UBC Faculty Research and Publications

In pursuit of quality : opportunities to improve patient experiences in British Columbian emergency departments Watson, Diane E.; Peterson, Sandra; Black, Charlyn, 1954- Jan 31, 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-Watson_D_et_al_Pursuit_quality.pdf [ 556.59kB ]
Metadata
JSON: 52383-1.0048321.json
JSON-LD: 52383-1.0048321-ld.json
RDF/XML (Pretty): 52383-1.0048321-rdf.xml
RDF/JSON: 52383-1.0048321-rdf.json
Turtle: 52383-1.0048321-turtle.txt
N-Triples: 52383-1.0048321-rdf-ntriples.txt
Original Record: 52383-1.0048321-source.json
Full Text
52383-1.0048321-fulltext.txt
Citation
52383-1.0048321.ris

Full Text

Diane E. Watson PhD MBASandra Peterson MScCharlyn Black MD ScDIn Pursuit of QualityOpportunities to improve patient experiences in British Columbian emergency departmentsJanuary 2009I N  P U R S U I T  O F  Q U A L I T Y2Library and Archives Canada Cataloguing in PublicationWatson, Diane E.        In pursuit of quality [electronic resource] : opportunities to improvepatient experiences in British Columbian emergency departments / DianeE. Watson, Sandra Peterson, Charlyn Black.Includes bibliographical references.ISBN 978-1-897085-15-8        1. Hospitals--Emergency services--British Columbia.  2. Emergencymedical services--British Columbia.  3. Patient satisfaction--BritishColumbia.  4. Hospital care--British Columbia.  5. Health servicesaccessibility--British Columbia.  I. Black, Charlyn, 1954-  II. Peterson,Sandra  III. University of British Columbia. Centre for Health Servicesand Policy Research  IV. Title.RA975.5.E5W3 2008                     362.18’09711                  C2008-907887-X   4 List of fi gures and tables 5 About CHSPR 6 Acknowledgements 7 Executive Summary 10 Introduction 11 What did we do? 14 Overall ratings of quality of care in emergency departments and factors    that underlie it 16 Improve quality by focusing on care experiences that underlie overall    ratings of quality 23 Improve quality by emulating care experiences of patients who rated their    quality as excellent 25 Improve quality by addressing factors that underlie negative experiences    to prevent similar experiences in the future 26 Conclusions 27 References 28 Appendix A: Survey methods and statistical analyses 32 Appendix B: Survey results by predisposing characteristics, presenting    characteristics and care experiences 37 Appendix C: Results of logistic regression statistical model to predict    positive patient ratings of overall quality of care 42 Appendix D: Results of logistic regression statistical model to predict     negative patient ratings of overall quality of care 48 Appendix E: Results of logistic regression statistical model to predict    patient views on waiting too long to see a doctor 51 Appendix F: Relationship between wait times and patient views on     waiting too long 52 Appendix G: Results of statistical analyses to account for differences in    patient characteristics and their presentation at emergency departments on    differences in patient ratings of overall quality of care  Appendices are available at www.chspr.ubc.caContents3U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H 13 Figure 1: Overall ratings of quality of care in emergency departments in BC     and elsewhere in Canada, 2007 15 Figure 2: Overall ratings of quality of care in emergency departments, by health    region in 2007 17 Figure 3: Patient ratings of staff courtesy in emergency departments, by type of    hospital in 2007 17 Figure 4: Patient ratings of staff courtesy in emergency departments, by health    region in 2007 18 Figure 5: Patient ratings of the degree to which doctors and nurses work together    in emergency departments, by type of hospital in 2007 18 Figure 6: Patient ratings of the degree to which doctors and nurses work together    in emergency departments, by health region in 2007 19 Figure 7: Patient ratings of the degree to which they received all the services they    needed when they visited an emergency department, by type of hospital in 2007 19 Figure 8: Patient ratings of the degree to which they received all the services they    needed when they visited an emergency department, by health region in 2007 20 Figure 9: Patients’ assessments about the degree to which they waited too long to    see a doctor when they visited an emergency department, by type of hospital in    2007 20 Figure 10: Patients’ assessments about the degree to which they waited too long    to see a doctor when they visited an emergency department, by health region in    2007 21 Table 1: Per cent of patients who said the wait for doctor was “defi nitely too    long”, by pain level and length of wait time for a doctor in 2007 22 Figure 11: Patients’ assessments of the availability of nurses when they visited an    emergency department, by type of hospital in 2007 22 Figure 12: Patients’ assessments of the availability of nurses when they visited an    emergency department, by health region in 2007 23 Figure 13: A closer look at patients’ care experiences among those that offer    positive or negative ratings of overall quality of care received in an emergency    department in 2007List of fi gures and tablesI N  P U R S U I T  O F  Q U A L I T Y45U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C HAbout CHSPRTh e Centre for Health Services and Policy Research (CHSPR) is an independent research centre based at the University of British Columbia. CHSPR’s mission is to advance scientifi c enquiry into issues of health in population groups, and ways in which health services can best be organized, funded and delivered. Our research-ers carry out a diverse program of applied health services and population health research under this agenda. Th e Centre’s work is:Independent• Population-based• Policy relevant• Interdisciplinary• Privacy sensitive• CHSPR aims to contribute to the improvement of population health by ensuring our research is relevant to contemporary health policy concerns and by working closely with decision makers to actively translate research fi ndings into policy options. Our researchers are active participants in many policy-making forums and provide advice and assistance to both government and non-gov-ernment organizations in British Columbia (BC), Canada and abroad. Funding and supportCHSPR receives core funding from the BC Ministry of Health Services, and ongoing support from the University of British Columbia. Th is enables the Centre to focus on research that has a direct role in informing policy and health reform, and facilitates CHSPR’s continuing devel-opment of the BC Linked Health Database.Our researchers are also funded by competitive external grants from provincial, national and international fund-ing agencies. Th ey include the Canadian Health Services Research Foundation, the Canadian Institutes of Health Research, the Commonwealth Fund, Health Canada, the Michael Smith Foundation for Health Research, and WorkSafeBC.Data services: BC Linked Health DatabaseMuch of CHSPR’s research is made possible through the BC Linked Health Database, a valuable resource of data relating to the encounters of BC residents with various health care and other systems in the province. Th ese data are used in a de-identifi ed form for applied health services and population health research deemed to be in the public interest.CHSPR has developed strict policies and procedures to protect the confi dentiality and security of these data hold-ings and fully complies with all legislative acts governing the protection and use of sensitive information. CHSPR has over 30 years of experience in handling data from the BC Ministry of Health Services and other professional bodies, and acts as the access point for researchers wish-ing to use these data for research in the public interest.I N  P U R S U I T  O F  Q U A L I T Y6AcknowledgementsUnder the direction of the Deputy Minister of Health Services and Chief Executive Offi  cers of the health authorities of British Columbia (BC), a Patient Satisfac-tion Steering Committee has undertaken to learn and share information about experiences BC residents have with health care they receive in the province. In 2008, that Steering Committee engaged our team to glean insights from a survey they conducted in BC emergency depart-ments in 2007 under a contract with NRC+Picker (www.nrcpicker.com). Th is report relies on survey data col-lected by NRC+Picker and analyses by CHSPR faculty and staff . Between February and April 2007 more than 50,000 people sought care in BC emergency departments. Over 16,800 of these patients completed a survey in order to share infor-mation about their fi rst-hand experiences with that care. Th eir contributions made this report possible. Th is project benefi ted from the contributions of many other individuals. In particular, we would like to ac-knowledge the expert advice provided by Lena Cuth-bertson, Co-Chair BC Patient Satisfaction Steering Committee and Kevin Samra from the Ministry of Health Services. We are also thankful for feedback from Michael Murray who authored the fi rst descriptive report using the same survey data. Th anks to a number of faculty at the UBC School of Population and Public Health for as-sistance in statistical modeling and interpretation. Alicia Priest assisted with copy-editing and Dawn Mooney assisted with graphics, layout and design. Th e BC health authorities provided funding for this project under an agreement between them and the BC Ministry of Health Services. Th e Behavioural Research Ethics Board of the University of British Columbia ap-proved of the analyses. Th e conclusions are those of the authors and no offi  cial endorsement by health authorities or the BC Ministry of Health Services is intended or should be inferred. U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H7Canadians have become increasingly concerned about lack of timely access to needed health services and the quality of their care. Th ese views have not gone unnoticed—health care policy-makers, administrators, and clinicians are taking action to ensure that many more people have positive experiences with health care services. Every day in British Columbia (BC), thousands of pa-tients who suff er from life-threatening or severe illnesses and injuries receive care in emergency departments. Many people also use emergency departments as a regu-lar source of care or when their regular medical doctor is not available. Whatever the reason they visit, patients’ experiences infl uence their views about quality of care. While clinical experts can best judge the degree to which patients receive the care that experts recommend, patients are best placed to judge the degree to which services meet their needs and expectations.  