Browse the corpus

Introduction A growing body of evidence shows that a more engaged patient experiences better health outcomes and lower use of healthcare services. The term ‘patient engagement’ encompasses a number of related concepts, including ‘patient-centred care’1 and ‘shared decision-making’,2 all of which build on the idea of involving patients as partners in their care. Under this broad umbrella term, there is evidence that patient engagement is associated with fewer adverse events,3 better patient self-management,4 5 fewer diagnostic tests,6 decreased use of healthcare services7 and shorter lengths of stay in hospitals.8 There is also evidence that patients benefit when family members play an active part in the patient's care. For example, one study found that family members gave new information 46% of the time when present during rounds.9 Family members can take on many roles, such as participating in care coordination and assessing care practices for consistency, accuracy and safety.10 They play an especially important role when patients are not physically or cognitively able to participate in their own care, and family members become the surrogate decision-makers.11 Further, organisational policies supporting family involvement, such as extended visiting hours, may affect health outcomes; one study reported that longer visiting hours in the intensive care unit were linked to a reduction in cardiovascular complications, possibly through patients’ reduced anxiety and better hormonal profiles.12

anisational policies supporting family involvement, such as extended visiting hours, may affect health outcomes; one study reported that longer visiting hours in the intensive care unit were linked to a reduction in cardiovascular complications, possibly through patients’ reduced anxiety and better hormonal profiles.12 While continuing to grow, the evidence is not entirely positive; for example, one study found that patient-centred communication was associated with longer visit length.6 In addition, a focus group study found that some people would like their providers to tell them what to do rather than engage in shared decision-making with them.13

g to grow, the evidence is not entirely positive; for example, one study found that patient-centred communication was associated with longer visit length.6 In addition, a focus group study found that some people would like their providers to tell them what to do rather than engage in shared decision-making with them.13 Multiple mechanisms have been proposed to link patient and family engagement (PFE) to better outcomes. One hypothesis is that when clinicians engage patients and their family members as active partners in their care, the patients and family members can provide information missing from medical charts and can recognise and speak up about errors in care delivery.9 Another is that when patients help make the decisions about which treatment options would be the most favourable, they are more likely to follow the selected treatment plans.14 Additionally, when patients can better communicate their questions and concerns, they may be better able to understand, and more likely to adhere to, their treatment protocols.4 15 It is also possible that when clinicians give patients the confidence and skills to manage their conditions, patients will play more effective roles in self-care.16 Furthermore, when patients and family members serve as advisors to hospitals, they can help improve the experiences of patients and families in ways that lead to favourable outcomes.17 The comparative importance of these proposed mechanisms is unknown, and there is currently limited or conflicting evidence to support them.

hermore, when patients and family members serve as advisors to hospitals, they can help improve the experiences of patients and families in ways that lead to favourable outcomes.17 The comparative importance of these proposed mechanisms is unknown, and there is currently limited or conflicting evidence to support them. However, regardless of the mixed evidence regarding specific mechanisms and outcomes, there is growing agreement that effective PFE is foundational to improving patient experience and clinical outcomes and decreasing use of unnecessary healthcare services. This agreement has led to an increased emphasis on and expansion of PFE practices by healthcare leaders and providers. This expansion has been reinforced by financial incentives for hospitals. As part of the Patient Protection and Affordable Care Act, Medicare payments to hospitals are affected by Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) scores.18 HCAHPS includes some items that directly relate to patient engagement, specifically those regarding doctor and nurse communication, and evidence links greater patient engagement to higher patient ratings of hospital quality.3

ffected by Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) scores.18 HCAHPS includes some items that directly relate to patient engagement, specifically those regarding doctor and nurse communication, and evidence links greater patient engagement to higher patient ratings of hospital quality.3 In this study, we build on the framework proposed by Carman et al19 and define PFE as (1) a set of beliefs and behaviours by patients, family members and health professionals; (2) features of organisational design and procedures; and (3) a set of organisational policies, all designed to ensure the inclusion of patients and their family members as active members of the healthcare team as well as encourage collaborative partnerships with providers, patients and their families. Given the evidence that PFE is associated with better health outcomes, higher patient ratings of hospital quality and lower use of healthcare services, it is surprising how little is known about actual practices and policies of hospitals in the USA.

