % and PEEP > 10 cm H2O) Severe burn injury (≥10 % TBSA) Severe dysnatremia (<120 or > 155 mmol/l) Hepatic coma Mentally disabled undergoing forced treatment. Pregnancy/breast feeding Lack of commitment for on-going life support including RRT Lack of informed consent Fig. 46 (abstract A921). FFAKI algorithm - Rescucitation Fig. 47 (abstract A921). FFAKI algorithm - Furosemide Fig. 48 (abstract A921). FFAKI algorithm - CRRT

Known allergy to furosemide or sulphonamides Known pre-hospitalization advanced chronic kidney disease. (eGFR < 30 mL/minute/1.73 m2 or chronic RRT) Severe hypoxic respiratory failure (use of invasive ventilation and FiO2 > 80 % and PEEP > 10 cm H2O) Severe burn injury (≥10 % TBSA) Severe dysnatremia (<120 or > 155 mmol/l) Hepatic coma Mentally disabled undergoing forced treatment. Pregnancy/breast feeding Lack of commitment for on-going life support including RRT Lack of informed consent Fig. 46 (abstract A921). FFAKI algorithm - Rescucitation Fig. 47 (abstract A921). FFAKI algorithm - Furosemide Fig. 48 (abstract A921). FFAKI algorithm - CRRT A922 Sequential measurement of 1 hour creatinine clearance (1-CRCL) in critically ill patients at risk of acute kidney injury (AKI) T. Bucknall1, J. Dixon2, F. Boa3, I. MacPhee4, B.J. Philips5, AKI Research Group, St George's University of London 1St George's, University of London, London, United Kingdom; 2St Georges NHS Foundation Trust, General Intensive Care, London, United Kingdom; 3St Georges NHS Foundation Trust, Clinical Blood Sciences, London, United Kingdom; 4St George's, University of London, Renal Medicine, London, United Kingdom; 5St George's, University of London, General Intensive Care, London, United Kingdom Correspondence: T. Bucknall – St George's, University of London, London, United Kingdom Introduction: Measuring renal function in acute kidney injury (AKI) is difficult. The criteria on which a clinical diagnosis of AKI is made are broad and open to interpretation. More objective measures remain elusive. A truly robust biomarker has not been discovered and exogenous markers are difficult to use in critically ill patients. We have previously used measures of renal clearance of Iohexol (rICl) to evaluate AKI (data submitted) and have shown that 1-hour renal creatinine clearances (1-CrCl) correlate well with rICl (correlation r = 0.99; Bland Altman bias −5.6, SD 24.1 %) performing better than 4 hourly CrCl measurements (bias 11.4 % SD 56.7 %).

ave previously used measures of renal clearance of Iohexol (rICl) to evaluate AKI (data submitted) and have shown that 1-hour renal creatinine clearances (1-CrCl) correlate well with rICl (correlation r = 0.99; Bland Altman bias −5.6, SD 24.1 %) performing better than 4 hourly CrCl measurements (bias 11.4 % SD 56.7 %). Objective: To use sequential 1-CrCl measurements to measure renal function in critically ill patients with or at risk of AKI over 72 hours and to compare this with traditional methods of assessing AKI. This was part of a BSc project. Methods: Consent was obtained from the next of kin according to ethical review (NREC: 15/LO/1720). 1-hour urine samples were collected at 7 am and 7 pm. Serum creatinine (SCr) samples were taken 30 min after the urine collections were started. SCr and urine creatinine (UCr) concentrations were measured using the enzymatic method by the dedicated researcher. 1-CrCl was calculated by the equation: CrCl=(UCr×vol)1.73/((SCr/min)xBSA). 1-CrCls were compared with the ´kidney disease improving global outcomes´ (KDIGO) criteria for AKI and estimated glomerular filtration rates (eGFR) calculated by the Cockcroft-Gault (CG), modification of diet in renal disease 7 (MDRD7) and chronic kidney disease epidemiology collaboration (CKI-EPI) methods. Kruskal Wallis, Spearman correlation and the Bland-Altman method were used (IBM® SPSS® version 22) in analysis.

nd estimated glomerular filtration rates (eGFR) calculated by the Cockcroft-Gault (CG), modification of diet in renal disease 7 (MDRD7) and chronic kidney disease epidemiology collaboration (CKI-EPI) methods. Kruskal Wallis, Spearman correlation and the Bland-Altman method were used (IBM® SPSS® version 22) in analysis. Results: 17 patients were included with a total of 87 1-CrCl measurements. Median (IQR) age was 67 (50–76) and admission APACHE II score 22 (22–32). 14 patients were medical admissions and 3 had emergency surgery. Observed overtime the 1-CrCl changed in similar patterns to the calculated eGFR (Figs. 49 and 50, examples) and correlations were good (1-CrCl vs CG r = 0.857, 1-CrCl vs CKD-epi = 0.861, 1-CrCl vs MDRD7 = 0.845). However the eGFR did not predict actual values. The Bland Altman comparisons for CKD-epi and MDRD7 are in Table 28. Fig 51 compares 1-CrCl with the KDIGO criteria Conclusions: Despite good correlations, the eGFR calculated values were not good measures of actual renal function. 1-CrCl was often although not always markedly lower than the eGFR. The KDIGO criteria did reflect the measured 1-CrCl but there was marked overlap in all categories. References 1. Endre ZH et al., Nat Rev Nephrol, 2014; 10: 683–385. 2. Bagshaw SM et al., Nephrol Dial Transplant 2008, 23:1569–74 3. Kirwan CJ et al., Critical care research and practice 2013, 1–8. 4. Dixon J et al., J Transl Med, 2015; 12:Fig. 49 (abstract A922). Individual example 1 Fig. 50 (abstract A922). Individual example 2 Table 28 (abstract A922). Bland Altman comparison of 1-CrCl with eGFR

2. Bagshaw SM et al., Nephrol Dial Transplant 2008, 23:1569–74 3. Kirwan CJ et al., Critical care research and practice 2013, 1–8. 4. Dixon J et al., J Transl Med, 2015; 12:Fig. 49 (abstract A922). Individual example 1 Fig. 50 (abstract A922). Individual example 2 Table 28 (abstract A922). Bland Altman comparison of 1-CrCl with eGFR Comparison Bias (%) SD(%) Limits of Agreement (%) 1CrCl vs CKD-epi −6.5 21.3 −39.4 to 44.1 1CrCl vs MDRD7 2.4 20.5 −46.5 to 33.9 Fig. 51 (abstract A922). Boxplot of KDIGO and 1 hour CrCl

4. Dixon J et al., J Transl Med, 2015; 12:Fig. 49 (abstract A922). Individual example 1 Fig. 50 (abstract A922). Individual example 2 Table 28 (abstract A922). Bland Altman comparison of 1-CrCl with eGFR Comparison Bias (%) SD(%) Limits of Agreement (%) 1CrCl vs CKD-epi −6.5 21.3 −39.4 to 44.1 1CrCl vs MDRD7 2.4 20.5 −46.5 to 33.9 Fig. 51 (abstract A922). Boxplot of KDIGO and 1 hour CrCl A923 Does the implementation of a quality improvement care bundle reduce the incidence of acute kidney injury in those undergoing an emergency laparotomy J. Doyle1, F. Saadat1, T. Samuels1, S. Huddart2, B. McCormick3, R. DeBrunnar3, J. Preece3, M. Swart4, C. Peden5, S. Richardson6, L. Forni1 1Department of Intensive Care Medicine and Surrey Peri-Operative Anaesthesia and Critical Care Collaborative Research Group (SPACER), Guildford, United Kingdom; 2Department of Anaesthesia, Royal Surrey County Hospital NHS Foundation Trust, Guildford, United Kingdom; 3Royal Devon and Exeter Hospital, Exeter, United Kingdom; 4South Devon Healthcare NHS Foundation Trust, Torquay, United Kingdom; 5Department of Anaesthesia and Intensive Care, Royal United Hospital Bath NHS Trust, Bath, United Kingdom; 6Department of Surgery, Royal United Hospital Bath NHS Trust, Bath, United Kingdom Correspondence: J. Doyle – Department of Intensive Care Medicine and Surrey Peri-Operative Anaesthesia and Critical Care Collaborative Research Group (SPACER), Guildford, United Kingdom Introduction: Enhanced recovery pathways have been a focus for patient optimisation of morbidiy and mortality in the post-operative patient. Significant mortality improvement was seen following the implementation of the emergency laparotomy pathway quality improvement care (ELPQuiC) bundle with an adjusted risk of death from 15.6 % to 9.6 % (1). The first national emergency laparotomy audit (NELA) has since been published demonstrating a 30-day mortality of 11 % and recommending access to pathways that identify need to escalate care (2). However acute kidney injury (AKI) in critically unwell patients remains a major source of mortality, of up to 60 %, and morbidity (3). It is not yet clear whether enhanced recovery pathways, specifically those that utilise early goal directed therapy, affect the incidence of AKI.

hat identify need to escalate care (2). However acute kidney injury (AKI) in critically unwell patients remains a major source of mortality, of up to 60 %, and morbidity (3). It is not yet clear whether enhanced recovery pathways, specifically those that utilise early goal directed therapy, affect the incidence of AKI. Objectives: To determine if there was a difference in incidence of combined AKI pre and post implementation of an enhanced recovery protocol, one that had already demonstrated a significant mortality benefit. Methods: A subgroup analysis of the data gathered via the ELPQuiC bundle was performed (1). We obtained buy-in from the 4 participating centres and requested an extrapolation of values from their raw data. If required further data was obtained via the hospital's electronic path system. All data was reviewed by a second investigator. We defined the baseline creatinine as the best available preoperative creatinine from the past 1 year. The data recorded included creatinine at baseline, post-op, worse recorded creatinine between day1 and day30, MAKE 30, P-Possum and 30-day mortality data. CKD stage was identified via MDRD equation with age, gender and baseline creatinine. Patients with AKI were stratified according to KIDGO stages of AKIN. Primary outcome was the incidence of AKI in each of combined pre and post ELQUIC patient population. Secondary outcome included the stage specific incidence of AKI.