In this report, we share new insights about what 16,800 patients said about their experiences with the care they received in BC emergency departments in 2007. We take a close look at the factors that drive patient ratings of quality, as well as the degree to which patients reported very positive experiences, in order to identify what health care professionals do well and should continue to do. We also take a close look at the minority of patients who report negative experiences, to provide perspective about what can be done to address the factors underlying these experiences and prevent similar experiences in the future. Executive summaryI N  P U R S U I T  O F  Q U A L I T Y8What did we learn about factors that underlie patient views on overall quality of care in BC emergency departments?In 2007, the vast majority of patients in BC said that quality of care in emergency departments was excel-lent (27%), very good (32%) or good (24%). Th e most important factor infl uencing those who said their overall quality of care was excellent, was the degree to which they considered staff  to be courteous. What mattered also, although to a lesser extent, was: teamwork, comprehen-siveness of services and availability of nurses. A minority of patients in the province (16%) had negative experiences in emergency departments, reporting the overall quality of care they received as fair (11%) or poor (5%). Similar to those who reported positive experiences, those who reported negative ratings cited staff  courtesy as the most important factor. Other things also mattered, but to a lesser extent: comprehensiveness of services, teamwork and waiting too long to see a doctor.*Together, these fi ndings contain the following important lessons for ensuring that most patients in BC continue to report positive experiences and fewer patients report negative experiences:the factors that underlie patient ratings of both posi-• tive and negative reports of the overall quality of care in emergency departments are remarkably similar;the degree to which staff  are considered to be courte-• ous is the most important factor infl uencing patient ratings of quality;when health care professionals do well on factors • that underlie these ratings, then patients off er high ratings of overall quality of care; and when health care professionals do poorly in those ar-• eas, patients are very likely to off er negative ratings of overall quality of the care they receive in emergency departments. What did we learn about the performance of emergency departments in BC on factors that matter to patients? In 2007, the vast majority of patients in BC said that overall quality of care was excellent, very good or good and that health care professionals do well on the list of things that infl uence their views on quality. For example, many patients said that staff  courtesy was excellent (31%), very good (33%) or good (23%). Similarly, many said that teamwork was excellent (27%), very good (35%) or good (25%) and that the availability of nurses was excellent (19%), very good (29%) or good (28%). A majority said that they completely (60%) or somewhat (30%) agree that they received all the services they need at the emergency department. When asked if they waited too long to see a doctor, one in fi ve said “defi nitely” (18%) and slightly more (28%) said “yes, somewhat”. Half said “no” (52%).     Th ere is variation across types of hospitals and health regions about patient views on overall quality and ratings on the factors that matter to them. Accordingly, we off er an array of graphics in this report so that health care professionals can learn what their patients have to say. We found that patients’ characteristics (e.g. age, ethnicity) and their presentation at emergency departments (e.g. time of day, acuity) had some infl uence on their views about overall quality of care. Since patients from diff er-ent types of hospitals and health regions diff er in these characteristics, we used statistical methods to risk-adjust performance metrics. Importantly, we found that the rank order of highest and lowest performance across dif-ferent types of hospitals and health regions remained un-changed even aft er sophisticated analysis was conducted to account for diff erences in the characteristics of patients and their presentation at emergency departments. Th is is true for other measures profi led in this report.   * Th is is the patients’ view on the length of their wait for doctors, not the length of wait for nurses or the overall time in emergency departments.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H9We explored the data to better understand views of the majority of patients who said they had positive experi-ences in order to identify what could be done to emulate these circumstances with other patients. Across BC, 27 per cent of patients in BC who said overall quality of care was excellent also rated staff  courtesy as excellent or very good (99%). Similarly, almost all of these patients rated teamwork as excellent or very good (97%). Virtually all said they received all the services they needed (98%). Few waited more than two hours for a doctor (3%) though some said their wait was too long (16%). In contrast, 16% of patients off er negative ratings of overall quality of care and the majority of these individuals off ered negative ratings of staff  courtesy (62%) and teamwork (53%). Additionally, four in 10 said they did not receive all the services they needed (39%), one-third waited more than two hours for a doctor (29%) and the vast majority said their wait was too long (89%).Among patients who reported negative experiences, those who presented in pain represent a unique group. Th e more pain patients experience, the more likely they are to rate overall quality of care negatively. Patients in severe pain are twice as likely to off er negative overall ratings and patients who are in moderate/mild pain are, in turn, 1.4 times as likely as those in no pain to report negative overall ratings of quality of care. Importantly, patients who said they were in severe or moderate/mild pain represent the majority (74%) of all patients who off er negative ratings of overall quality of care. Th us, one approach to improving ratings of overall quality in emer-gency departments would be to pay particular attention to improving the experiences of patients who experience pain. Given recent policy initiatives to reduce wait times in emergency departments across BC and elsewhere in Canada, we also attempted to better understand issues related to patient views on ‘waiting too long’ to see a doc-tor. Across all patients, we determined that most patients in BC who waited less than half an hour did not mind the wait and patients who waited between half an hour and two hours were tolerant of the wait. But 96 per cent of the patients who waited more than two hours to see a doctor reported that the wait was defi nitely (71%) or somewhat (25%) too long. Th is suggests that patients seem to reach a tipping point aft er which they become less tolerant. A fi nal refl ection is that despite the attention paid by policy-makers, clinicians and the media to reducing wait times in emergency departments, we found that patient perceptions of staff  courtesy emerged as the most impor-tant factor infl uencing their reports of overall quality of care. Th is suggests that eff orts to improve quality should focus more broadly so that other factors such as staff  courtesy to patients, ensuring that patients receive the full set of services they need, and supporting team-based interaction with patients, are also addressed—in addition to wait times. We hope that this report will be used to congratulate health care professionals on the work they do with patients in emer-gency departments and to give them more information about where to target future eff orts to ensure that fewer and fewer patients have negative views regarding the overall quality of care they receive. I N  P U R S U I T  O F  Q U A L I T Y1 0IntroductionCanadians have become increasingly concerned about lack of timely access to needed health services and the quality of that