ollaborative partnerships with providers, patients and their families. Given the evidence that PFE is associated with better health outcomes, higher patient ratings of hospital quality and lower use of healthcare services, it is surprising how little is known about actual practices and policies of hospitals in the USA. Thus, we undertook a national survey of hospitals about their use of a range of recommended strategies, including patient and family advisory councils, online access to medical records, health education materials in other languages, 24 h visitation policies, nurse shift-change reports at patients’ bedsides, decision aids, and physician and nurse training in patient engagement. We also surveyed hospitals about their perceived barriers to adopting PFE practices. The purpose of this project was twofold: (1) to describe the degree to which a core set of recommended PFE practices is currently being used in a sample of US hospitals; and (2) to determine the relative significance of perceived barriers to PFE. Methods Overview We conducted a cross-sectional survey of a random sample of hospitals in the USA. The survey was developed in conjunction with a panel of experts on PFE and mailed to 3500 randomly selected hospitals. We report on the survey results regarding adoption of PFE practices and the barriers to PFE.

Thus, we undertook a national survey of hospitals about their use of a range of recommended strategies, including patient and family advisory councils, online access to medical records, health education materials in other languages, 24 h visitation policies, nurse shift-change reports at patients’ bedsides, decision aids, and physician and nurse training in patient engagement. We also surveyed hospitals about their perceived barriers to adopting PFE practices. The purpose of this project was twofold: (1) to describe the degree to which a core set of recommended PFE practices is currently being used in a sample of US hospitals; and (2) to determine the relative significance of perceived barriers to PFE. Methods Overview We conducted a cross-sectional survey of a random sample of hospitals in the USA. The survey was developed in conjunction with a panel of experts on PFE and mailed to 3500 randomly selected hospitals. We report on the survey results regarding adoption of PFE practices and the barriers to PFE. Survey To develop the questionnaire, we reviewed the literature and identified a number of PFE strategies to be assessed, and searched the literature and the internet for existing surveys of PFE. We then convened a technical expert panel (TEP) of 12 external experts with experience or expertise in providing patient-centred care; implementing patient-centred models of care; patient safety awareness; organisational development and management; healthcare quality research; and PFE in planning, delivery and evaluation of healthcare services (see online supplementary appendix A for list of experts). We presented the TEP with a draft survey; suggestions from the TEP were incorporated into a revision, which was then shared again with the TEP; the process repeated for multiple iterations. The final instrument was comprised of 38 items, which measured PFE on 20 distinct topics (see online supplementary appendix B). Items included 37 questions on the implementation of PFE strategies and one multiple-response item concerning perceived barriers to additional implementation.

repeated for multiple iterations. The final instrument was comprised of 38 items, which measured PFE on 20 distinct topics (see online supplementary appendix B). Items included 37 questions on the implementation of PFE strategies and one multiple-response item concerning perceived barriers to additional implementation. Sampling Surveyed hospitals were chosen at random from the 2012 American Hospital Association (AHA) Annual Survey.20 We excluded hospitals that were not acute care hospitals, which left 5290 hospitals. We calculated that to detect a difference of 10% between a subset of 25% of the hospitals reporting and the remaining 75%, we would need 1028 responses. Our initial sample was based on an anticipated 50% response rate. The sampling was done in two stages; an initial sample of 2000 hospitals was surveyed in July 2013. Because of a lower-than-expected response rate, this was followed by a second sample of 1500 in October 2013. The two samples were pooled for analysis. Because of closures, mergers and inaccurate mailing information in the AHA survey file, the number of actual hospitals sampled was 3442. The survey was sent to CEOs at each hospital; hospitals could return the paper form or fill out an electronic version online. Sampling stage and mode of administration had no substantive relationship to reported findings. We used a number of strategies to increase the response rate, including a two-page pre-announcement about the survey at the Health Forum/AHA Annual Summit; second, third and fourth mailings to hospitals; calls to all non-responders after the fourth mailings; announcements about the survey in AHA newsletters; survey promotion by the TEP and the AHA/Health Research & Educational Trust Hospital Engagement Network; and letters to hospital CEOs from state hospital associations.