We defined the baseline creatinine as the best available preoperative creatinine from the past 1 year. The data recorded included creatinine at baseline, post-op, worse recorded creatinine between day1 and day30, MAKE 30, P-Possum and 30-day mortality data. CKD stage was identified via MDRD equation with age, gender and baseline creatinine. Patients with AKI were stratified according to KIDGO stages of AKIN. Primary outcome was the incidence of AKI in each of combined pre and post ELQUIC patient population. Secondary outcome included the stage specific incidence of AKI. Results: There was no significant difference between the cumulative incidence of AKIN pre and post ELQUIC implementation on day 1 post-op (18.4 % vs 19.8 %, P = 0.6920) or day 30 post-op (14.3 % vs 8.6 %, P = 0.0647). Conclusion: This multi-centre cohort subgroup analysis demonstrates that the implementation of a quality improvement care bundle does not affect the incidence of AKI. This is in contrast to the clear mortality benefit that such a care bundle has provided and provides stimulus to discover what factors may yet improve AKI, and so further improve these patients outcome. References 1. Huddart S, Peden CJ, Swart M, et al. Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. Br J Surg. 2015;102(1):57 2. The First Patient Report of the National Emergency Laparotomy Audit - Oct 2015. 3. Chang CH, Fan PC, Chang MY, et al. Acute kidney injury enhances outcome prediction ability of sequential organ failure assessment score in critically ill patients. PLoS One. 2014;9(10):e109649

1. Huddart S, Peden CJ, Swart M, et al. Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. Br J Surg. 2015;102(1):57 2. The First Patient Report of the National Emergency Laparotomy Audit - Oct 2015. 3. Chang CH, Fan PC, Chang MY, et al. Acute kidney injury enhances outcome prediction ability of sequential organ failure assessment score in critically ill patients. PLoS One. 2014;9(10):e109649 Grants - Nil received. QUALITY OF LIFE ISSUES IN THE ICU A924 Reduction of self-perceived discomforts in critically ill patients in french intensive care units: a cluster-randomized controlled trial P. Kalfon1, K. Baumstarck2, P. Estagnasie3, M.-A. Geantot4, A. Berric5, G. Simon6, B. Floccard7, T. Signouret8, M. Boucekine2, M. Fromentin9, M. Nyunga10, A. Sossou11, M. Venot12, R. Robert13, A. Follin14, A. Renault15, M. Garrouste16, O. Collange17, Q. Levrat18, I. Villard19, D. Thévenin20, J. Pottecher21, R.-G. Patrigeon22, N. Revel23, C. Vigne24, O. Mimoz25, P.

-A. Geantot4, A. Berric5, G. Simon6, B. Floccard7, T. Signouret8, M. Boucekine2, M. Fromentin9, M. Nyunga10, A. Sossou11, M. Venot12, R. Robert13, A. Follin14, A. Renault15, M. Garrouste16, O. Collange17, Q. Levrat18, I. Villard19, D. Thévenin20, J. Pottecher21, R.-G. Patrigeon22, N. Revel23, C. Vigne24, O. Mimoz25, P. Auquier2, IPREA Study Group 1CH Chartres, Réanimation, Chartres, France; 2Aix-Marseille Université, Unité de Recherche EA 3279, Marseille, France; 3Clinique Ambroise Paré, Réanimation, Neuilly/Seine, France; 4CHU Dijon Bourgogne, Anesthésie Réanimation, Dijon, France; 5CHI Toulon/La Seyne sur Mer, Réanimation Polyvalente, Toulon, France; 6CH Troyes, Réanimation, Troyes, France; 7CHU Edouard Herriot, Hospices Civils de Lyon, Réanimation Polyvalente, Lyon, France; 8Hôpital Européen de Marseille, Réanimation, Marseille, France; 9CHU Cochin, AP-HP, Réanimation Chirurgicale, Paris, France; 10CH Victor Provo, Réanimation Polyvalente, Roubaix, France; 11CH Emile Roux, Réanimation, Le Puy-en-Velay, France; 12CHU Saint-Louis, Réanimation Médicale, Paris, France; 13CHU La Milétrie, Réanimation Médicale, Poitiers, France; 14Hôpital Européen Georges Pompidou, AP-HP, Réanimation Chirugicale, Paris, France; 15CHU Brest, Réanimation Médicale, Brest, France; 16Groupe Hospitalier Paris Saint-Joseph, Médecine Intensive et Réanimation, Paris, France; 17CHU Strasbourg, Hôpital Civil, Réanimation Chirugicale Polyvalente, Strasbourg, France; 18Groupe Hospitalier de La Rochelle - Ré - Aunis, Réanimation, La Rochelle, France; 19CHU Beaujon, AP-HP, Anesthésie Réanimation, Clichy, France; 20CH Lens, Réanimation, Lens, France; 21CHU Strasbourg, Hôpital Hautepierre, Réanimation Chirugicale, Strasbourg, France; 22CH Auxerre, Réanimation, Auxerre, France; 23CHU Nice, Hôpital Pasteur, Réanimation Médico-Chirurgicale, Nice, France; 24CHU Hôpital Nord, AP-HM, Réanimation Chirugicale, Marseille, France; 25CHU La Milétrie, Réanimation Chirugicale, Poitiers, France Correspondence: P. Kalfon – CH Chartres, Réanimation, Chartres, France Introduction: It is now well documented that critically ill patients are exposed to stressful conditions and experience discomforts from multiple sources. Improved identification of the discomforts of patients in intensive care units (ICUs) may have implications for managing their care, including consideration of ethical issues, and may assist clinicians in choosing the most appropriate interventions.

stressful conditions and experience discomforts from multiple sources. Improved identification of the discomforts of patients in intensive care units (ICUs) may have implications for managing their care, including consideration of ethical issues, and may assist clinicians in choosing the most appropriate interventions. Objectives: The primary objective of this study was to assess the effectiveness of a multicomponent program (MCP) of discomfort reduction in critically ill patients. The secondary objectives were to assess the sustainability of the impact of the program and the potential seasonality effect.

stressful conditions and experience discomforts from multiple sources. Improved identification of the discomforts of patients in intensive care units (ICUs) may have implications for managing their care, including consideration of ethical issues, and may assist clinicians in choosing the most appropriate interventions. Objectives: The primary objective of this study was to assess the effectiveness of a multicomponent program (MCP) of discomfort reduction in critically ill patients. The secondary objectives were to assess the sustainability of the impact of the program and the potential seasonality effect. Methods: We conducted a multicenter, cluster-randomized, controlled, single (patient)-blind study involving 34 French adult ICUs. The experimental intervention was the implementation of the MCP including the following steps: identification of discomforts, immediate feedback to the healthcare team, and implementation of targeted interventions under control of local champions who received monthly feedback and organized monthly meetings with their healthcare team. All ICUs started with a 1-month period with no intervention, and then they were randomized to one of two groups: 17 ICUs with MCP implemented during a 6-month period (experimental group) and 17 ICUs without any programm during the same period (control group). To assess the sustainabilty of the impact of the MCP, the study was completed with a second 6-month period during which the MCP was no longer applied in the experimental group. The primary endpoint was the monthly overall score of self-reported discomfort from the French 16-item questionnaire on discomforts in ICU patients (IPREA)1(range from 0 to 100, the lowest possible level of discomfort to the highest). The secondary endpoints were the scores of each item of IPREA.

he experimental group. The primary endpoint was the monthly overall score of self-reported discomfort from the French 16-item questionnaire on discomforts in ICU patients (IPREA)1(range from 0 to 100, the lowest possible level of discomfort to the highest). The secondary endpoints were the scores of each item of IPREA. Results: At the end of the 6-month period, taking into account the clustering design, the monthly overall discomfort score was lower in the experimental group (median 16.7, IQR [14.6-18.7], N = 398] in comparison with the control group (median 23.7, IQR [21.6-25.7], N = 360]) ; D = −7.3, p < 0.001. This finding is confirmed after adjustment on age, gender, and IGS2: beta (95 % IC) = −6.6 (4.09;9.04), p < 0.001. For each item of IPREA, except pain, the score was lower in the experimental group in comparison with the control group. Conclusions: The MCP of discomfort reduction decreased discomfort perceived by unselected ICU patients and may pave the way for a new strategy in the management of care for ICU patients. References 1. Kalfon P, Mimoz O, Auquier P, Loundou A, Gauzit R, Lepape A, et al. Development and validation of a questionnaire for quantitative assessment of perceived discomforts in critically ill patients. Intensive Care Med. 2010;36(10):1751–8. Grant acknowledgment Supported by institutional grants from the French 2012 Programme Hospitalier Recherche Clinique National.

1. Kalfon P, Mimoz O, Auquier P, Loundou A, Gauzit R, Lepape A, et al. Development and validation of a questionnaire for quantitative assessment of perceived discomforts in critically ill patients. Intensive Care Med. 2010;36(10):1751–8. Grant acknowledgment Supported by institutional grants from the French 2012 Programme Hospitalier Recherche Clinique National. A925 Evaluation of cognitive and emotional states during multidisciplinary clinical simulation sessions S. Pawar1, T. Jacques2, K. Deshpande2, R. Pusapati2, B. Wood2 St George Hospital, Intensive Care, Sydney, Australia; 2St George Hospital, Sydney, Australia Correspondence: S. Pawar – St George Hospital, Intensive Care, Sydney, Australia Introduction: Simulation is increasingly used in healthcare to teach management of critical and life-threatening situations. The heightened emotions and cognitive overload caused by interaction with high-fidelity simulators may impact on participants' learning ability. We conducted this study to assess emotion during simulation and to explore its association with cognitive load. Design: Prospective observational study. Setting. Intensive Care Unit (ICU), St George Hospital, Sydney.