care.1  Th ese views have not gone unnoticed—health care policy-makers, administrators and clinicians are taking action to ensure that patients have more positive experiences with health services (see Learning from British Columbian’s Experiences with Health Care, below).Quality of health care is a multi-faceted concept and measuring it requires an assessment from many diff erent perspectives. Clinical quality and patient safety require valid and reliable measurements of the degree to which patients receive the care that experts recommend. Simi-larly, the degree to which care is patient-centred requires standardized and scientifi cally sound measurements of experiences. Patient-centred care has been identifi ed as one of six* domains of quality. Th is domain focuses on “the patient’s experiences of illness and health care and on the systems that work or fail to work to meet individual patients’ needs”.2Every day in BC, thousands of patients who suff er from life-threatening or severe illnesses and injuries receive care in emergency departments. Many people also use emergency departments as a regular source of care3 or when their regular medical doctor is not available.4  Not surprisingly, this is oft en true in many rural areas but it also occurs in urban communities where small groups of people frequently use emergency departments for issues such as mental health care.5  Whatever the reason they visit, patients have fi rst-hand experiences that infl uence their views about overall quality of care in emergency departments.   * Th e other fi ve domains of quality include: safety, eff ectiveness, timeliness, effi  ciency and equity.LEARNING FROM BRIT ISH COLUMBIAN’S EXPERIENCES WITH HEALTH CAREIn 2003, the Deputy Minister of Health Services, other Ministry executives, and the Chief Executive Offi  cers of the health authorities struck a steering committee to commission and oversee surveys of patients across BC to obtain information for quality improvement initiatives. In 2003, that BC Patient Satisfaction Steering Committee conducted its fi rst survey to understand patient experiences with health care in emergency departments. Results were released in October 2004. Be-tween 2003 and 2007, the steering committee surveyed patients who received other types of health care services and then health care workers used that information to improve services. In 2007, the spotlight was directed again toward patients’ understanding and reporting on the accessibility and quality of emergency department services. Th e Ministry of Health Services coordinated the public release of these results in January 2008.6In 2008, the BC Patient Satisfaction steering committee requested that CHSPR conduct analyses of the 2007 survey of patient-reported experiences to identify opportunities to improve patient ratings of quality of care in emergency depart-ments so that they could best target improvement initiatives in those areas. Th is report is the result.  U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H1 1What did we do?infl uence patient ratings of quality. Th roughout the report we include graphics to illuminate the distribution of pa-tient ratings across all response categories; select graphics include notations that summarize where there are statisti-cally signifi cant diff erences between facilities and/or regions aft er accounting for diff erences between patient characteristics and/or their presentation at emergency departments. Th is information is intended to provide a baseline against which future performance can be gauged.Th ere is variation across types of hospitals and health regions about patient views on overall quality and rat-ings on the factors that matter to them. We found that patients’ characteristics (e.g. age, ethnicity) and their presentation at emergency departments (e.g. time of day, acuity) had some infl uence on their views about overall quality of care. Since patients from diff erent types of hospitals and health regions diff er in these characteristics, we used statistical methods to risk-adjust performance metrics. Importantly, we found that the rank order of highest and lowest performance across diff erent types of hospitals and health regions remained unchanged even aft er sophisticated analysis was conducted to account In order to identify factors that drive positive or negative patient ratings of the overall quality of care in emergency departments, we used results of a province-wide survey completed by more than 16,800 patients who visited one of 110 facilities in BC in 2007 (see About the Emergency Care Sector Survey, following page). We included data from the subset of survey respondents who completed 80 per cent or more of questions and also answered a ques-tion regarding overall quality of care. Th is cohort is not substantively diff erent from all individuals who complet-ed the survey. A detailed summary of the analyses meth-ods is provided in Appendix A, a detailed list of all survey items used in these analyses is provided in Appendix B, and statistical fi ndings are provided in Appendices C to G. All appendices are available at www.chspr.ubc.ca.      In addition to describing patients’ overall assessments of the quality of their care, we used statistics to identify: patient characteristics (age, gender, ethnicity, etc.); their presenting characteristics, referring to the circum-stances at the time of the visit (time of day, reason for visit, etc.); and their care experiences, referring to patient perceptions about the nature and process of their care (waiting, staff  courtesy, how well staff  work as a team, availability of nurses, etc.). All patient characteristics, presenting characteristics and care experiences are listed in Appendix B.We fi rst assessed the degree to which patient and present-ing characteristics underlie patient ratings of overall qual-ity of care. Th en, we considered these factors in tandem with information on care experiences to determine: (a) which experiences most infl uence the likelihood that a patient will report positive or negative ratings of overall quality, (b) the magnitude of infl uence that care experi-ences have on positive or negative patient ratings and (c) the relative magnitude of infl uence of patient and presenting characteristics as well as experiences with care. Finally, we illustrated the degree to which facilities and/or regions vary on the kind of care experiences that most I N  P U R S U I T  O F  Q U A L I T Y1 2for diff erences in the characteristics of patients and their presentation at emergency departments. Th is is true for other measures profi led in this report (Appendix G). In order to identify the factors that drive positive patient ratings we focused on the group of patients that reported that the overall quality of care they received in an emer-gency department was excellent (27% of respondents who completed the vast majority of survey questions). Th en, we used statistical techniques to identify the factors and experiences that diff erentiate this group from patients who reported that the overall quality of care was very good, good, fair or poor (all remaining respondents). To identify the factors that drive negative patient ratings of the overall quality of care, we used the same approach in reverse. Th at is, we identifi ed the factors and experi-ences that diff erentiate the group of patients who re-ported fair or poor ratings of overall quality of care (16% of people who completed the vast majority of survey questions) from those that off ered excellent, very good or good ratings (all remaining respondents). Because we found that patients who reported having pain rated their experiences diff erently than those who did not, we con-ducted special analyses of this group of patients.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H1 3ABOUT THE EMERGENCY CARE SECTOR SURVEYBetween February 1 and April 30, 2007 thousands of people visited one of 110 emergency departments across BC. A stratifi ed, random sample of patients who had scheduled or unscheduled visits were mailed an Emergency Care Sector Survey (n=55,613). Privacy offi  cers for all health regions approved of this project and the Offi  ce of the Information and Privacy Commissioner was notifi ed of this initiative. Over 16,800 individuals (32 per cent of delivered surveys) elected to complete the survey. In January 2008, the Ministry of Health Services coordinated the public release of results of the survey as well as compari-sons between BC and elsewhere in Canada. Th ey determined that:Th e general health of patients who visit emergency departments is lower than that reported by the adult BC popu-• lation. When asked to rate their general health, patients reported it to be excellent (16%), very good (29%), good (32%), fair (17%) or poor (6%). Patients were less likely to rate their general health as excellent or very good and more likely to rate their general health as good, fair or poor relative to other BC residents. In the month prior to visiting the emergency department, 45 per cent of patients reported that illness or injury kept them in bed for one or more days, with nine per cent reporting more than 10 days in bed.