econd, third and fourth mailings to hospitals; calls to all non-responders after the fourth mailings; announcements about the survey in AHA newsletters; survey promotion by the TEP and the AHA/Health Research & Educational Trust Hospital Engagement Network; and letters to hospital CEOs from state hospital associations. Data sources and variables We linked the PFE surveys to the AHA Annual Survey in order to determine hospital teaching status (none, residency programme, Council of Teaching Hospitals (COTH)); ownership (private not-for-profit, private for-profit, US Department of Veterans Affairs (VA) or other government); number of beds (≤25, 26–100, 101–200, 201–300, >300); critical access hospital (CAH) status; urban-continuum status (division, metro, micro, rural); and geographic region. We separated VA hospitals from other government hospitals because an initial inspection of survey results showed a strong pattern of differences between the two.

s (≤25, 26–100, 101–200, 201–300, >300); critical access hospital (CAH) status; urban-continuum status (division, metro, micro, rural); and geographic region. We separated VA hospitals from other government hospitals because an initial inspection of survey results showed a strong pattern of differences between the two. PFE scoring To facilitate evaluation and comparison of hospitals with respect to PFE, we used the original survey to construct a set of summary items. This approach allowed us to combine items that had dependencies and categorise items that had multiple response options. The PFE summary items were constructed in consultation with the TEP in regards to which survey items, if any, could be excluded. The summary items were structured according to the topics of the questionnaire, with items selected from each topic. We grouped these items into three broad categories: (1) organisational practices, (2) bedside practices and (3) patient and family access to information and support for shared decision-making. Analysis We summarised the characteristics of hospitals that responded to the survey and compared them to those of the hospitals that did not respond or were not sampled, using χ2-tests to test for differences between the two groups.

PFE scoring To facilitate evaluation and comparison of hospitals with respect to PFE, we used the original survey to construct a set of summary items. This approach allowed us to combine items that had dependencies and categorise items that had multiple response options. The PFE summary items were constructed in consultation with the TEP in regards to which survey items, if any, could be excluded. The summary items were structured according to the topics of the questionnaire, with items selected from each topic. We grouped these items into three broad categories: (1) organisational practices, (2) bedside practices and (3) patient and family access to information and support for shared decision-making. Analysis We summarised the characteristics of hospitals that responded to the survey and compared them to those of the hospitals that did not respond or were not sampled, using χ2-tests to test for differences between the two groups. The survey contained questions about 17 specific barriers to hospital adoption of PFE practices, each with a response scale of 1–5, in which 1 was ‘no barrier’ and 5 was ‘significant barrier’. To facilitate analysis and interpretation, we dichotomised the responses for each of the 17 items into 1–3 (‘not a significant barrier’) or 4–5 (‘significant barrier’), and reported the frequency of hospitals citing each as a significant barrier. All analyses were done using Stata version 13.1 (Stata Corp, 2014, College Station, Texas, USA).

The survey contained questions about 17 specific barriers to hospital adoption of PFE practices, each with a response scale of 1–5, in which 1 was ‘no barrier’ and 5 was ‘significant barrier’. To facilitate analysis and interpretation, we dichotomised the responses for each of the 17 items into 1–3 (‘not a significant barrier’) or 4–5 (‘significant barrier’), and reported the frequency of hospitals citing each as a significant barrier. All analyses were done using Stata version 13.1 (Stata Corp, 2014, College Station, Texas, USA). Results Of the 3442 hospitals surveyed, 1457 (42.4%) responded. The 1457 hospitals differed significantly from non-surveyed acute care hospitals on all characteristics except CAH status, with moderately higher likelihood of being COTH, public or not-for-profit, larger bed size, urban and in New England or the South Atlantic (see table 1). Eighteen per cent of respondents were hospital CEOs, with the remaining surveys being completed by other hospital staff. Table 1 Characteristics of hospitals included compared with those not surveyed