A925 Evaluation of cognitive and emotional states during multidisciplinary clinical simulation sessions S. Pawar1, T. Jacques2, K. Deshpande2, R. Pusapati2, B. Wood2 St George Hospital, Intensive Care, Sydney, Australia; 2St George Hospital, Sydney, Australia Correspondence: S. Pawar – St George Hospital, Intensive Care, Sydney, Australia Introduction: Simulation is increasingly used in healthcare to teach management of critical and life-threatening situations. The heightened emotions and cognitive overload caused by interaction with high-fidelity simulators may impact on participants' learning ability. We conducted this study to assess emotion during simulation and to explore its association with cognitive load. Design: Prospective observational study. Setting. Intensive Care Unit (ICU), St George Hospital, Sydney. Methods: We enrolled medical and nursing staff with varying levels of ICU experience after obtaining informed consent. We used a standardised clinical scenario and the same human patient simulator for all simulation sessions. The actions of the participants determined the outcome of the scenario. We assessed the emotional state of each participant before and after the completion of the scenario using an eight-item scale containing bipolar oppositional descriptors of emotion.

and the same human patient simulator for all simulation sessions. The actions of the participants determined the outcome of the scenario. We assessed the emotional state of each participant before and after the completion of the scenario using an eight-item scale containing bipolar oppositional descriptors of emotion. 1. We asked the participants to rate their emotion on a 5-point Likert scale with anchors at −2 and +2 representing the bipolar opposites. We assessed the cognitive load of each participant after the completion of the scenario using a validated subjective rating tool2. We asked the participants to rate their cognitive load on a scale of 1 to 9, with higher scores indicating higher mental effort. Results: A total of 103 medical and nursing staff were studied. Compared to pre-simulation, the participants felt more relaxed (−0.28 ± 1.15 vs. 0.14 ± 1, p < 0.005; d = 0.39), excited (0.25 ± 0.89 vs. 0.55 ± 0.92, P < 0.005, d = 0.35) and alert (0.85 ± 0.87 vs. 1.28 ± 0.73, P < 0.00001, d = 0.54) following simulation. There was no difference in the mean scores for the remaining 5 items. The mean cognitive load for all participants was high (6.67 ± 1.41) and was not different in medical and nursing staff. There was no association between pre or post-simulation emotional state and cognitive load. Conclusion: Simulation had a positive effect on the emotional state despite high cognitive load. Further studies are needed to determine their relationship with learning ability. References

Results: A total of 103 medical and nursing staff were studied. Compared to pre-simulation, the participants felt more relaxed (−0.28 ± 1.15 vs. 0.14 ± 1, p < 0.005; d = 0.39), excited (0.25 ± 0.89 vs. 0.55 ± 0.92, P < 0.005, d = 0.35) and alert (0.85 ± 0.87 vs. 1.28 ± 0.73, P < 0.00001, d = 0.54) following simulation. There was no difference in the mean scores for the remaining 5 items. The mean cognitive load for all participants was high (6.67 ± 1.41) and was not different in medical and nursing staff. There was no association between pre or post-simulation emotional state and cognitive load. Conclusion: Simulation had a positive effect on the emotional state despite high cognitive load. Further studies are needed to determine their relationship with learning ability. References 1. Fraser K, Huffman J. The emotional and cognitive impact of unexpected simulated patient death. CHEST 2014; 145(5):958–963. 2. Paas FGWC, Van Merriënboer JJG. The efficiency of instructional conditions: an approach to combine mental effort and performance measures. Hum Factors. 1993; 35 (4): 737–743.

Conclusion: Simulation had a positive effect on the emotional state despite high cognitive load. Further studies are needed to determine their relationship with learning ability. References 1. Fraser K, Huffman J. The emotional and cognitive impact of unexpected simulated patient death. CHEST 2014; 145(5):958–963. 2. Paas FGWC, Van Merriënboer JJG. The efficiency of instructional conditions: an approach to combine mental effort and performance measures. Hum Factors. 1993; 35 (4): 737–743. A926 Quality of life of children 12 months after emergency admission to intensive care R.A. Pulham1, J. Wray1, K. Brown1, C. Pierce2, S. Nadel3, P. Ramnarayan4 1Great Ormond Street Hospital, Critical Care and Cardiorespiratory Division, London, United Kingdom; 2Great Ormond Street Hospital, Paediatric Intensive Care Unit, London, United Kingdom; 3St Mary's Hospital and Charing Cross Hospital Imperial College NHS trust, Paediatric Intensive Care Unit, London, United Kingdom; 4Great Ormond Street Hospital, Children's Acute Transport Service, London, United Kingdom Correspondence: R.A. Pulham – Great Ormond Street Hospital, Critical Care and Cardiorespiratory Division, London, United Kingdom Introduction: Increasing numbers of children are now surviving after an intensive care admission but a proportion of these have significant morbidities. Although there is a growing body of evidence concerning the impact of intensive care stay on children and families, little is known specifically about the long-term status of children who are admitted to intensive care in an emergency.

an intensive care admission but a proportion of these have significant morbidities. Although there is a growing body of evidence concerning the impact of intensive care stay on children and families, little is known specifically about the long-term status of children who are admitted to intensive care in an emergency. Objectives: The aim of this study was to collect data from parents about their child's quality of life 12 months after an emergency intensive care admission. Methods: Eligible children were those retrieved from district hospitals to 2 paediatric intensive care (PICU) units in the UK. Parents were asked to consent to being contacted 12 months after discharge, at which point they were asked to complete the PedsQL, a generic measure of quality of life. The PedsQL enables a total score, physical health summary score and psychosocial health summary score to be calculated, with possible scores ranging from 0–100 and higher scores equating to better quality of life.

er discharge, at which point they were asked to complete the PedsQL, a generic measure of quality of life. The PedsQL enables a total score, physical health summary score and psychosocial health summary score to be calculated, with possible scores ranging from 0–100 and higher scores equating to better quality of life. Results: Parents of 60 children aged 1–16.5 years (median age: 2.1 years; 37 (62 %) males), the majority of whom had had an emergency PICU admission due to sepsis (n = 25, 42 %) or respiratory problems (n = 18; 30 %), completed the PedsQL 12 months after discharge from PICU. For the group overall the total score was 78.07 (SD 20.89), physical health summary score was 81.18 (SD 21.93) and psychosocial health summary score 77.60 (SD 21.63). Babies aged 13–23 months (n = 25) had total scores (m = 85.80, SD =12.39) comparable to those of healthy norms (m = 85.55, SD = 8.74). However older children in all age groups had lower total scores than healthy norms. Whilst 25 % (6/25) of babies had scores of more than one standard deviation below the score of healthy norms, which is recognised as being of clinical significance, this rose to 54 % (7/15) of children aged 5–18 years and 60 % (12/20) of children aged 2–4 years. Of note is that 6 children (30 %) aged 2–4 years had been admitted to PICU for reasons related to trauma or neurological concerns whereas no child aged 13–24 months had been admitted for those reasons.

inical significance, this rose to 54 % (7/15) of children aged 5–18 years and 60 % (12/20) of children aged 2–4 years. Of note is that 6 children (30 %) aged 2–4 years had been admitted to PICU for reasons related to trauma or neurological concerns whereas no child aged 13–24 months had been admitted for those reasons. Conclusions: Children who have had an emergency admission to PICU are at risk for impaired quality of life 12 months after discharge. The risk appears to be greater for children of 2 years and older which is likely to be at least partly attributable to the underlying reason for their admission. Evaluating quality of life outcomes in the longer term after PICU discharge is warranted and identification of potential risk factors will enable interventions to be targeted to optimise outcomes after an emergency admission to PICU. A927 Long-term cognitive outcomes in survivors of critical illness J.R. Azevedo1, W.S. Montenegro1, D.P. Rodrigues2, S.C. Sousa1, V.F. Araujo1, A.L. Leitao1, P.H. Prazeres1, A.V. Mendonca1, M.P. Paula2 1Hospital Sao Domingos, ICU, Sao Luis, Brazil; 2Hospital Sao Domingos, Sao Luis, Brazil Correspondence: J.R. Azevedo – Hospital Sao Domingos, ICU, Sao Luis, Brazil Introduction: Cognitive dysfunction is an important long-term complication of critical illness associated with reduced quality of life, increase in healthcare costs and institutionalization. Delirium, an acute form of brain dysfunction that is common during critical illness has been shown to be associated with long-term cognitive dysfunction(1).

function is an important long-term complication of critical illness associated with reduced quality of life, increase in healthcare costs and institutionalization. Delirium, an acute form of brain dysfunction that is common during critical illness has been shown to be associated with long-term cognitive dysfunction(1). Objectives: The aim of this prospective cohort study was to estimate the prevalence and severity of cognitive dysfunction in survivors of critical illness and to evaluate if delirium duration is an independent determinant of the severity of cognitive dysfunction.

function is an important long-term complication of critical illness associated with reduced quality of life, increase in healthcare costs and institutionalization. Delirium, an acute form of brain dysfunction that is common during critical illness has been shown to be associated with long-term cognitive dysfunction(1). Objectives: The aim of this prospective cohort study was to estimate the prevalence and severity of cognitive dysfunction in survivors of critical illness and to evaluate if delirium duration is an independent determinant of the severity of cognitive dysfunction. Methods: Included were all adult patients admitted to a 45-bed medical surgical ICU over a 12-month period(from March 2014 to February 2015).We excluded patients with preexisting cognitive dysfunction; those that in the evaluation by the psychologist on admission to the ICU had evidence of impaired cognition through the Mini Mental State Examination and patients who could not be reliably assessed for delirium owing to blindness, deafness or language deficit and patients for whom informed consent could not be obtained. After at least 3 months of hospital discharge patients were assessed for cognition using a validated battery of tests including:1)the Digit Span, forward and backward;2) the Rey Auditory Verbal Learning Test (RAVLT);3) the Clock Drawing Test (CDT);4) the Verbal Fluency Test; and the Mini Mental State Examination. We classified patients as having mild or moderate impairment if they had either two cognitive test scores 1.5 standard deviation (SD) below the mean or one cognitive test score 2 SD below the mean; we classified patients as having severe cognitive impairment if they had 3 or more cognitive test scores 1.5 SD below the mean or two or more cognitive test scores 2 SD below the mean.