*Most patients reported that they have a regular family physician or general practitioner who they see when they have • health problems (94%). Among the general population in BC, 89 per cent reported that they have a regular medical doctor.7Patients reported that the injury or illness that prompted them to go to the emergency department was extremely • serious (12%), very serious (27%), moderately serious (38%), slightly serious (17%) or not at all serious (5%). Th e vast majority did not have an appointment for their most recent visit (93%).When asked to think about the overall emergency department services they received in 2007, a large majority of • patients rated the quality of care as excellent, very good or good. Th is level of overall satisfaction with quality of care is virtually identical to that reported by patients in other provinces (Figure 1).Figure 1» Overall ratings of quality of care in emergency departments in BC and elsewhere in Canada, 200784%84%0% 20% 40% 60% 80% 100%% of patients who rated care positivelyBCRest of Canada99%Best sites for care in BC97%Best sites for carein rest of CanadaExcellentVery goodGood  * In this section, information from all survey respondents was used to derive estimates. In the rest of this report we used information from survey questionnaires that were 80 per cent complete. A detailed description of methods can be found in Appendix A.From: Murray MA. Patient experiences with emergency departments in British Columbia, 2007. Victoria: Ministry of Health.I N  P U R S U I T  O F  Q U A L I T Y1 4Overall ratings of quality of care in emergency departments and the factors that underlie itIn 2007, the vast majority of patients in BC and in all health regions said that quality of care in emergency departments was excellent, very good or good. Patients from Interior Health were more likely to report more positive ratings of overall quality of care than patients in Fraser, Vancouver Island or Northern health regions. Patients from Fraser Health were less likely to report more positive ratings of overall quality of care than in all other health regions (Figure 2). Th e values in Figure 2 represent actual performance ratings but notations have been made to identify where performance ratings diff er signifi cantly (from a statistical perspective) from other facilities and regions aft er accounting for diff erences in patients’ predisposing and presenting characteristics.Among those who said their overall quality of care was excellent (27%), the degree to which they considered staff  to be courteous was the most important factor infl uenc-ing their rating. Other things that mattered, but to a lesser extent, included: teamwork, comprehensiveness of services and availability of nurses. Among those who rated their overall quality of care as fair (11%) or poor (5%), staff  courtesy was the most important infl uencing factor. Other things also mattered, but to a lesser extent: comprehensiveness of services, reporting too long a wait to see a doctor* and teamwork.Prospectively, we conducted separate analyses to identify factors that drive positive or negative patient ratings of the overall quality of care (Appendix C and D, respec-tively). Why? We expected that diff erent factors could underlie each experience so that health care professionals might need to do some things to promote positive patient ratings and other things to avoid negative ratings. In general, however, this is not the case. Importantly, the fac-tors that drive positive patient ratings underlie negative ones. Th is suggests that the types of quality improvement eff orts necessary to pursue high ratings of overall quality of care or to avoid negative ones are, essentially, the same.While patients’ characteristics (e.g. age, ethnicity) and their presentation at emergency departments (e.g. time of day, acuity) infl uence their views about overall quality of care, these characteristics had little power in predicting overall patient ratings of quality of care. More important, however, are issues within the scope of control of health care professionals, such as staff  courtesy, teamwork, comprehensiveness of services, the length of wait for doc-tors and the availability of nurses. What is the strongest factor? Th e degree to which patients view staff  as being courteous is the strongest factor infl uencing their ratings of overall quality of care (Appendix C and D). Quality improvement eff orts targeted at these areas, therefore, are most likely to infl uence patient ratings of overall quality of care in emergency departments.    * Th is is the patients’ view on the length of their wait for doctors, not the length of wait for nurses or the overall time in emergency departments.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H1 5Figure 2» Overall ratings of quality of care in emergency departments, by health region in 2007Vancouver Island (20 EDs)Northern (27 EDs)Interior (37 EDs)Fraser (12 EDs)Vancouver Coastal (13 EDs)3% 9% 22% 34% 30%6% 15% 27% 30% 20%5% 10% 24% 32% 28%5% 10% 22% 33% 30%4% 11% 23% 32% 29%Poor Fair Good Very good Excellent0% 20% 40% 60% 80% 100%% of patients who rated careBC overall (110 EDs) 5% 11% 24% 32% 27%F VI NI VC VI NFI FI FThe distribution of patient rat-ings have not been standardized, but letters indicate signifi cant differences between regions (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. I=different from Interior; F=Fraser; VC=Vancouver Coastal; VI=Vancouver Island; N=Northern.ED counts do not add to 110 be-cause the ED administered by the Provincial Health Services Author-ity is not included in a region. Pa-tient ratings from that facility are included in the BC overall (n=110) patient ratings.In 2007, the vast majority of patients in BC and in all health regions said that quality of care in emergency departments was excellent, very good or good. I N  P U R S U I T  O F  Q U A L I T Y1 6Improve quality by focusing on care experiences that underlie overall ratings of quality In this section, we take a close look at the factors that most profoundly infl uence patient ratings of quality—factors that both matter to patients and are amenable to being infl uenced, managed and improved by system leaders and practitioners. We created graphics to illustrate variation across types of facilities and health regions so that health care professionals can see how their patients rate the factors most associated with overall quality of care. Th en, we profi le the care experiences of patients who rated their quality as excellent; in order to identify what health care professionals do well and should con-tinue to do.Importantly, the values in these fi gures represent actual performance ratings for the purposes of baseline mea-surement, but notations have been made to identify where performance ratings diff er signifi cantly (from a statistical perspective) from other facilities and regions aft er accounting for diff erences in patients’ predisposing and presenting characteristics.   Patient experiences with staff courtesy  Th e degree to which patients rate staff  as being courteous is among the most important issues vis-à-vis positive or negative ratings of overall quality of care in emergency departments. In order to assist health care workers to see where they are doing well and identify areas for improve-ment, Figures 3 and 4 off er a baseline against which future performance can be gauged. Figures 3 and 4 illustrate that that the vast majority of patients in BC and across types of hospitals and health regions off er positive ratings of staff  courtesy. Addition-ally, both before and aft er accounting for diff erences in patients’ predisposition and their presentation at emer-gency departments, patients rate staff  courtesy at outpost hospitals the highest and rate staff  courtesy at community hospitals the lowest. Patients in Interior Health report the highest ratings of staff  courtesy and those in Fraser Health report the lowest.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H1 7Figure 3» Patient ratings of staff courtesy in emergency departments, by type of hospital in 2007Figure 4» Patient ratings of staff courtesy in emergency departments, by health region in 2007Urgent care centre (24 EDs)Outpost hospital (9 EDs)BC overall (110 EDs)Teaching hospital (5 EDs)Community hospital (38 EDs)Small hospital (34 EDs)2 9% 24% 34% 30%3% 10% 24% 33% 28%2 7% 19% 34% 37%2 6% 15% 32% 44%12 9% 18% 71%Poor Fair Good Very good Excellent0% 20% 40% 60% 80% 100%% of patients who rated courtesy3% 9% 23% 33% 31%C U OT S U OC U OT C S OT C S UVancouver Island (20 EDs)Northern (27 EDs)Interior (37 EDs)Fraser (12 EDs)Vancouver Coastal (13 EDs)2 8% 21% 34% 35%4% 13% 26% 33% 23%2 9% 23% 33% 31%3% 8% 20% 33% 33%4% 9% 22% 33% 31%Poor Fair Good Very good Excellent0% 20% 40% 60% 80% 100%% of patients who rated courtesyBC overall (110 EDs) 3% 9% 23% 33% 31%F VI NI VC VI NFI FI FThe distribution of patient ratings have not been standardized, but letters indicate signifi cant dif-ferences between facility types (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. T=different from Teaching hospital; C=Community; S=Small; U=Urgent; O=Outpost.The distribution of patient rat-ings have not been standardized, but letters indicate signifi cant differences between regions (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. I=different from Interior; F=Fraser; VC=Vancouver Coastal; VI=Vancouver Island; N=Northern.ED counts do not add to 110 be-cause the ED administered by the Provincial Health Services Author-ity is not included in a region. Pa-tient ratings from that facility are included in the BC overall (n=110) patient ratings.I N  P U R S U I T  O F  Q U A L I T Y1 8Figure 5» Patient ratings of the