Results Of the 3442 hospitals surveyed, 1457 (42.4%) responded. The 1457 hospitals differed significantly from non-surveyed acute care hospitals on all characteristics except CAH status, with moderately higher likelihood of being COTH, public or not-for-profit, larger bed size, urban and in New England or the South Atlantic (see table 1). Eighteen per cent of respondents were hospital CEOs, with the remaining surveys being completed by other hospital staff. Table 1 Characteristics of hospitals included compared with those not surveyed Characteristic Not included Responded All eligible p Value N 3833 (100.0) 1457 (100.0) 5290 (100.0) Teaching 0.000 None 3215 (83.9) 1116 (76.6) 4331 (81.9) Residency 449 (11.7) 185 (12.7) 634 (12.0) COTH 169 (4.4) 156 (10.7) 325 (6.1) Ownership 0.000 Government 832 (21.7) 351 (24.1) 1183 (22.4) Non-profit 2008 (52.4) 866 (59.4) 2874 (54.3) Profit 900 (23.5) 208 (14.3) 1108 (20.9) VA 93 (2.4) 32 (2.2) 125 (2.4) Beds (category) 0.000 ≤25 927 (24.2) 267 (18.3) 1194 (22.6) 26–100 1192 (31.1) 384 (26.4) 1576 (29.8) 101–200 742 (19.4) 294 (20.2) 1036 (19.6) 201–300 439 (11.5) 154 (10.6) 593 (11.2) 301+ 460 (12.0) 358 (24.6) 818 (15.5) Missing 73 (1.9) 0 (0.0) 73 (1.4) Critical access hospital 0.241 No 2863 (74.7) 1111 (76.3) 3974 (75.1) Yes 970 (25.3) 346 (23.7) 1316 (24.9) Urban status 0.036 Division 535 (14.0) 228 (15.6) 763 (14.4) Metro 1766 (46.1) 662 (45.4) 2428 (45.9) Micro 629 (16.4) 267 (18.3) 896 (16.9) Rural 903 (23.6) 300 (20.6) 1203 (22.7) Region 0.000 New England 125 (3.3) 86 (5.9) 211 (4.0) Mid Atlantic 329 (8.6) 125 (8.6) 454 (8.6) South Atlantic 509 (13.3) 255 (17.5) 764 (14.4) East North Central 536 (14.0) 183 (12.6) 719 (13.6) East South Central 543 (14.2) 243 (16.7) 786 (14.9) West North Central 350 (9.1) 95 (6.5) 445 (8.4) West South Central 654 (17.1) 186 (12.8) 840 (15.9) Mountain 304 (7.9) 140 (9.6) 444 (8.4) Pacific 428 (11.2) 140 (9.6) 568 (10.7) Other 54 (1.4) 4 (0.3) 58 (1.1) Missing 1 (0.0) 0 (0.0) 1 (0.0) COTH, Council of Teaching Hospitals; VA, US Department of Veterans Affairs.

) West North Central 350 (9.1) 95 (6.5) 445 (8.4) West South Central 654 (17.1) 186 (12.8) 840 (15.9) Mountain 304 (7.9) 140 (9.6) 444 (8.4) Pacific 428 (11.2) 140 (9.6) 568 (10.7) Other 54 (1.4) 4 (0.3) 58 (1.1) Missing 1 (0.0) 0 (0.0) 1 (0.0) COTH, Council of Teaching Hospitals; VA, US Department of Veterans Affairs. The 25 items and their frequencies are given in table 2. Of the 25 items, 49% of hospitals had 9 or more fully implemented. Table 2 Constructed composite items