f they had either two cognitive test scores 1.5 standard deviation (SD) below the mean or one cognitive test score 2 SD below the mean; we classified patients as having severe cognitive impairment if they had 3 or more cognitive test scores 1.5 SD below the mean or two or more cognitive test scores 2 SD below the mean. Results: Enrolled in the clinical trial were 724 patients and 667 patients were eligible for the cohort (Fig. 52). Four hundred and thirteen patients were tested 11 (03–18) months after discharge. Table 29 shows demographic and clinical data of these patients. Cognitive impairment was identified in 206(49.9 %) patients; 120 (29.1 %) had mild or moderate and 86(20 %) severe cognitive dysfunction (Table 30). Eleven(34.3 %) patients with delirium for 3 days or more presented severe cognitive dysfunction. In logistic regression analysis the duration of delirium for 3 days or more was not an independent predictor of cognitive dysfunction(p = 0.76). Conclusions: This investigation in an unselected population of critically ill medical and surgical patients demonstrates that cognitive dysfunction is a frequent and severe long-term complication in survivors of critical illness. On the other hand, unlike other studies we couldn't demonstrate that the duration of delirium is an independent determinant of cognitive impairment. References 1. Pandharipande PP,Girard TD,Jackson JC, et al. Long-term cognitive impairment after critical illness.N Engl J Med 2013;369:1306–1316.Fig. 52 (abstract A927). Enrollement and follow-up Table 29 (abstract A927). Demograhic and Clinical Characteristics

Conclusions: This investigation in an unselected population of critically ill medical and surgical patients demonstrates that cognitive dysfunction is a frequent and severe long-term complication in survivors of critical illness. On the other hand, unlike other studies we couldn't demonstrate that the duration of delirium is an independent determinant of cognitive impairment. References 1. Pandharipande PP,Girard TD,Jackson JC, et al. Long-term cognitive impairment after critical illness.N Engl J Med 2013;369:1306–1316.Fig. 52 (abstract A927). Enrollement and follow-up Table 29 (abstract A927). Demograhic and Clinical Characteristics In - Hospital Cohort (n = 724) Follow-up Cohort (n = 413) Age, Yr Median (IQR) 59 (47–73) 57 (46–72) Male sex n(%) 374 (51.7) 206 (50.1) Education, yr Median (IQR) 11 (11–14) 11 (11–14) APACHE IV Score Median (IQR) 35 (23–53) 32 (21–48) SOFA Score at enrollment Median (IQR) 1 (0–3) 1 (1–3) Delirium N° of Pacients (%) 80 (11.6) 55 (13.3) Delirium N° of days Median (IQR) 4 (2–5) 3 (2–5) Duration of hospital stay, days Median (IQR) 11 (6–23) 10 (5–19) IQR = Interquatile Range; APACHE = Acute Physiology and Chronic Health Evaluation; SOFA = Sequential Organ Failure Assessment; Table 30 (abstract A927). Diagnosis at admission

In - Hospital Cohort (n = 724) Follow-up Cohort (n = 413) Age, Yr Median (IQR) 59 (47–73) 57 (46–72) Male sex n(%) 374 (51.7) 206 (50.1) Education, yr Median (IQR) 11 (11–14) 11 (11–14) APACHE IV Score Median (IQR) 35 (23–53) 32 (21–48) SOFA Score at enrollment Median (IQR) 1 (0–3) 1 (1–3) Delirium N° of Pacients (%) 80 (11.6) 55 (13.3) Delirium N° of days Median (IQR) 4 (2–5) 3 (2–5) Duration of hospital stay, days Median (IQR) 11 (6–23) 10 (5–19) IQR = Interquatile Range; APACHE = Acute Physiology and Chronic Health Evaluation; SOFA = Sequential Organ Failure Assessment; Table 30 (abstract A927). Diagnosis at admission Diagnosis at admission n(%) In-Hospital Cohort (n = 724) Follow-up Cohort (n = 413) AMI, CHF, Arrhythmia 146 (20.2) 98 (23.7) Acute respiratory failurea 109 (15.1) 58 (14.0) Other surgical proceduresb 99 (13.7) 56 (13.6) Neurologic disease 93 (12.8) 53 (12.8) Severe sepsis, septic shock 77 (10.6) 37 (9.0) Digestive Surgery 69 (9.6) 37 (9.0) Digestive disease 67 (9.3) 39 (9.4) Other disease 64 (8.7) 35 (8.5) AMI Acute Myocardial Infarction, CHF Congestive Heart Failure aAcute respiratory failure included acute respiratory distress syndrome, pneumonia, acute exacerbation of chronic pulmonary disease, asthma, pulmonary edema, embolism bOther surgical procedures included orthopedic, vascular, urologic surgery Table 31 (abstract A927). Cognitive outcomes during follow-up Follow-up Asjessment (n = 413) No Impairment n(%) 207 (50.1) Mild/Moderate n(%) 120 (29.1) Severe Impairment n (%) 86 (20)

aAcute respiratory failure included acute respiratory distress syndrome, pneumonia, acute exacerbation of chronic pulmonary disease, asthma, pulmonary edema, embolism bOther surgical procedures included orthopedic, vascular, urologic surgery Table 31 (abstract A927). Cognitive outcomes during follow-up Follow-up Asjessment (n = 413) No Impairment n(%) 207 (50.1) Mild/Moderate n(%) 120 (29.1) Severe Impairment n (%) 86 (20) A928 Predictors of long-term quality of life after ICU discharge: the Argentinian Caviuci study (evaluation of quality of life after ICU in Argentine) A. Das Neves1, C.I. Loudet1, M. Busico2, D. Vazquez3, D. Villalba4, A. Lischinsky5, M. Veronesi6, M. Emmerich7, E. Descotte8, A. Juliarena9, M. Carboni Bisso10, M. Grando11, A. Tapia12, M. Camargo13, D. Villani Ulla14, L. Corzo15, H. Placido dos Santos16, A. Ramos17, J.A. Doglia18, E. Estenssoro1 1HIGA Gral San Martin, La Plata, Argentina; 2Hospital de Trauma F Abete, Municipio Islas Malvinas, Argentina; 3Sanatorio Anchorena, Buenos Aires, Argentina; 4Hospital Municipal Chivilcoy, Chivilcoy, Argentina; 5Instituto De Neurología Cognitiva-INECO, Buenos Aires, Argentina; 6Instituto Fleni, Escobar, Argentina; 7Sanatorio Güemes, Buenos Aires, Argentina; 8Hospital Británico de Buenos Aires, Buenos Aires, Argentina; 9Hospital Austral, Pilar, Argentina; 10Hospital San Antonio, Gualeguay, Argentina; 11Clinica Bazterrica, Buenos Aires, Argentina; 12Hospital Domingo Funes, Santa Maria de Punilla, Argentina; 13Hospital San Luis, San Luis, Argentina; 14HIGA San José, Pergamino, Argentina; 15Hospital Artemides Zatti, Viedma, Argentina; 16Hospital de Gastroenterologia Udaondo, Buenos Aires, Argentina; 17Sanatorio Parque, Rosario, Argentina; 18Hospital Dr. Carlos Macías, Mar de Ajó, Argentina Correspondence: A. Das Neves – HIGA Gral San Martin, La Plata, Argentina Introduction: Long-term impact of critical illness on quality of life has been scarcely studied in Latin-America.

rologia Udaondo, Buenos Aires, Argentina; 17Sanatorio Parque, Rosario, Argentina; 18Hospital Dr. Carlos Macías, Mar de Ajó, Argentina Correspondence: A. Das Neves – HIGA Gral San Martin, La Plata, Argentina Introduction: Long-term impact of critical illness on quality of life has been scarcely studied in Latin-America. Objectives: To describe health-related quality of life (HRQOL) after ICU discharge and to identify its predictors. Methods: The Argentinian Society of Critical Care launched a multicenter cohort study which enrolled patients receiving mechanical ventilation (MV) >72 hr.(3/5-8/31/2014). Epidemiological data and events were collected. Pre-ICU, and at 2, 6 and 12 months after discharge, HRQOL was recorded with the EQ-5D questionnaire (EQ-Index and EQ-EVA) together with relevant physical, neuropsychological and other variables. Values at different time points were compared versus pre ICU time with Wilcoxon signed rank test. We performed longitudinal data analysis to evaluate the effect of time on EQ-Index adjusting for important covariates, building a marginal model with generalized estimating equations. Results: 320 patients from 26 ICUs were included:age 50[29–66], 41 % had Charlson score ≥ 2, 48 % married; 51 % occupied; 65 % from public hospitals; APACHE II 18 ± 7, SOFA24hs 7 ± 4; 51 % medical admissions; 23 % trauma; 51 % shock, 45 % ARDS. 1-year mortality was 22 %. EQ-index dimensions and evolution in time of EQ-index and EQ-EVA are shown in Figs. 53 and 54. Symptoms are displayed in Table 32

score ≥ 2, 48 % married; 51 % occupied; 65 % from public hospitals; APACHE II 18 ± 7, SOFA24hs 7 ± 4; 51 % medical admissions; 23 % trauma; 51 % shock, 45 % ARDS. 1-year mortality was 22 %. EQ-index dimensions and evolution in time of EQ-index and EQ-EVA are shown in Figs. 53 and 54. Symptoms are displayed in Table 32 Positive determinants of the evolution of the EQ-index were time and admission Glasgow score (p 0.032 and 0.005 respectively) while age, duration of MV and weakness were negatively associated (p 0.000, 0.014 and 0.000) ]. EQ-EVA paralleled EQ-index changes. Conclusions: After ICU discharge, patients suffered frequent long-term consequences that negatively affect their HRQOL. Alterations in mobility, daily activities and personal care exhibited the greatest deterioration. Prevalence of pain, anxiety and depression was high even before ICU admission, aggravated after 1-year post-discharge (40 % of patients) Duration of VM was the only intra-ICU variable that affected HRQOL. Pre-ICU conditions as age and the extent of neurological injury and, after ICU, time and weakness, were also independent determinants. Grant acknowledgment The present study was supported by the Argentinian Society of Critical Care (SATI)Fig. 53 (abstract A928). Evolution of dimensions of EuroQol-5D Fig. 54 (abstract A928). EQ-VAS vs EQ Index Table 32 (abstract A928). Follow-up questionnaire