degree to which doctors and nurses work together in emergency departments, by type of hospital in 2007Figure 6» Patient ratings of the degree to which doctors and nurses work together in emergency departments, by health region in 2007Urgent care centre (24 EDs)Outpost hospital (9 EDs)*BC overall (110 EDs)Teaching hospital (5 EDs)Community hospital (38 EDs)Small hospital (34 EDs)2 8% 26% 33% 27%3% 9% 26% 35% 25%2 7% 21% 34% 33%1 5% 17% 33% 41%12 7% 22% 60%Poor Fair Good Very good Excellent0% 20% 40% 60% 80% 100%% of patients who rated teamwork3% 8% 25% 35% 27%U OS U OC U OT C S OT C S UVancouver Island (20 EDs)Northern (27 EDs)Interior (37 EDs)Fraser (12 EDs)Vancouver Coastal (13 EDs)2 7% 22% 36% 30%4% 11% 29% 33% 21%2 9% 23% 35% 28%3% 7% 23% 35% 29%2 9% 24% 34% 29%0% 20% 40% 60% 80% 100%% of patients who rated teamworkBC overall (110 EDs) 3% 8% 25% 35% 27%Poor Fair Good Very good ExcellentFI VC VI NFFFPatient experiences with teamwork  Th e degree to which patients rate how well doctors and nurses work together (i.e. teamwork) drives positive and negative ratings of overall quality of care in emergency departments. In order to assist health care workers to see where they are doing well and identify areas for improve-ment, Figures 5 and 6 off er a baseline against which future performance can be gauged. Figures 5 and 6 illustrate that that the vast majority of patients in BC and across types of hospitals and health regions off er positive ratings of teamwork. Addition-ally, both before and aft er accounting for diff erences in patients’ predisposition and presentation at emergency departments, patients rate staff  teamwork at outpost hospitals the highest and teamwork at teaching and com-munity hospitals the lowest. Patients in Fraser Health rate staff  teamwork the lowest while those in all other regions report similar ratings.The distribution of patient ratings have not been standardized, but letters indicate signifi cant dif-ferences between facility types (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. T=different from Teaching hospital; C=Community; S=Small; U=Urgent; O=Outpost.* Total doesn’t add to 100% as many respondents didn’t answer this question. It could be that these patients did not see a doctor when they visited the emergency depart-ment at an outpost hospital.The distribution of patient rat-ings have not been standardized, but letters indicate signifi cant differences between regions (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. I=different from Interior; F=Fraser; VC=Vancouver Coastal; VI=Vancouver Island; N=Northern.ED counts do not add to 110 be-cause the ED administered by the Provincial Health Services Author-ity is not included in a region. Pa-tient ratings from that facility are included in the BC overall (n=110) patient ratings.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H1 9Figure 7» Patient ratings of the degree to which they received all the services they needed when they visited an emergency department, by type of hospital in 2007Figure 8» Patient ratings of the degree to which they received all the services they needed when they visited an emergency department, by health region in 2007Urgent care centre (24 EDs)Outpost hospital (9 EDs)BC overall (110 EDs)Teaching hospital (5 EDs)Community hospital (38 EDs)Small hospital (34 EDs)10% 31% 57%9% 32% 57%6% 25% 68%5% 21% 73%3% 13% 82%0% 20% 40% 60% 80% 100%% of patients who received all needed services9% 30% 60%No Yes, somewhat Yes, completelyS U OS U OT CT CT CVancouver Island (20 EDs)Northern (27 EDs)Interior (37 EDs)Fraser (12 EDs)Vancouver Coastal (13 EDs)8% 28% 64%12% 34% 53%8% 31% 59%9% 29% 60%7% 29% 62%0% 20% 40% 60% 80% 100%% of patients who received all needed servicesBC overall (110 EDs) 9% 30% 60%No Yes, somewhat Yes, completelyF VII VC VI NFI FFPatient experiences with comprehensive services  Th e degree to which patients indicate that they received all the services they needed drives their positive and negative ratings of overall quality of care in emergency departments. In order to assist health care workers to see where they are doing well and identify areas for improve-ment, Figures 7 and 8 off er a baseline against which future performance can be gauged. Figures 7 and 8 illustrate that that the vast majority of patients in BC and across types of hospitals and health regions off er positive ratings of comprehensiveness of ser-vices. Additionally, both before and aft er accounting for diff erences in patients’ predisposition and their presenta-tion at emergency departments, patients rate comprehen-siveness of services at outpost hospitals the highest and rate comprehensiveness of services at teaching and com-munity hospitals the lowest. Patients in Interior Health report the highest ratings and those in Fraser Health report the lowest. The distribution of patient ratings have not been standardized, but letters indicate signifi cant dif-ferences between facility types (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. T=different from Teaching hospital; C=Community; S=Small; U=Urgent; O=Outpost.The distribution of patient rat-ings have not been standardized, but letters indicate signifi cant differences between regions (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. I=different from Interior; F=Fraser; VC=Vancouver Coastal; VI=Vancouver Island; N=Northern.ED counts do not add to 110 be-cause the ED administered by the Provincial Health Services Author-ity is not included in a region. Pa-tient ratings from that facility are included in the BC overall (n=110) patient ratings.I N  P U R S U I T  O F  Q U A L I T Y2 0Figure 9» Patients’ assessments about the degree to which they waited too long to see a doctor when they visited an emergency department, by type of hospital in 2007Figure 10» Patients’ assessments about the degree to which they waited too long to see a doctor when they visited an emergency department, by health region in 2007Urgent care centre (24 EDs)Outpost hospital (9 EDs)*BC overall (110 EDs)Teaching hospital (5 EDs)Community hospital (38 EDs)Small hospital (34 EDs)19% 32% 48%20% 29% 49%11% 24% 63%9% 22% 66%5% 7% 68%0% 20% 40% 60% 80% 100%% of patients who waited too long to see a doctor18% 28% 52%Yes, definitely Yes, somewhat NoS U OS U OT C OT CT C S0% 20% 40% 60% 80% 100%% of patients who waited too long to see a doctorBC overall (110 EDs)Yes, definitely Yes, somewhat NoVancouver Island (20 EDs)Northern (27 EDs)Interior (37 EDs)Fraser (12 EDs)Vancouver Coastal (13 EDs)14% 27% 57%26% 30% 43%15% 28% 55%17% 27% 54%14% 27% 56%18% 28% 52%F VI NI VC VI NF VII F VCI FPatient experiences with waiting too long to see a doctorTh e degree to which patients report that they waited too long to see an emergency department doctor drives their negative ratings of overall quality of care. Given recent policy initiatives to reduce wait times in emergency departments across BC and elsewhere in Canada, we also attempted to better understand issues related to patient views on ‘waiting too long’ to see a doctor (See A Closer Look: How Long is Too Long to Wait for a Doctor in an Emergency Department? on the following page).In order to assist health care workers to see where they are doing well and identify areas for improvement, Figures 9 and 10 off er a baseline against which future performance can be gauged. Figure 9 illustrates that that the majority of patients in BC and across types of hospitals and health regions off er positive ratings of wait times. Additionally, both before and aft er accounting for diff erences in patients’ predisposition and presentation at emergency departments, more patients at teaching and community hospitals report that they waited too long to see a doctor than those at small, urgent care or outpost hospitals. Figure 10 illustrates that before and aft er ac-counting for diff erences in patients’ predisposition and presentation at emergency departments, signifi cantly more in Fraser Health report that they waited too long to see a doctor than those in other health regions.The distribution of patient ratings have not been standardized, but letters indicate signifi cant dif-ferences between facility types (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. T=different from Teaching hospital; C=Community; S=Small; U=Urgent; O=Outpost.* Total doesn’t add to 100% as many respondents didn’t answer this question. It could be that these patients did not see a doctor when they visited the emergency depart-ment at an outpost hospital.The distribution of patient rat-ings have not been standardized, but letters indicate signifi cant differences between regions (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. I=different from Interior; F=Fraser; VC=Vancouver Coastal; VI=Vancouver Island; N=Northern.ED counts do not add to 110 be-cause the ED administered by the Provincial Health Services Author-ity is not included in a region. Pa-tient ratings from that facility are included in the BC overall (n=110) patient ratings.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H2 1A CLOSER LOOK:  HOW LONG IS  TOO LONG TO WAIT FOR A DOCTOR IN  AN EMERGENCY DEPARTMENT?What is it about patients’ emergency department experience that makes them feel their wait to see the doctor was too long? Is it the amount of time they waited, or do other experiences in the emergency department plus their predisposing or presentation characteristics contribute to their view that the wait was too long? We found that aft er accounting for pa-tients’ predisposition and characteristics of their presentation to emergency departments, the issue most highly associated with their view that the wait to see a doctor was defi nitely too long was—not surprisingly—the time they waited to see a doctor. Other types of experiences that underlie patient views on waiting too long are, in rank order, the availability of nurses, not getting help when needed, and waiting too long for test results (Appendix E).  