) West North Central 350 (9.1) 95 (6.5) 445 (8.4) West South Central 654 (17.1) 186 (12.8) 840 (15.9) Mountain 304 (7.9) 140 (9.6) 444 (8.4) Pacific 428 (11.2) 140 (9.6) 568 (10.7) Other 54 (1.4) 4 (0.3) 58 (1.1) Missing 1 (0.0) 0 (0.0) 1 (0.0) COTH, Council of Teaching Hospitals; VA, US Department of Veterans Affairs. The 25 items and their frequencies are given in table 2. Of the 25 items, 49% of hospitals had 9 or more fully implemented. Table 2 Constructed composite items Count (%) 1. Organisational practices to support PFE Formal self-assessment of PFE practices No 902 (61.9) >12 months ago 168 (11.5) Within 12 months 344 (23.6) Missing 43  (3.0) Patient and family advisory council No 883 (60.6) Some units 195 (13.4) Hospital-wide 364 (25.0) Missing 15  (1.0) Patient and family advisory council meetings within the last 12 months None 256 (17.6) 1–3 27  (1.9) ≥4 278 (19.1) Missing 896 (61.5) Percent of patient and family advisory council who are patients/family members None/no council 1114 (76.5) 1–50% 96  (6.6) 51–100% 234 (16.1) Missing 13  (0.9) Patients/family members in hospital committees Low 633 (43.4) Medium 222 (15.2) High 31  (2.1) Missing 571 (39.2) Written policy on patients’ rights to specify which family members or other partners in care will be actively involved in their care No 74  (5.1) Yes 1281 (87.9) Missing 102  (7.0) Policy facilitating unrestricted access No 153 (10.5) Some units 411 (28.2) All units 841 (57.7) Missing 52  (3.6) Formal policy for disclosing/apologising for errors No 355 (24.4) Yes 971 (66.6) Missing 131  (9.0) Patients/families routinely interviewed for root-cause analysis No 636 (43.7) Yes 684 (46.9) Missing 137  (9.4) Hospital provides training to physicians on partnering No/Unknown 918 (63.0) Yes 392 (26.9) Missing 147 (10.1) Hospital provides training to nurses on partnering No/Unknown 490 (33.6) Yes 823 (56.5) Missing 144  (9.9) Proportion of staff receiving training on partnering None 480 (32.9) 1–50% 565 (38.8) 51–100% 168 (11.5) Missing 244 (16.7) Patients/family members involved in training clinical staff (either as educators or content developers) Low 210 (14.4) Medium 72  (4.9) High 35  (2.4) Missing 1140 (78.2) Clinical staff trained in how to communicate with patients Low 185 (12.7) Medium 412 (28.3) High 221 (15.2) Missing 639 (43.9) Metrics used to track implementation of PFE strategies None 537 (36.9) 1–4 metrics 499 (34.2) 5+ metrics 211 (14.5) Missing 210 (14.4) 2.

.4) Medium 72  (4.9) High 35  (2.4) Missing 1140 (78.2) Clinical staff trained in how to communicate with patients Low 185 (12.7) Medium 412 (28.3) High 221 (15.2) Missing 639 (43.9) Metrics used to track implementation of PFE strategies None 537 (36.9) 1–4 metrics 499 (34.2) 5+ metrics 211 (14.5) Missing 210 (14.4) 2. PFE practices at the bedside Patients/family encouraged to participate in nurse shift-change report No 359 (24.6) Some units 409 (28.1) All units 583 (40.0) Missing 106  (7.3) Multidisciplinary rounds conducted with patients/family members No 454 (31.2) Some units 583 (40.0) All units 312 (21.4) Missing 108  (7.4) Teach-back used with patients None 227 (15.6) Some 586 (40.2) All units 406 (27.9) Missing 238 (16.3) White board for patients’ daily care Seldom/never 148 (10.2) Sometimes 1140 (78.2) Often/always 30  (2.1) Missing 139  (9.5) 3. Access to information and support for shared decision-making Online access to personal health information No 946 (64.9) Yes 404 (27.7) Missing 107  (7.3) Patients are given information on accessing health records No 274 (18.8) Yes 1070 (73.4) Missing 113  (7.8) Patients can examine their health records No 94  (6.5) Appointment/consult only 639 (43.9) Anytime while in the hospital 610 (41.9) Missing 114  (7.8) Health literacy and language issues addressed None 238 (16.3) Some 701 (48.1) All 361 (24.8) Missing 157 (10.8) Patients are provided with decision aids No 425 (29.2) Yes 1032 (70.8) Patients and families can activate a rapid response team No rapid response team 280 (19.2) No 234 (16.1) Through hospital staff 225 (15.4) Directly 604 (41.5) Missing 114  (7.8) PFE, patient and family engagement.