Duration of VM was the only intra-ICU variable that affected HRQOL. Pre-ICU conditions as age and the extent of neurological injury and, after ICU, time and weakness, were also independent determinants. Grant acknowledgment The present study was supported by the Argentinian Society of Critical Care (SATI)Fig. 53 (abstract A928). Evolution of dimensions of EuroQol-5D Fig. 54 (abstract A928). EQ-VAS vs EQ Index Table 32 (abstract A928). Follow-up questionnaire Pre ICU 2 months 6 months 12 months p Caregiver 66 % 50 % 50 % 39 % 6 m vs. 2 m: p ≤ 0.000 12 m vs 2: p ≤ 0.000 Married/partnership 73 % 73 % 75 % 68 % 2 m vs. 0: p = 0.9 6 m vs 0: p = 0.9 12 m vs 0: p = 0.33 Weight 75 ± 18 69 ± 18 72 ± 16 73 ± 17 2 m vs. 0: p ≤ 0.000 6 m vs. 0: p ≤ 0.000 12 m vs 0: p ≤ 0.001 Paresia/paralysis 48 % 36 % 33 % 2 m vs. 6: p ≤ 0.000 2 m vs. 12: p < 0.000 Weakness 53 % 34 % 27 % 6 m vs. 2: p ≤ 0.000 12 m vs. 2: p < 0.000 Mood disorders 44 % 35 % 35 % 6 m vs. 2: p ≤ 0.05 12 m vs 2: p = 0.07 Psychological/Psychiatric therapy 24 % 25 % 24 % 6 vs. 2: p = 0.4 12 vs 2: p = 0.36 Memory problems 40 % 33 % 31 % 6 m vs 2: p = 0.16 12 m vs 2: p = 0.29 Occupation (Works or studies) 54 % 22 % 31 % 44 % Pre UCI vs 2 m: p ≤ 0.000 Pre UCI vs 6 m: p ≤ 0.000 Pre UCI vs 12 m:p ≤ 0.001

m vs. 2: p ≤ 0.05 12 m vs 2: p = 0.07 Psychological/Psychiatric therapy 24 % 25 % 24 % 6 vs. 2: p = 0.4 12 vs 2: p = 0.36 Memory problems 40 % 33 % 31 % 6 m vs 2: p = 0.16 12 m vs 2: p = 0.29 Occupation (Works or studies) 54 % 22 % 31 % 44 % Pre UCI vs 2 m: p ≤ 0.000 Pre UCI vs 6 m: p ≤ 0.000 Pre UCI vs 12 m:p ≤ 0.001 NON-INASIVE NEUROMONITORING A929 Longitudinal whole-brain analysis of white matter integrity in severe traumatic brain injury patients M. Carbonara1, S. Magnoni2, C.L. Mac Donald3, J.S. Shimony4, V. Conte2, F. Triulzi5, F. Stretti6, M. Macrì6, A.Z. Snyder4, N. Stocchetti2,6, D.L. Brody7 1Fondazione IRCCS Cà Granda-Ospedale Maggiore Policlinico, Department of Anaesthesiology and Intensive Care, Milan, Italy; 2Fondazione IRCCS Cà Granda-Ospedale Maggiore Policlinico, Department of Anaesthesiology and Intensive Care, Milano, Italy; 3University of Washington School of Medicine, Department of Neurological Surgery, Seattle, United States; 4Washington University, Mallinckrodt Institute of Radiology, St. Louis, United States; 5Fondazione IRCCS Cà Granda-Ospedale Maggiore Policlinico, Department of Neuroradiology, Milano, Italy; 6Milan University, Milano, Italy, 7Washington University, Department of Neurology, St. Louis, United States Correspondence: M. Carbonara – Fondazione IRCCS Cà Granda-Ospedale Maggiore Policlinico, Department of Anaesthesiology and Intensive Care, Milan, Italy Introduction: Diffuse axonal injury (DAI) is a common event following traumatic brain injury (TBI), which is likely related to worst long term outcome. Diffusion tensor imaging (DTI), a magnetic resonance imaging (MRI) technique that investigates white matter integrity, is recognized as a useful tool to quantify DAI extent in TBI and possibly predict outcome. Few studies explored whole brain longitudinal changes of DTI-derived parameters in single subjects following TBI.

fusion tensor imaging (DTI), a magnetic resonance imaging (MRI) technique that investigates white matter integrity, is recognized as a useful tool to quantify DAI extent in TBI and possibly predict outcome. Few studies explored whole brain longitudinal changes of DTI-derived parameters in single subjects following TBI. Methods: Patients with severe TBI underwent brain MRI including DTI (32 directions, b = 1000, voxel size 2x2x2) 2–3 weeks and 1 year after trauma. 31 age-matched healthy controls underwent the same DTI protocol. We used region of interest (ROI) automated analysis (www.mristudio.org) covering the entire brain to quantify white matter integrity. The ROI fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD) and radial diffusivity (RD) were extracted. Abnormalities were defined as DTI values more than 2 standard deviations below or above the mean values of controls for each ROI. Results: 14 TBI patients with a median age of 34 (IQR 20–38) and a median GCS score of 5 (IQR 4–7) were included. 11 had diffuse injury according to Marshall classification.

Methods: Patients with severe TBI underwent brain MRI including DTI (32 directions, b = 1000, voxel size 2x2x2) 2–3 weeks and 1 year after trauma. 31 age-matched healthy controls underwent the same DTI protocol. We used region of interest (ROI) automated analysis (www.mristudio.org) covering the entire brain to quantify white matter integrity. The ROI fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD) and radial diffusivity (RD) were extracted. Abnormalities were defined as DTI values more than 2 standard deviations below or above the mean values of controls for each ROI. Results: 14 TBI patients with a median age of 34 (IQR 20–38) and a median GCS score of 5 (IQR 4–7) were included. 11 had diffuse injury according to Marshall classification. Regions with increased MD and reduced FA were more than expected in both early and late scan (p < 0.001 binomial test), while AD and RD abnormalities were less common. More than 50 % of the patients had increased MD in the early scan in the fronto-basal girae, corona radiata and thalami; in late scans MD abnormalities were larger and more diffuse, affecting also all frontal and temporal girae and corpus callosum. FA was frequently reduced in the corpus callosum, internal capsule and fronto-basal girae in early scan, while in late phase reductions were similar but more widespread, also including the central girae, cerebellum and inferior longitudinal fascicles. The number of regions with abnormal MD increased over time (p < 0.01 Mann-Whitney), whereas for FA it was not statistically different.

Regions with increased MD and reduced FA were more than expected in both early and late scan (p < 0.001 binomial test), while AD and RD abnormalities were less common. More than 50 % of the patients had increased MD in the early scan in the fronto-basal girae, corona radiata and thalami; in late scans MD abnormalities were larger and more diffuse, affecting also all frontal and temporal girae and corpus callosum. FA was frequently reduced in the corpus callosum, internal capsule and fronto-basal girae in early scan, while in late phase reductions were similar but more widespread, also including the central girae, cerebellum and inferior longitudinal fascicles. The number of regions with abnormal MD increased over time (p < 0.01 Mann-Whitney), whereas for FA it was not statistically different. An inverse correlation between the number of ROI with altered MD at early scan and outcome evaluated with GOSE was found (p < 0.05, Spearman r). Conclusion: The present results indicate that early alterations of mean diffusivity and fractional anisotropy persist or worsen (for MD) at 1 year after TBI, suggesting an ongoing loss of white matter integrity and gliosis. The more frequently affected regions were the frontal girae, corpus callosum, corona radiata, inferior longitudinal fascicles and cerebellum. The number of ROI with early abnormal mean diffusivity is inversely correlated with outcome.

t 1 year after TBI, suggesting an ongoing loss of white matter integrity and gliosis. The more frequently affected regions were the frontal girae, corpus callosum, corona radiata, inferior longitudinal fascicles and cerebellum. The number of ROI with early abnormal mean diffusivity is inversely correlated with outcome. A930 Application of evoked potentials for predicting swallowing disorders in early period after posterior fossa surgery V. Podlepich, V. Shimanskiy, I. Savin, K. Lapteva, A. Chumaev Burdenko Neurosurgery Institute, Moscow, Russian Federation Correspondence: V. Podlepich – Burdenko Neurosurgery Institute, Moscow, Russian Federation Introduction: Neurogenic dysphagia after removal of fossa posterior tumors (FPT) appears in 17 % of cases. The necessary level of upper airway protection depends on the severity of dysphagia. In our neuro-ICU, the clinical screening test (CST) and fibrooptic evaluation of swallowing test (FEST) are performed to determine the severity of swallowing disorders. The CST and FEST, however, require the patient´s collaboration during examination. Objectives: We have evaluated evoked potentials using different modes (SSEP, AEP, TMS) to predict swallowing disorders in early FPT post-operative period in sedated patients.

A930 Application of evoked potentials for predicting swallowing disorders in early period after posterior fossa surgery V. Podlepich, V. Shimanskiy, I. Savin, K. Lapteva, A. Chumaev Burdenko Neurosurgery Institute, Moscow, Russian Federation Correspondence: V. Podlepich – Burdenko Neurosurgery Institute, Moscow, Russian Federation Introduction: Neurogenic dysphagia after removal of fossa posterior tumors (FPT) appears in 17 % of cases. The necessary level of upper airway protection depends on the severity of dysphagia. In our neuro-ICU, the clinical screening test (CST) and fibrooptic evaluation of swallowing test (FEST) are performed to determine the severity of swallowing disorders. The CST and FEST, however, require the patient´s collaboration during examination. Objectives: We have evaluated evoked potentials using different modes (SSEP, AEP, TMS) to predict swallowing disorders in early FPT post-operative period in sedated patients. Methods: The prospective study was carried out during the period from December 2013 to June 2014 and included 132 patients operated for non infiltrative paraxial fossa posterior tumors. All patients underwent SSEP, AEP and TMS the day before operation. After operation, all patients were delivered to ICU intubated and mechanically ventilated. 111 patients demonstrated full recovery from anesthesia with regaining consciousness, passed spontaneous breathing test (SBT) and gained 6 points on CST without deficiency. These patients had none or low level of dysphagia and were successfully extubated after operation. These patients formed 1st group. 21 patients had a neurogenic dysphagia and formed 2nd group. We performed SSEP, AEP and TMS on all patients immediately after admission to ICU.