Additional analyses suggest that most patients prefer a wait of less than half an hour but are tolerant of waits between half an hour and two hours. However, the overwhelming majority of patients (96%) who waited more than two hours report that their wait was defi nitely/somewhat too long. More specifi cally, among the people who did not wait or waited less than half an hour to see a doctor, the vast majority felt that the wait was not too long (90% and 80% respectively). Among patients who waited between half an hour and one hour, 40 per cent felt the wait was not too long, 47 per cent felt the wait was somewhat long and 13 per cent felt it was defi nitely too long. Among those who waited from one to two hours, 15 per cent felt the wait was not too long, 51 per cent felt the wait was somewhat too long, and 35 per cent felt the wait was defi -nitely too long. Finally, among those who waited more than two hours, 71 per cent felt the wait was defi nitely too long, 25 per cent felt the wait was somewhat too long, and the remaining four per cent felt the wait was not too long (Appendix F). A closer look at patients who experience pain reinforces the fi nding that most patients in pain prefer a wait of less than half an hour but become less tolerant of waits between half an hour and two hours. More specifi cally, among the people who did not wait or waited less than half an hour to see a doctor, four to fi ve per cent said the wait was “defi nitely too long”. Th e same is true whether or not the patient reported experiencing pain in the emergency department. Among patients who waited between half an hour and one hour, a minority felt the wait was “defi nitely too long”. Th at is, 12 per cent of those with no pain, 13 per cent of those with moderate/mild pain and 16 percent of those with severe pain felt the wait was “defi nitely” too long. Among those who waited from one to two hours, 43 per cent of those in severe pain felt the wait was “defi nitely too long”. In comparison, 28 per cent those in no pain felt similarly. Finally, among those who waited more than two hours, 77 percent of those in severe pain felt the wait was “defi nitely too long” with 70 per cent of those in moderate/mild pain feeling the same and 66% of those in no pain feeling the same (Table 1).Wait time for a doctor was “defi nitely too long”< 30 minutes 30–60 minutes 1–2 hours 2+ hoursSevere pain 5% 16% 43% 77%Moderate/mild pain 5% 13% 34% 70%No pain 4% 12% 28% 66%Table 1» Per cent of patients who said the wait for doctor was “defi nitely too long”, by pain level and length of wait time for a doctor in 2007I N  P U R S U I T  O F  Q U A L I T Y2 2Figure 11» Patients’ assessments of the availability of nurses when they visited an emergency department, by type of hospital in 2007*Figure 12» Patients’ assessments of the availability of nurses when they visited an emergency department, by health region in 2007*Urgent care centre (24 EDs)Outpost hospital (9 EDs)*BC overall (110 EDs)Teaching hospital (5 EDs)Community hospital (38 EDs)Small hospital (34 EDs)7% 14% 28%8% 17% 28%4% 12% 31%3% 8% 32%14% 23%0% 20% 40% 60% 80% 100%% of patients who rated availability of nurses7% 15% 29%18%16%24%29%60%19%29%29%26%25%11%28%Poor Fair Good Very good ExcellentS U OS U OT C U OT C S OT C S U0% 20% 40% 60% 80% 100%% of patients who rated availability of nursesBC overall (110 EDs)Vancouver Island (20 EDs)Northern (27 EDs)Interior (37 EDs)Fraser (12 EDs)Vancouver Coastal (13 EDs)5% 13% 31%10% 19% 25%6% 15% 28%6% 16% 29%7% 15% 29%7% 15% 29%21%14%19%19%20%19%27%30%28%28%27%28%Poor Fair Good Very good ExcellentF VC VI NI VC VI NI FI FI FPatient experiences with availability of nursesTh e degree to which patients rate the availability of nurses drives their positive ratings of overall quality of care. In order to assist health care workers to see where they are doing well and identify areas for improvement, Figures 11 and 12 off er a baseline against which future performance can be gauged. Figures 11 and 12 illustrate that that the majority of patients in BC and across types of hospitals and health regions off er positive ratings regarding the availability of nurses. Additionally, both before and aft er accounting for diff erences in patients’ predisposition and presentation at emergency departments, patients rate the availability of nurses at outpost hospitals the highest and rate the avail-ability of nurses at teaching and community hospitals the lowest. Patients in Fraser Health off er the lowest rat-ings of availability and those in Interior Health off er the highest. Patients in all other health regions off er similar ratings.The distribution of patient ratings have not been standardized, but letters indicate signifi cant dif-ferences between facility types (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. T=different from Teaching hospital; C=Community; S=Small; U=Urgent; O=Outpost.* Many outpost hospitals are prin-cipally or solely staffed by nurses.The distribution of patient rat-ings have not been standardized, but letters indicate signifi cant differences between regions (p<.005) after standardizing for differences in the predisposition of patients and their presentation at EDs. I=different from Interior; F=Fraser; VC=Vancouver Coastal; VI=Vancouver Island; N=Northern.ED counts do not add to 110 be-cause the ED administered by the Provincial Health Services Author-ity is not included in a region. Pa-tient ratings from that facility are included in the BC overall (n=110) patient ratings.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H2 3Figure 13» A closer look at patients’ care experiences among those that offer positive or negative ratings of overall quality of care received in an emergency department in 2007Patient ratings ofoverall quality of carePoor Good Very good ExcellentFair Poor Good Very good ExcellentFairPoor Good Very good ExcellentFair>2 hrs ½-1 hr <½ hr No wait1-2 hrs >2 hrs ½-1 hr <½ hr No wait1-2 hrsPoor Good Very good ExcellentFairNo Yes, somewhat Yes, completelyYes, definitely Yes, somewhat No Yes, definitely Yes, somewhat NoNo Yes, somewhat Yes, completelyPoor Fair Good Very good ExcellentRatings ofstaff courtesy 17% 45% 28% 7% 2 8% 91%Ratings ofstaff teamwork 16% 36% 34% 6% 2 2 20% 77%Waiting timeto see doctor 15% 226%22%29% 53% 18%17%7%3Felt waited toolong to see doctor 58% 31% 10% 3 13% 82%Ratings of comprehen-siveness of services 9%51%39% 93%5%15% 11% 24% 32% 27%Among the 16%who offered poor andfair ratings of overall qualityAmong the 27%who offered an excellentrating of overall qualityIn 2007, almost all patients in BC who said overall quality of care was excellent (27%) also rated staff  courtesy as excellent (91%) or very good (8%). Similarly, almost all of these patients rated teamwork as excellent (77%) or very good (20%). Virtually all said they received all the services they needed (98%). Few waited more than two hours for a doctor (3%) though some said their wait was too long (16%). When health care professionals do well on factors that underlie these ratings, then patients off er high ratings of overall quality of care (Figure 13).Improve quality by emulating care experiences of patients who rated their quality as excellent I N  P U R S U I T  O F  Q U A L I T Y2 4Th is story is similar across health regions and illustrates the degree to which these factors underlie positive rat-ings.In the Interior Health region, almost all patients who • said overall quality of care was excellent (30%) also rated staff  courtesy as excellent (93%) or very good (6%). Similarly, almost all of these patients rated teamwork as excellent (78%) or very good (19%). Virtually all of these patients said they received all the services they needed (98%). Few waited more than two hours for a doctor (2%), though some said their wait was too long (14%). In the Vancouver Island Health region, almost all • patients who said overall quality of care was excellent (30%) also rated staff  courtesy as excellent (91%) or very good (7%). Similarly, almost all of these patients rated teamwork as excellent (76%) or very good (20%). Virtually all of these patients said they re-ceived all the services they needed (98%). Few waited more than two hours for a doctor (4%) though some said their wait was too long (17%). In the Northern Health region, almost all patients • who said overall quality of care was excellent (29%) also rated staff  courtesy as excellent (90%) or very good (9%). Similarly, almost all of these patients rat-ed teamwork as excellent (80%) or very good (18%). Virtually all of these patients said they received all the services they needed (99%). Few waited more than two hours for a doctor (2%) though some said their wait was too long (18%). In the Vancouver Coastal Health region, almost all • patients who said overall quality of care was excellent (28%) also rated staff  courtesy as excellent (91%) or very good (8%). Similarly, almost all of these patients rated teamwork as excellent (79%) or very good (19%). Virtually all of these patients said they re-ceived all the services they needed (97%). Few waited more than two hours for a doctor (1%) though some said their wait was too long (13%). In Fraser Health region, almost all patients who said • overall quality of care was excellent (20%) also rated staff  courtesy as excellent (87%) or very good (11%). Similarly, almost all of these patients rated teamwork as excellent (73%) or very good (25%). Virtually all of these patients said they received all the services they needed (98%). Few waited more than two hours for a doctor (4%) though some said their wait was too long (20%).U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H2 5Improve quality by addressing factors that underlie negative experiences to prevent similar experiences in the futureIn 2007, a minority of patients in BC (16%) said that overall quality of care they received in an emergency department was fair or poor. Th e key factors underlying those ratings were staff  courtesy, teamwork, compre-hensiveness of services and waiting too long for doctors. Among these individuals, a majority off er negative ratings of staff  courtesy (62%) and teamwork (53%).  