) All 361 (24.8) Missing 157 (10.8) Patients are provided with decision aids No 425 (29.2) Yes 1032 (70.8) Patients and families can activate a rapid response team No rapid response team 280 (19.2) No 234 (16.1) Through hospital staff 225 (15.4) Directly 604 (41.5) Missing 114  (7.8) PFE, patient and family engagement. Barriers The barriers included in the PFE survey are listed in table 3 in order of the frequency with which respondents rated them as significant barriers. Fifty-one per cent of respondents identified competing organisational priorities as a significant barrier, followed by time to set up and implement advisory programmes, time required for rounds and shift changes that engage patients, and financial support for PFE activities. Table 3 Barriers to patient and family engagement (PFE)

Barriers The barriers included in the PFE survey are listed in table 3 in order of the frequency with which respondents rated them as significant barriers. Fifty-one per cent of respondents identified competing organisational priorities as a significant barrier, followed by time to set up and implement advisory programmes, time required for rounds and shift changes that engage patients, and financial support for PFE activities. Table 3 Barriers to patient and family engagement (PFE) Barriers Significant barrier n (%) Competing organisational priorities 652/1276 (51.1) Time it takes to set up and implement advisory programmes 545/1284 (42.4) Time available for rounds, shift changes, etc. 409/1282 (31.9) Amount of financial support for PFE activities 411/1289 (31.9) Training of clinical providers in how to engage with patients 327/1282 (25.5) Degree of transparency of medical cost information enabling patients to compare 319/1277 (25.0) Availability of clinical information systems 284/1283 (22.1) Support for PFE from clinicians 224/1282 (17.5) Patient's willingness to participate in care activities 210/1280 (16.4) Risk management concerns 200/1282 (15.6) Degree of transparency of quality information enabling patients to compare 184/1278 (14.4) Privacy/Health Insurance Portability and Accountability Act concerns 180/1284 (14.0) Differences in language between patient and healthcare provider 156/1289 (12.1) Cultural differences between patient and healthcare provider 119/1289 (9.2) Leadership commitment to PFE activities 119/1288 (9.2) Support for PFE from hospital administrators 109/1285 (8.5) The literature/evidence supporting the usefulness of PFE 98/1275 (7.7) Discussion Findings In this first ever survey of the PFE strategies and policies implemented at hospitals across the USA, we found a wide variation in practices. Among the most widely adopted organisational practices, 88% had written policies on patients’ rights to identify which of their personal contacts they would like to have actively involved in their care, 86% had a policy for unrestricted visitor access in at least some units and 67% had formal policies for disclosing and apologising for medical errors. Within this same domain, the least frequently adopted included the involvement of patients and family members as either educators or content developers when training clinical staff (7%), patient and family advisory councils meeting within the last 12 months (21%), and patient and family members sitting on the patient and family advisory councils (23%).

frequently adopted included the involvement of patients and family members as either educators or content developers when training clinical staff (7%), patient and family advisory councils meeting within the last 12 months (21%), and patient and family members sitting on the patient and family advisory councils (23%). Among the most widely adopted PFE bedside practices, 80% at least sometimes employed the use of white boards in patients’ rooms, and 68% practised teach-back with patients in at least some units. The least frequently adopted practice was conducting multidisciplinary rounds with patients and families (61%).

frequently adopted included the involvement of patients and family members as either educators or content developers when training clinical staff (7%), patient and family advisory councils meeting within the last 12 months (21%), and patient and family members sitting on the patient and family advisory councils (23%). Among the most widely adopted PFE bedside practices, 80% at least sometimes employed the use of white boards in patients’ rooms, and 68% practised teach-back with patients in at least some units. The least frequently adopted practice was conducting multidisciplinary rounds with patients and families (61%). Among practices related to patient and family access to information and support for shared decision-making, the most widely adopted practice was allowing patients to examine their health records either anytime (42%) or by appointment or consult (44%). The second most common was taking steps to address health literacy and language issues (73%). The least commonly reported practice was giving patients 24 h online access to their personal health information (28%). These findings are difficult to interpret, in part because of the inherent limitations of a survey. Responses may not correspond to actual practices at the hospital, and there may be differences across hospitals in what a particular practice actually means. To study this further, we contacted some of the 71% of hospitals that said they provided decision aids to patients and asked them for copies of decision aids they used. A large majority of the materials they provided were simply educational materials, such as pamphlets that did not fit the technical definition provided in the survey (‘informational health materials and literature that help people become involved in decision-making by making explicit the decision that needs to be made, providing information about treatment options and outcomes, and helping the patient clarify personal values’). There are likely other item responses that similarly overstate or understate actual PFE; however, as suggested by the decision aid example, we think it is more likely that our results overstate rather than understate the use of PFE practices.