d gained 6 points on CST without deficiency. These patients had none or low level of dysphagia and were successfully extubated after operation. These patients formed 1st group. 21 patients had a neurogenic dysphagia and formed 2nd group. We performed SSEP, AEP and TMS on all patients immediately after admission to ICU. Results: We revealed no clinical or electrophysiological points that could have predicted neurogenic dysphagia before operation. In our research, we found the EP values which were different for the first group and for the second group. The AEP and TMS data were not informative. We found instrumental the SSEP values that reflected perioperative CCT dynamics, lat p45, amp n13, AUC n11-n13, AUC n13-n18. These SSEP values were used to create a prognostic rule through logistic regression and ROC-curves. As a result, we were able to predict neurogenic dysphagia in early postoperative period in sedated patients with sensitivity 78,6 % and specificity 85 %. Conclusions: Our study allows to predicting swallowing disorders in sedation in patients with paraxial FPT immediately after surgery through SSEP. References 1. Goryachev A.S., Savin I.A., Putsillo M.V.; Scale evaluation and therapeutic strategy in violation of swallowing in patients with damage to the brain stem. Voprosy neyrokhirurgii №4 2006 str. 24–28 (in Russian). 2. Kang De-Zhi, Wu Zan-Yi, Lan; Combined monitoring of evoked potentials during microsurgery for lesions adjacent to the brainstem and intracranial aneurysms Chinese Medical Journal 2007; 120(18):1567–1573

1. Goryachev A.S., Savin I.A., Putsillo M.V.; Scale evaluation and therapeutic strategy in violation of swallowing in patients with damage to the brain stem. Voprosy neyrokhirurgii №4 2006 str. 24–28 (in Russian). 2. Kang De-Zhi, Wu Zan-Yi, Lan; Combined monitoring of evoked potentials during microsurgery for lesions adjacent to the brainstem and intracranial aneurysms Chinese Medical Journal 2007; 120(18):1567–1573 3. Georg Neuloh, M.D., And Johannes Schramm, M.D; Monitoring of motor evoked potentials compared with somatosensory evoked potentials and microvascular Doppler ultrasonography in cerebral aneurysm surgery. J Neurosurg 100:389–399, 2004

2. Kang De-Zhi, Wu Zan-Yi, Lan; Combined monitoring of evoked potentials during microsurgery for lesions adjacent to the brainstem and intracranial aneurysms Chinese Medical Journal 2007; 120(18):1567–1573 3. Georg Neuloh, M.D., And Johannes Schramm, M.D; Monitoring of motor evoked potentials compared with somatosensory evoked potentials and microvascular Doppler ultrasonography in cerebral aneurysm surgery. J Neurosurg 100:389–399, 2004 A931 The revised cerebral recovery index for prediction of neurological outcome after cardiac arrest at the bedside M.C. Tjepkema-Cloostermans1, J. Hofmeijer2,3, A. Beishuizen4, H. Hom4, M.J. Blans5, M.J.A.M. van Putten1,3 1Medisch Spectrum Twente, Clinical Neurophysiology and Neurology, Enschede, Netherlands; 2Rijnstate Hospital, Neurology, Arnhem, Netherlands; 3University of Twente, Clinical Neurophysiology, Enschede, Netherlands; 4Medisch Spectrum Twente, Intensive Care Medicine, Enschede, Netherlands; 5Rijnstate Hospital, Intensive Care Medicine, Arnhem, Netherlands Correspondence: M.C. Tjepkema-Cloostermans – Medisch Spectrum Twente, Clinical Neurophysiology and Neurology, Enschede, Netherlands Introduction: EEG monitoring during the first 24 hours robustly contributes to the prediction of either poor or good outcome in comatose patients after cardiac arrest [1]. Quantitative EEG (qEEG) measures can be useful to visualize evolution of the EEG over hours. We recently proposed the Cerebral Recovery Index (CRI), an index based on a combination of five qEEG measures grading the severity of hypoxic brain damage on a scale from zero to one to facilitate prognostication [2].

rest [1]. Quantitative EEG (qEEG) measures can be useful to visualize evolution of the EEG over hours. We recently proposed the Cerebral Recovery Index (CRI), an index based on a combination of five qEEG measures grading the severity of hypoxic brain damage on a scale from zero to one to facilitate prognostication [2]. Objectives: To evaluate the prognostic accuracy of a revised CRI, after optimalization by the use of a random forest classifier instead of a manually chosen feature combination and the addition of four qEEG measures, resuscitation parameters and patient characteristics. Methods: In this prospective cohort study, 283 consecutive comatose patients after cardiac arrest were included in two intensive care units. Continuous EEG was recorded during the first three days. Outcome at 6 months was dichotomized as good (CPC 1–2) or poor (CPC 3–5). Nine qEEG measures were extracted: alpha to delta ratio, signal power, Shannon entropy, delta coherence, regularity, the number of burst/min, mean and max burst correlation, and fraction of burst correlation >0.8. These measures were combined with patient characteristics and resuscitation data, including sex, age, initial heart rhythm, in- versus out-of-hospital-cardiac-arrest, and presumed cause of cardiac arrest.

rence, regularity, the number of burst/min, mean and max burst correlation, and fraction of burst correlation >0.8. These measures were combined with patient characteristics and resuscitation data, including sex, age, initial heart rhythm, in- versus out-of-hospital-cardiac-arrest, and presumed cause of cardiac arrest. Patients were randomly divided over a training and a validation set of respectively 143 and 140 patients. Within the training set, a random forest classifier was fitted for each hour after cardiac arrest. Based on results in the test set, two thresholds were chosen: one for predicting poor neurological outcome and one for predicting good neurological outcome. Subsequently, the revised CRI was evaluated in the validation set. Results: Poor outcome could reliable be predicted with the revised CRI (with 100 % specificity) in the validation set with a sensitivity of 67 and 37 % at respectively 12 and 24 hours after cardiac arrest. Good neurological outcome could be predicted with a sensitivity of 60 and 62 % at a specificity of 94 and 93 %.

Patients were randomly divided over a training and a validation set of respectively 143 and 140 patients. Within the training set, a random forest classifier was fitted for each hour after cardiac arrest. Based on results in the test set, two thresholds were chosen: one for predicting poor neurological outcome and one for predicting good neurological outcome. Subsequently, the revised CRI was evaluated in the validation set. Results: Poor outcome could reliable be predicted with the revised CRI (with 100 % specificity) in the validation set with a sensitivity of 67 and 37 % at respectively 12 and 24 hours after cardiac arrest. Good neurological outcome could be predicted with a sensitivity of 60 and 62 % at a specificity of 94 and 93 %. Conclusions: Here we show that a combination of qEEG and clinical measures, extracted and combined by a random forest classifier, provides reliable, objective prognostic information. This revised CRI can be used for the prediction of both poor and good neurological outcome, thereby poor outcome can be reliable predicted (without false positives) with relatively high sensitivity. The revised CRI is expressed as a single index between 0 and 1, which can be used in real time at the bedside, even by professionals who are not trained in EEG interpretation. References [1] Hofmeijer et al., Neurology, 2015 [2] Tjepkema-Cloostermans et al., Crit Care, 2013Table 33 (abstract A931). Sensitivity and specificity of the revised CRI

Conclusions: Here we show that a combination of qEEG and clinical measures, extracted and combined by a random forest classifier, provides reliable, objective prognostic information. This revised CRI can be used for the prediction of both poor and good neurological outcome, thereby poor outcome can be reliable predicted (without false positives) with relatively high sensitivity. The revised CRI is expressed as a single index between 0 and 1, which can be used in real time at the bedside, even by professionals who are not trained in EEG interpretation. References [1] Hofmeijer et al., Neurology, 2015 [2] Tjepkema-Cloostermans et al., Crit Care, 2013Table 33 (abstract A931). Sensitivity and specificity of the revised CRI Predicting poor outcome Predicting good outcome Sensitivity Specificity Sensitivity Specificity At 12 h after CA 0.67 1.00 0.60 0.94 At 24 h after CA 0.37 1.00 0.62 0.93

[1] Hofmeijer et al., Neurology, 2015 [2] Tjepkema-Cloostermans et al., Crit Care, 2013Table 33 (abstract A931). Sensitivity and specificity of the revised CRI Predicting poor outcome Predicting good outcome Sensitivity Specificity Sensitivity Specificity At 12 h after CA 0.67 1.00 0.60 0.94 At 24 h after CA 0.37 1.00 0.62 0.93 A932 From observation to monitoring: implementation of quantitative electroencephalography (QEEG) in neurointensive care L. Longhi1, B. Frigeni2, M. Curinga3, D. Mingone4, S. Beretta5, A. Patruno6, L. Gandini6, A. Vargiolu6, F. Ferri1, R. Ceriani1, M.R. Rottoli2, L. Lorini1, G. Citerio4,6 1Azienda Socio Sanitaria Territoriale Papa Giovanni XXIII, Anesthesia and Critical Care Medicine, Bergamo, Italy; 2Azienda Socio Sanitaria Territoriale Papa Giovanni XXIII, Neurology, Bergamo, Italy; 3University of Milan, Milano, Italy; 4University of Milan - Bicocca, Monza, Italy; 5Ospedale San Gerardo, Neurology, Monza, Italy; 6Ospedale San Gerardo, Anesthesia and Critical Care Medicine, Monza, Italy Correspondence: L. Longhi – Azienda Socio Sanitaria Territoriale Papa Giovanni XXIII, Anesthesia and Critical Care Medicine, Bergamo, Italy Introduction: Continuous electroencephalography (cEEG) allows real-time monitoring critically-ill patients neurophysiology and to detect non-clinical seizures in comatose patients, delayed cerebral ischemia after subarachnoid haemorrhage, and guide therapies for status epilepticus. The application of cEEG is still limited because it requires awkward analysis by experienced neurophysiologists of huge amount of EEG tracings. Quantitative EEG (qEEG) techniques, i.e.amplitude integrated EEG (aEEG) and Density Spectra Array (DSA), have been developed to simplify the complexity of EEG interpretation, to allow rapid evaluation of cerebral background electrical activity and the power spectrum of the EEG frequencies derived from raw data EEG. These developments offer the potentiality to transform an instrument interpreted by neurophysiologist afterwards in a monitoring tool useful to ICU staff.