Four in 10 said they did not get all the services they needed (39%). One-third waited more than two hours for a doctor (29%). Th e vast majority said their wait was too long (89%). Clearly, when health care professionals do poorly in the areas that matter to overall ratings of quality, patients are very likely to off er negative ratings of overall quality of the care they receive in emergency departments (Figure 13). Th is story is similar across health regions and illustrates the degree to which these factors underlie negative rat-ings and diff er between patients who off er positive and negative views on overall quality of care in emergency departments. In the Interior Health region, 12% of patients off er • negative ratings of overall quality of care. Many of these patients off er negative ratings of staff  courtesy (59%) and teamwork (51%). Some of these patients said they did not get all the services they needed (39%). One-quarter waited more than two hours for a doctor (26%). Nine in 10 said their wait was too long (89%). In the Vancouver Island Health region, 15% of • patients off er negative ratings of overall quality of care. Many of these patients off er negative ratings of staff  courtesy (62%) and teamwork (50%). Some of these patients said they did not get all the services they needed (44%). One-third waited more than two hours for a doctor (32%). Nine in 10 said their wait was too long (87%). In the Northern Health region, 15% of patients off er • negative ratings of overall quality of care. Many of these patients off er negative ratings of staff  courtesy (66%) and teamwork (55%). Some of these patients said they did not get all the services they needed (34%). One-fi ft h waited more than two hours for a doctor (21%). Eight in 10 said their wait was too long (84%). In the Vancouver Coastal Health region, 15% of • patients off er negative ratings of overall quality of care. Many of these patients off er negative ratings of staff  courtesy (61%) and teamwork (55%). Some of these patients said they did not get all the services they needed (40%). One-quarter waited more than two hours for a doctor (25%). Almost nine in 10 said their wait was too long (87%). In Fraser Health region, 22% of patients off er nega-• tive ratings of overall quality of care. Many of these patients off er negative ratings of staff  courtesy (65%) and teamwork (55%). Some of these patients said they did not get all the services they needed (39%). Four in 10 waited more than two hours for a doctor (37%). Nine in 10 said their wait was too long (93%). Due to the high number of people who indicated that they waited too long, we conducted analyses similar to that outlined in Appendix C and D to identify factors among patients in Fraser Health that underlie their positive or negative ratings regard-ing overall quality. We suspected, prospectively, that waiting too long may jump to the top of the queue in terms of priority. But the resultant analysis indicated that staff  courtesy remained the key driver underly-ing positive and negative ratings in Fraser Health. One patient characteristic stands out in analyses of fac-tors that underlie negative patient ratings. Th e more pain patients experience, the more likely they are to rate over-all quality of care negatively. Patients who said they were in severe or moderate/mild pain represent the majority I N  P U R S U I T  O F  Q U A L I T Y2 6Conclusions(74%) of all patients that off er negative ratings of overall quality of care. One in fi ve patients in severe pain (20%) off er negative ratings of staff  courtesy and one-quarter off er negative ratings of overall quality of care (24%). Patients in severe pain are twice as likely to off er negative ratings and patients in moderate/mild pain are, in turn, 1.4 times as likely as those in no pain to report negative overall ratings of quality of care. When patients experi-ence pain in emergency departments and, then view staff  as not being courteous they are more likely to rate overall quality of care negatively (See Appendix D). Th us, one strategy to shrink the number of patients who hold nega-tive views of overall quality in emergency departments is to target eff orts toward those patients who experience pain.Every day thousands of British Columbians receive care in emergency departments. Th eir experiences with that care matters to them and their families, as well as to oth-ers who want to learn about their fi rst-hand experiences with our health care system. While expert clinicians can best judge the degree to which patients receive high qual-ity clinical services, patients are best placed to judge the degree to which services are patient-centred. Perhaps the most important fi nding of this work is that we now know that the degree to which staff  are courte-ous, particularly to patients in pain, is the key driver of patient ratings of overall quality of care in emergency departments. Indeed, staff  courtesy is the single greatest infl uence on the likelihood that patients will report positive overall  ratings of quality of care. Teamwork, comprehensive services, wait times to see a doctor, views on the reasonableness of that wait and availability of nurses matter also but not as much as the courteousness of staff . U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H2 71 Soroka SN. A report to the Health Council of Canada. Canadian perceptions of the health care system. Toronto: Health Council of Canada; 2007. 2  Institute of Medicine. Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academy Press; 2001. p. 48.3  Statistics Canada. Canadian Community Health Survey, 2007. Th e Daily. 2008 June 18;3-9.4  Health Council of Canada. Fixing the foundation: An update on primary health care and home care renewal in Canada. Toronto: Health Council of Canada; 2007. 5  Doupe M, Kozyrskyj A, Soodeen RA, Derksen S, Burchill C, Huq S. An initial analyses of emer-gency departments and urgent care in Winnipeg. Winnipeg, MB: Manitoba Centre for Health Policy; 2008 May.6  Murray MA. Patient experiences with emergency departments in British Columbia, 2007. Victoria: BC Ministry of Health; 2008.7  Statistics Canada. Canadian Community Health Survey. Cycle 3.1. Ottawa: Statistics Canada; 2005.ReferencesI N  P U R S U I T  O F  Q U A L I T Y2 8APPENDIX  ASurvey methods and statistical analysesUnder the direction of the Deputy Minister of Health Services and Chief Executive Offi  cers of the health au-thorities of British Columbia (BC), a Patient Satisfaction Steering Committee has undertaken to learn and share information about the experiences that BC residents have with health care they receive in the province. In 2006 and 2007, that Steering Committee oversaw the implementa-tion of a patient experiences survey—the Emergency Care Sector Survey under a contract with NRC+Picker (www.nrcpicker.com). Th is report describes a secondary analysis of this survey data.Study participants, sampling and sample weightsTh e Steering Committee established and implemented a strategy to ensure a representative sample size was obtained from all participating emergency departments. Patients were randomly selected to participate, with the sample drawn from the records of patient visits at the facility level. Diff erent sampling fractions were used for youth (ages 12 to 19) versus non-youth (less than 12 years and older than 19 years) due to the need for a diff erent survey technique for these two populations. Diff erent sampling strategies were used for facilities depending on their size (extra small, small, medium and large) to ensure large enough sample sizes from each.*Patients were excluded from the survey if they had no fi xed address, were infants up to 10 days old, had experi-enced a miscarriage or therapeutic abortion, were fl agged as “do not announce” or a similar designation, or were deceased in hospital. Where possible, patients presenting with sensitive issues were also excluded, such as those that presented with a confi rmed or suspected sexual abuse and/or domestic violence, or patients who died aft er discharge from hospital.  