ment options and outcomes, and helping the patient clarify personal values’). There are likely other item responses that similarly overstate or understate actual PFE; however, as suggested by the decision aid example, we think it is more likely that our results overstate rather than understate the use of PFE practices. It is also important to recognise that the survey's checklist format does not capture the depth of hospitals’ commitment to PFE. It is conceivable that a hospital has ‘fully implemented’ many of our 25 summary items but only in a cursory manner; on the other hand, a hospital may be more effective at PFE because of an active and integrated patient and family advisory council that has focused strongly on only a few key strategies.

mmitment to PFE. It is conceivable that a hospital has ‘fully implemented’ many of our 25 summary items but only in a cursory manner; on the other hand, a hospital may be more effective at PFE because of an active and integrated patient and family advisory council that has focused strongly on only a few key strategies. Yet even with these inherent restrictions on interpretation, we think the survey provides clear evidence that there is large variation in adoption of practices across hospitals in the USA. While few hospitals have adopted a wide range of practices, a number of practices are in widespread use, and some hospitals have adopted a large number of them. The availability of clinical information systems, support from clinicians or administrators, and concerns related to privacy or risk management were not identified by many respondents as significant barriers. However, over half of responding hospitals viewed competing organisational priorities as a significant barrier. A lack of financial support for PFE activities and a lack of time to implement advisory programmes and specific PFE practices were identified as significant barriers in >30% of responding hospitals.

barriers. However, over half of responding hospitals viewed competing organisational priorities as a significant barrier. A lack of financial support for PFE activities and a lack of time to implement advisory programmes and specific PFE practices were identified as significant barriers in >30% of responding hospitals. Implications There are no prior comprehensive surveys of PFE, so we have few benchmarks against which to compare our results. The Centers for Medicare & Medicaid Services (CMS) records some aspects of PFE at hospitals across the country. It tracks whether lab results, tests and referrals are available electronically to patients; it also compiles patient ratings of how well nurses and doctors communicate with patients (including about medications and what to do during recovery at home).21 However, CMS does not collect information on the full range of hospital practices and policies reported here. Thus, our survey represents the best current evidence about what US hospitals are doing to engage patients and families and which barriers pose the most significant challenges. Moreover, it provides an important baseline to assess progress over time. While we believe that the use of PFE practices in hospitals is more extensive now than it was several years ago, significant opportunities to expand the use of PFE practices in American hospitals exist; such an expansion is likely to benefit both patients and the hospitals that care for them. Our findings suggest that this expansion will require changes in the priorities and resource-allocation decisions of hospital leaders. These changes are more likely if research continues to demonstrate the positive impacts of PFE practices on patients and hospital operations.