of EEG interpretation, to allow rapid evaluation of cerebral background electrical activity and the power spectrum of the EEG frequencies derived from raw data EEG. These developments offer the potentiality to transform an instrument interpreted by neurophysiologist afterwards in a monitoring tool useful to ICU staff. Objectives: To test the hypothesis that EEG-nonexpert neurointensivists can obtain real-time reliable information from qEEG after training under the supervision of an in-house neurophysiologist. To describe the implementation of qEEG monitoring in 2 neurointensive care units. Methods: The implementation occurred in sequential phases. cEEG was recorded using 8 surface electrodes according to the international 10–20 system, on a bipolar longitudinal montage in patients with brain injury. qEEG-naïve neurointensivists, after a short training from a neurophysiologist followed by daily supervision for the study period, were subjected to a baseline test evaluating aEEG and DSA traces. Each panel consisted of raw EEG data and 3 qEEG tools: the color density spectral array (DSA), amplitude integrated EEG (aEEG) and the burst suppression rate (BSR). After this evaluation, daily qEEG evaluation was performed by the neurointensivists and reviewed by the neurophysiologist.

evaluating aEEG and DSA traces. Each panel consisted of raw EEG data and 3 qEEG tools: the color density spectral array (DSA), amplitude integrated EEG (aEEG) and the burst suppression rate (BSR). After this evaluation, daily qEEG evaluation was performed by the neurointensivists and reviewed by the neurophysiologist. Results: From July 2015 to April 2016 we monitored 77 patients (57 ± 17 years, 34 male) admitted for brain trauma (41 %), stroke (7 %), intracerebral hemorrhage (19 %), subarachnoid hemorrhage (18 %) and other neurological conditions (15 %). Median admission motor Glasgow coma scale (GCS) was 5 (range 1–6). Indications for cEEG were: unexplained clinical status (69 %), status epilepticus management (17 %) and titrating sedation for ICP control (14 %). cEEG monitoring was done for a median time of 2 days (range 1–8). EEG seizures were detected in 17 % of 45 patients with unexplained clinical status despite they received sedation with Propofol and/or antiseizures prophylaxis. qEEG monitoring guided patients management in 68 % of cases. Conclusions: qEEG monitoring can be efficiently implemented as a real-time monitoring tool by EEG-nonexpert neurointensivists after proper training and under the supervision of a dedicated neurophysiologist. Further work is on-going to evaluate the learning curve of qEEG by naive neurointensivists.

Results: From July 2015 to April 2016 we monitored 77 patients (57 ± 17 years, 34 male) admitted for brain trauma (41 %), stroke (7 %), intracerebral hemorrhage (19 %), subarachnoid hemorrhage (18 %) and other neurological conditions (15 %). Median admission motor Glasgow coma scale (GCS) was 5 (range 1–6). Indications for cEEG were: unexplained clinical status (69 %), status epilepticus management (17 %) and titrating sedation for ICP control (14 %). cEEG monitoring was done for a median time of 2 days (range 1–8). EEG seizures were detected in 17 % of 45 patients with unexplained clinical status despite they received sedation with Propofol and/or antiseizures prophylaxis. qEEG monitoring guided patients management in 68 % of cases. Conclusions: qEEG monitoring can be efficiently implemented as a real-time monitoring tool by EEG-nonexpert neurointensivists after proper training and under the supervision of a dedicated neurophysiologist. Further work is on-going to evaluate the learning curve of qEEG by naive neurointensivists. A933 Early suspect of cerebral vasospasm using transcranial color coded sonography monitoring in patients with subarachnoid haemorrhage S. Pifferi1, M. Battistini1, V. Cordolcini1, A. Agarossi2, R. Di Rosso2, F. Ortolano1, N. Stocchetti1,2 1Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Neuroscience Intensive Care Unit, Milan, Italy; 2Università degli Studi di Milano, Milan, Italy Correspondence: S. Pifferi – Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Neuroscience Intensive Care Unit, Milan, Italy Introduction: Cerebral vasospasm (VSP) is the narrowing of cerebral vessels. It occurs in up to 67 % of patients with aneurysmal subarachnoid haemorrhage (aSAH) and contributes to delayed cerebral ischemia (DCI) and neurological deterioration (symptomatic VSP = sVSP) in 30 % of patients. Transcranial color coded sonography (TCCS) is a non-invasive bedside technique that allows the measure of mean blood flow velocity (mFV) of cerebral arteries. The range of normality is 55–85 cm/sec, whereas mFV >120 cm/s is currently used as a threshold to define TCCS vasospasm (1).

tic VSP = sVSP) in 30 % of patients. Transcranial color coded sonography (TCCS) is a non-invasive bedside technique that allows the measure of mean blood flow velocity (mFV) of cerebral arteries. The range of normality is 55–85 cm/sec, whereas mFV >120 cm/s is currently used as a threshold to define TCCS vasospasm (1). Objectives: To analyse TCCS mFV variation over time in patients with aSAH and evaluate the prognostic utility of a daily TCCS monitoring. Methods: All consecutive patients admitted to our ICU after aSAH were enrolled in the study. Mean blood flow velocity (mFV) of middle cerebral arteries on either side was obtained every 1–2 day for 14 days after the bleeding. We defined sVSP as a new neurological deterioration associated to radiological VSP evaluated with MRI angiography. Results: 80 patients (age 26–80 years old) were included; 714 TCCS exams were performed. Only 13 patients (16 %) had normal TCCS mFV throughout the study period. 67 patients (84 %) presented mFV increase over time, with the highest values between 8 and 12 days after the bleeding (median 9, IQR 6–11). Among these, 49 patients (61 %) had mFV ≥120 cm/s and 21 (26 %) had mFV ≥200 cm/s. Symptomatic VSP occurred in 20 patients (25 %), the median time of neurological deterioration was 7 days (IQR 5–10) after the bleeding. Patients who developed sVSP (sVSP yes) displayed significantly higher mFV starting from day 5 up to day 14 compared to patients without sVSP (sVSP no); the median time of TCCS vasospasm (mFV >120 cm/s) was 4 days (IQR 3–6).

(25 %), the median time of neurological deterioration was 7 days (IQR 5–10) after the bleeding. Patients who developed sVSP (sVSP yes) displayed significantly higher mFV starting from day 5 up to day 14 compared to patients without sVSP (sVSP no); the median time of TCCS vasospasm (mFV >120 cm/s) was 4 days (IQR 3–6). Accuracy of TCCS in diagnosing sVSP is displayed in the Table 34. Conclusions: MeanFV time variation is significantly different in patients with or without sVSP. Daily TCCS monitoring with a threshold of 120 cm/s permits to identify 95 % of patients who will have symptomatic VSP but specificity is low. Patients at greater risk of sVSP are those with early mFV increase (>120 cm/s within 9 days), rapid mFV increase (>50 cm/s in 48 hours) or higher mFV absolute values (>200 cm/sec). References 1) Zipper S et al. Clinical application of transcranial colour-coded duplex sonography - a review. Eur J Neurology 2002,9:1 ± 8Fig. 55 (abstract A933). mFV time variation in the middle cerebral arteries Table 34 (abstract A933). TCCS accuracy in predicting symptomatic vasospasm TCCS thresholds SENSITIVITY (CI 95 %) SPECIFICITY (CI 95 %) AUC mFV >120 cm/s 95 (75–100) 50 (37–63) 0.90 mFV >120 cm/s (day 4–6) 70 (48–88) 68 (54–79) 0.80 mFV >120 cm/s (day 7–9) 85 (62–97) 70 (57–81) 0.86 mFV >200 cm/s 75 (51–91) 90 (79–96) 0.90 ΔmFV >50 cm/s in 48 hrs 90 (68–99) 63 (50–75) 0.90

Table 34 (abstract A933). TCCS accuracy in predicting symptomatic vasospasm TCCS thresholds SENSITIVITY (CI 95 %) SPECIFICITY (CI 95 %) AUC mFV >120 cm/s 95 (75–100) 50 (37–63) 0.90 mFV >120 cm/s (day 4–6) 70 (48–88) 68 (54–79) 0.80 mFV >120 cm/s (day 7–9) 85 (62–97) 70 (57–81) 0.86 mFV >200 cm/s 75 (51–91) 90 (79–96) 0.90 ΔmFV >50 cm/s in 48 hrs 90 (68–99) 63 (50–75) 0.90 DIFFERENT ASPECTS ON OUTCOME IN THE ICU A934 Development and validation of a model to evaluate the application of human resources practices of high commitment, guided to the quality of service from the perspective of ICU healthcare staff C. Mora Lourido1, J.L. Santana Cabrera1, J.D. Martín Santana2, L. Melián Alzola2, C. García del Rosario1, H. Rodríguez Pérez1, R. Lorenzo Torrent1 1Hospital Insular Las Palmas GC, Las Palmas de Gran Canaria, Spain; 2Hospital Insular Las Palmas GC, Department of Economy, University of Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain Correspondence: H. Rodríguez Pérez – Hospital Insular Las Palmas GC, Las Palmas de Gran Canaria, Spain Introduction: It has been noted the importance of job satisfaction in healthcare services and the consequences resulting therefrom, such as increasing the quality of care services provided and satisfaction of their users. Objectives: To develop a model of influence of human resource management directed to the quality management and organizational excellence in the organizational results, from the perspective of healthcare staff.

DIFFERENT ASPECTS ON OUTCOME IN THE ICU A934 Development and validation of a model to evaluate the application of human resources practices of high commitment, guided to the quality of service from the perspective of ICU healthcare staff C. Mora Lourido1, J.L. Santana Cabrera1, J.D. Martín Santana2, L. Melián Alzola2, C. García del Rosario1, H. Rodríguez Pérez1, R. Lorenzo Torrent1 1Hospital Insular Las Palmas GC, Las Palmas de Gran Canaria, Spain; 2Hospital Insular Las Palmas GC, Department of Economy, University of Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain Correspondence: H. Rodríguez Pérez – Hospital Insular Las Palmas GC, Las Palmas de Gran Canaria, Spain Introduction: It has been noted the importance of job satisfaction in healthcare services and the consequences resulting therefrom, such as increasing the quality of care services provided and satisfaction of their users. Objectives: To develop a model of influence of human resource management directed to the quality management and organizational excellence in the organizational results, from the perspective of healthcare staff. Methods: We carried out a research study, of a transversal nature, whose study population were a total of 248 (12,9 % physicians, 49,2 % nurses and 37,9 % nurse assistants) ICU staff. A personal questionnaire was used to measure, through Likert scales of 7 points, the application of human resource practices of high commitment (HR), the quality of service provided to the patient (QUALITY), the satisfaction with the capacity of the service (CAPACITY), the personal satisfaction with the work done (SATISFACTION) and the affective commitment with the organization (COMMITMENT).

oints, the application of human resource practices of high commitment (HR), the quality of service provided to the patient (QUALITY), the satisfaction with the capacity of the service (CAPACITY), the personal satisfaction with the work done (SATISFACTION) and the affective commitment with the organization (COMMITMENT). Results: The measure models of these five constructs were validated by confirmatory factorial analysis, whose results were satisfactory. The measurement model of HR is a second order construct which is formed by 4 dimensions (training, participation, recognition, and communication). The QUALITY model is a model of second order formed by six dimensions (personal attention, ability of response, waiting room, tangibility, sensory experience, and meals). The CAPACITY model is a construct of second order formed by six dimensions referred to the satisfaction of different aspects (information, communication, facilities, additional services, decision maker's staff and non-decision maker's staff). The results of the PATH model indicate that it is an excellent model (CFI > 0,95 and RMFEA < 0,08). From the causal relationships of this model we can tell that (1) the HR influences positively and significantly on the QUALITY; (2) the HR and the QUALITY influence positively and significantly on the CAPACITY; (3) the RH and QUALITY influence on the SATISFACTION and (4) the CAPACITY and SATISFACTION influence positively and significantly on the COMMITMENT.

odel we can tell that (1) the HR influences positively and significantly on the QUALITY; (2) the HR and the QUALITY influence positively and significantly on the CAPACITY; (3) the RH and QUALITY influence on the SATISFACTION and (4) the CAPACITY and SATISFACTION influence positively and significantly on the COMMITMENT. Conclusions: The results of this model indicates that the implementation of a HR strategy supported by a quality management significantly influences on the performance of the ICUs in terms of staff satisfaction, just as with the personal satisfaction and the commitment of healthcare workers. References 1. Carrillo García C, Martínez Roche M, Gómez García CI y Meseguer de Pedro M. Satisfacción laboral de los profesionales sanitarios de un Hospital Universitario: análisis general y categorías laborales. Anales de Psicología 2015; 31 (2):645–650. 2. Newman K, Maylor U, Chansarkar B. “The nurse satisfaction, service quality and nurse retention chain: Implications for management of recruitment and retention. Journal of Management in Medicine 2002; 16 (4/5):271–291.

1. Carrillo García C, Martínez Roche M, Gómez García CI y Meseguer de Pedro M. Satisfacción laboral de los profesionales sanitarios de un Hospital Universitario: análisis general y categorías laborales. Anales de Psicología 2015; 31 (2):645–650. 2. Newman K, Maylor U, Chansarkar B. “The nurse satisfaction, service quality and nurse retention chain: Implications for management of recruitment and retention. Journal of Management in Medicine 2002; 16 (4/5):271–291. A935 Sequential implementation changes affect glucose regulation in the ICU S. Eslami1,2, A. Dalhuisen3, T. Fiks3, M.J. Schultz4, A. Abu Hanna1, P.E. Spronk3,4 1Academic Medical Center, University of Amsterdam, Medical Informatics, Amsterdam, Netherlands; 2Mashhad University of Medical Sciences, Pharmaceutical Research Center, Tehran, Islamic Republic of Iran; 3Gelre Hospitals Apeldoorn, Intensive Care Medicine, Apeldoorn, Netherlands; 4Academic Medical Center, University of Amsterdam, Intensive Care Medicine, Amsterdam, Netherlands Correspondence: P.E. Spronk – Academic Medical Center, University of Amsterdam, Medical Informatics, Amsterdam, Netherlands Introduction: In the period between 2007–2010, a successful implementation project was finished aimed at strict blood glucose level (BGL) regulation in the Intensive Care Unit (ICU) [1]. We hypothesized that glucose control would afterwards slack and that implementing other measures to modify behavior would be required to regain adequate glucose control.

2007–2010, a successful implementation project was finished aimed at strict blood glucose level (BGL) regulation in the Intensive Care Unit (ICU) [1]. We hypothesized that glucose control would afterwards slack and that implementing other measures to modify behavior would be required to regain adequate glucose control. Methods: A prospective study was performed in a 12-bed mixed medical-surgical ICU of a university affiliated teaching hospital. All BGL values were extracted from the ICU database in 4 years following the implementation project until December 2014. Following the project, BGL targets were set at a range of 80–135 mg/dl, Nurses' instructions for keeping BGL values in target were not changed. After 3.5 years, an automated warning system was implemented in the patient data management system that triggered a centrally placed monitor with feedback about the need for obtaining a BGL value, based on the actual value compared to the previous one. The primary outcome measure was mean BGL. Secondary endpoints were sampling frequency, BGL within pre-defined targets, incidences of severe hypoglycemia, and hyperglycemia. The analysis was restricted to patients with at least two blood glucose measurements. These indicators were analyzed over the course of time using the XMR control chart, a tool belonging to Statistical Process Control.

ling frequency, BGL within pre-defined targets, incidences of severe hypoglycemia, and hyperglycemia. The analysis was restricted to patients with at least two blood glucose measurements. These indicators were analyzed over the course of time using the XMR control chart, a tool belonging to Statistical Process Control. Results: Data of 3760 patient admissions were evaluated, which corresponded to 117,080 BGL measurements. The BGL sampling interval (Figure 56), mean BGL and percentage of severe hypoglycemia all increased after introducing nurses' instruction and decreased significantly after monitoring feedback (p < 0.05). Percentage of severe hypoglycemia events, which is associated tosafety, decreased with some delay after nurses' instruction and remained unchanged (0.18 % on average) and stable after introducing monitoring feedback. Percentage of “in range” measurements of both normoglycemia (80–110) and protocol recommended (80–135) decreased after nurses' instructions and then increased after feedback monitoring. Mean of per patient's standard deviation as a measurement of variability remained unchanged and stable after nurses' instruction and even decreased after monitoring feedback. Conclusion: even after successful implementation of a BGL control system, behavior changed within 3 months with inherent worsening of BGL control. An automated warning monitor in a central location was able to restore BGL control in the ICU. References 1. Schultz MJ et al. Minerva Anestesiol 2012; 78:982–995Fig. 56 (abstract A935). ᅟ

Conclusion: even after successful implementation of a BGL control system, behavior changed within 3 months with inherent worsening of BGL control. An automated warning monitor in a central location was able to restore BGL control in the ICU. References 1. Schultz MJ et al. Minerva Anestesiol 2012; 78:982–995Fig. 56 (abstract A935). ᅟ A936 Using robotic technology to precisely define the neurocognitive phenotype of ICU survivors M. Wood1, D. Maslove2, J. Muscedere2, S.H. Scott1, T. Saha3, A. Hamilton4, D. Petsikas4, D. Payne4, J.G. Boyd1,5 1Queen's University, Centre for Neuroscience Studies, Kingston, Canada; 2Queen's University, Medicine and Critical Care, Kingston, Canada; 3Queen's University, Anaesthesia and Perioperative Medicine, Kingston, Canada; 4Queen's University, Surgery, Kingston, Canada; 5Queen's University, Medicine (Neurology) and Critical Care, Kingston, Canada Correspondence: J.G. Boyd – Queen’s University, Centre for Neuroscience Studies, Kingston, Canada Introduction: Long-term cognitive impairment is common among ICU survivors, and is an important component of the post-ICU syndrome. There is mounting evidence that conventional screening tests are insufficient to capture the degree of cognitive dysfunction experienced by ICU survivors. The KINARM robot provides robust, objective, and quantitative metrics of the sensorimotor and neurocognitive impairments in humans. It has been used to describe impairments post stroke and importantly, deficits identified with the KINARM correlate well with reductions in quality of life.

n experienced by ICU survivors. The KINARM robot provides robust, objective, and quantitative metrics of the sensorimotor and neurocognitive impairments in humans. It has been used to describe impairments post stroke and importantly, deficits identified with the KINARM correlate well with reductions in quality of life. Objective: The overall objective of this research program is to use the KINARM to define the neurocognitive phenotype of ICU survivors (i.e. required invasive mechanical ventilation and/or vasoactive agents for hemodynamic support). This group is compared to healthy age- and gender-matched controls, as well as 3 active control groups. These active control groups were patients 1) pre- and 2) post-cardiac surgery, and 3) patients post-cardiac arrest. Methods: Participants performed 7 tasks on the KINARM that ranged from simple sensorimotor tasks to more complex executive tasks. For each task, 6–12 performance metrics were recorded. These metrics were compared to a normative database of age- and gender-matched controls and z-scores were generated. A composite score for each task was generated using a score derived from Maholanobis distance, with increasing scores representing worse performance. Cluster analysis was applied to these performance metrics using Euclidian distance.

a normative database of age- and gender-matched controls and z-scores were generated. A composite score for each task was generated using a score derived from Maholanobis distance, with increasing scores representing worse performance. Cluster analysis was applied to these performance metrics using Euclidian distance. Results: From February 2014-February 2016, we have performed 41 KINARM assessments (12 pre-cardiac surgery, 12 post-cardiac surgery, 10 post critical illness, 7 post cardiac arrest). Compared to all groups, survivors of critical illness performed worse on several tasks that assess the visuomotor control of the upper limb, interlimb coordination, proprioception, and visuospatial processing. Impairments in executive tasks were noted across all participants. While the KINARM tasks tended to cluster together (i.e. visuospatial, executive), the patient cohorts did not. Conclusions: Robotic technology can provide precise and quantitative metrics of the sensorimotor and cognitive deficits after critical illness. These strategies may be able to define a specific neurocognitive phenotype of the ICU survivor. Grant Acknowledgement This work is supported by Physician Services Incorporated and the Southeastern Ontario Medical Associations' New Clinician-Scientist Program.