ients and the hospitals that care for them. Our findings suggest that this expansion will require changes in the priorities and resource-allocation decisions of hospital leaders. These changes are more likely if research continues to demonstrate the positive impacts of PFE practices on patients and hospital operations. Limitations This study has the limitations of any observational study in that it cannot establish any causality between reported barriers and PFE practices. In addition, the sample of respondents is not random; while hospitals that were not sampled or did not respond to the survey were similar to the respondents with respect to important hospital characteristics, respondents may have differed systematically from non-respondents with respect to PFE. Respondent bias is a particular concern because of the low response rate, with fewer than half of surveyed hospitals replying. The response rate and the number of missing responses may also be related to the length and complexity of the survey and might have been improved with additional development and testing. Non-response is also likely correlated with reduced use of PFE, biasing our results towards higher reporting of PFE practices.22 More importantly, because all measures are self-reported by a single individual in the hospital, hospital responses may not correspond to actual implementation, particularly on items in which the PFE practice's definition may be unclear (such as in the example of decision aids discussed above). Related to this, the individual responding differed from hospital to hospital, and his or her position may be related to reporting bias; CEOs may under-report strategies of which they are unaware (eg, teach-back to patients). Thus, an important limitation is that our survey results may misstate the actual use of PFE strategies and policies in the surveyed hospitals; however, we think it is more likely that our results overstate rather than understate the use of PFE practices. A final limitation is that while we strove to identify all areas of PFE for inclusion in the survey by researching the literature and receiving input from external experts, in retrospect, there are PFE practices that were omitted (eg, live translation services).

verstate rather than understate the use of PFE practices. A final limitation is that while we strove to identify all areas of PFE for inclusion in the survey by researching the literature and receiving input from external experts, in retrospect, there are PFE practices that were omitted (eg, live translation services). Future research Further research examining PFE strategy use is warranted, given the variable rates of PFE reported. It is possible that some of these hospitals may engage patients and family members through informal processes not captured in the questions we asked. For example, a hospital could have a policy of limited visiting hours, but staff might allow family members to see patients at any time. Related to this, and as mentioned above, there may be additional practices at hospitals that were not included in our survey questions, and efforts should be made to identify and measure these. Parallel work by others, for instance, has identified additional practices.23 Research to better determine which PFE activities are genuinely helpful is also warranted. Measures of how well PFE strategies are being implemented will provide useful data to hospitals and allow researchers to better understand how well-executed PFE strategies are influencing outcomes and patient experiences. Despite the limitations of this study, it provides the clearest picture thus far of how extensively US hospitals are employing common PFE strategies and which barriers are the largest impediments to progress.

Research to better determine which PFE activities are genuinely helpful is also warranted. Measures of how well PFE strategies are being implemented will provide useful data to hospitals and allow researchers to better understand how well-executed PFE strategies are influencing outcomes and patient experiences. Despite the limitations of this study, it provides the clearest picture thus far of how extensively US hospitals are employing common PFE strategies and which barriers are the largest impediments to progress. Conclusion This is the first study to assess what tactics US hospitals currently employ in order to engage with patients and their family members in patient care. We found that about half of US hospitals were fully engaged in 9 or more of 25 PFE strategies for which there is expert consensus. Some key practices have been widely adopted. Efforts to promote these practices need to continue and be coupled with assistance to ensure that PFE activities are genuinely beneficial instead of simply perfunctory. Variance in adoption can be explained by factors associated with leadership choices and resource-allocation decisions. Because increased PFE use is associated with improved patient outcomes, higher patient ratings of hospital quality and decreased use of healthcare services, it is recommended that hospitals make PFE practices a priority. Moreover, creating measures that will allow hospital leaders to determine whether the PFE practices they are using have genuine value is essential. Aligning incentives for PFE to measures of the efficacy of PFE practices, rather than to the mere existence of PFE practices, will yield better outcomes for patients and better health policy.

measures that will allow hospital leaders to determine whether the PFE practices they are using have genuine value is essential. Aligning incentives for PFE to measures of the efficacy of PFE practices, rather than to the mere existence of PFE practices, will yield better outcomes for patients and better health policy. Supplementary Material Web supplement Contributors: JH performed data analysis and interpretation and helped draft the manuscript. KH helped with data acquisition efforts and drafting the manuscript. KK helped create the survey and assisted in sending out the survey and data acquisition. He also helped with data analysis, data interpretation and the drafting of the manuscript. SH helped create the survey and assisted in data analysis, data interpretation and drafting the manuscript. MJ helped create the survey and assisted in data analysis and drafting the manuscript. DF helped create the survey and assisted in drafting the manuscript. Funding: The Gordon and Betty Moore Foundation (grant number 3903). Competing interests: None declared. Ethics approval: Health Research & Educational Trust IRB. Provenance and peer review: Not commissioned; externally peer reviewed.