Patients between the ages of 12 and 19 years (termed “youth”) were included in the study. Youth were sampled separately and mailed surveys in unmarked envelopes; that is, with no health authority or facility logo showing to mitigate privacy concerns of including this population. Mailing of surveys to youth in unmarked envelopes was undertaken as a risk mitigation strategy and was ap-proved by both the Offi  ce of the Information and Privacy Commissioner and the Provincial Information, Privacy and Security Working Group.Surveys with accompanying cover letters and return envelopes were mailed by NRC+Picker to patients’ home addresses starting February 22, 2007. A reminder letter and survey were sent 24 days later to those who had not yet responded. Th e mailed survey was in English but Chinese, Punjabi and French versions were available by calling a 1-866 number. A web based response option was also off ered via a unique access code.Weights were calculated so that mailed surveys would be representative of the province as a whole. Th e weights adjust for the diff ering sampling fractions between youth and non-youth and diff erent sampling fractions among facilities. Th ese weights have been applied in all analyses unless otherwise noted. Post-stratifi cation weighting has not been done to adjust for the age/sex diff erences be-tween, for example, those who were mailed surveys and those who responded.   * Due to a sampling error only a subset of the patient population for St. Paul’s Hospital (SPH) was selected to receive a survey. By the time this error was discovered it was too late to re-survey a correct sample and using the returned surveys from the incorrect sample would not provide results from a representative population. Fortunately, SPH had undertaken a survey of its emergency patient population aged 20+ years that had made visits from April through September 2006 using the same survey tool and vendor. SPH is a major source of emergency department care, so it was deemed important to include data for SPH in the analyses. Th us, although the time frame was slightly diff erent, the 2006 responses have been used in place of the 2007 responses for SPH. Th ese data are weighted to refl ect SPH’s 2007 volumes within the province.  U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H2 9Survey instrumentTh e Emergency Care Sector Survey (ECSS) has, as its core content, questions developed and widely used in the United States. Th e American version was modifi ed and tested in Canada in 2002 and fi eld testing included three BC hospital emergency departments.* It has also been used in New Brunswick (n=353), Nova Scotia (n=4,164), Manitoba (n=385), Ontario (n=106,098) and Yukon (n=490). Th e survey comes from a family of questionnaires initial-ly developed by the Picker Institute in Boston. Th e Picker Institute developed a suite of questionnaires to under-stand the patient’s experiences with health care and not just their satisfaction with it. Th at is, patients are asked to report on whether something good (or bad) happened or not, or to give an evaluation of some aspect of care.6Th e survey is a 66-item questionnaire, covering six dimensions of quality (access and coordination of care, respect for patient preferences, emotional support, information and education, continuity and transition and physical comfort), as well as demographics, health status questions, and other questions such as reason for and seriousness of visit. Survey responses were linked with se-lect administrative data such as age and gender, Canadian Triage Acuity Score (CTAS)** and time and date of visit. Emergency department facilities were grouped according to type, size, and location.Response rates***Th e overall response rate to the survey was 32% (n=16,837); this varied from 28% to 38% across health authorities in the province. Th e age and gender structure of respondents diff ered from the population who were mailed surveys. Th e sample of respondents included in this report is slightly biased toward women and substan-tially biased toward older persons.6  Among completed surveys, 78% were done by the patient and 17% by some-one else. Five percent did not respond to this question. Completeness of data***Among returned surveys, the completeness of survey questions, excluding questions within a skip pattern, ranged from 0% to 100%. Ninety-three per cent of surveys were over 80% complete, 4% were 50% to 80% complete, and a further 3% were less than 50% complete. To ensure robustness of results, only those surveys that were 80% or more complete were included in the analyses (n=15,619). A comparison of the distribution of key vari-ables between all returned surveys and those that were 80% or more complete showed no substantial diff erences (Appendix B).Statistical analysesFigures 2 to 13 were created using responses from surveys that were 80% or more complete.  Th e values in Figures 2 to 13 represent actual performance ratings for the purposes of baseline measurement, but notations have been made in fi gures 2 to 12 to identify where mean performance ratings are signifi cantly diff erent than other facilities and regions aft er accounting for diff erences in predisposing and presenting characteristics of the patients they serve. When making multiple comparisons between types of hospitals or health regions, a p value of 0.005 was used, rather than the 0.05 convention, to refl ect a Bonferroni correction factor which is used in situations of multiple testing in order to maintain the overall level of Type I error at 95%.  Figures by health region do not include Provincial Health Services Authority because it only has one emergency department, but patient ratings at that facility were included in the BC overall profi les.    * Validation of the Picker Emergency Care Survey in Canada, National Research Corporation, January 2003.  ** Only 63% of weighted cases contained CTAS in the data. Th is varied by hospital type: CTAS was in the data for 74% of weighted cases in teaching hospitals, 63% in community hospitals, 54% in small hospitals, 57% in urgent care centres, and 0% in outpost hospitals. *** Numbers and percents presented in this section are not weighted.I N  P U R S U I T  O F  Q U A L I T Y3 0Patients who came to the emergency department because they had a scheduled appointment (1,047 respondents) were excluded from the following analyses in order to focus on a group of non-scheduled emergency patients.  In addition, because the experience of patients who were in severe pain, moderate pain, mild pain or no pain was expected to be diff erent, 198 respondents who indicated they had pain but did not indicate the severity were excluded. Th e fi nal cohort used in the following analyses was further restricted to those respondents who answered the question on overall care in the Emergency Depart-ment and consisted of 14,207 respondents (weighted n = 14,572).In order to focus attention on patients who have posi-tive or negative views of their experiences in emergency departments, we selected two metrics or outcomes of interest for some of the analyses in this report, both based on the question: Overall, how would you rate the care you received in the Emergency Department?  Potential answers were: Poor, Fair, Good, Very Good or Excellent. Positive experience: adults who rate their experience in emergency department as excellent. Th is group repre-sents 27% of all survey participants in the analyses. Negative experience: adults who rate their experience in the emergency department as fair or poor. Th is group represents 16% of all survey participants in the analyses.  Independent variables were grouped into four categories: Predisposing (patient characteristics such as age, • gender and general health), Presenting (visit characteristics such as day/time of • visit and reason for visit), Pain level (severe, moderate/mild or no pain/miss-• ing*), andExperience (system response such as wait for nurse • and doctor and courtesy of nurses and doctors).A full list of these variables can be found in Appendix B.Categorizing the variables in this way allows us to de-termine separately (and combined) the infl uence each category has on patient experience.Correlation matrices of variables within categories, and specifi c variables between categories were fi rst examined as the variables were, potentially, highly correlated and as such could not be included simultaneously in multivari-ate regression models. Bivariate survey logistic regres-sions were performed between each independent variable and the two outcome variables.  To develop a parsimonious set of predisposing and presenting factors, variables within each category were entered into forward stepwise logistic regressions with the two outcome variables. Only those variables that entered into the models were retained for further analysis (reduced set). Th e reduced set of predisposing and pre-senting variables, as well as pain level, were then entered into a survey logistic procedure with the two outcome variables.  To determine which experience variables most infl uence patient rating of quality of care, a forward stepwise logis-tic regression was performed for the experience variables for each outcome aft er inclusion of the reduced set of pre-  * Th e layout of the skip pattern for the pain questions in the survey caused a high percent (10.6%) of respondents in our fi nal cohort to not answer the question on if they had pain. (Respondents who had pain were instructed to go to the next question. Th e placement of that instruction was prior to the bubble to fi ll in for a ‘no’ response, making it logical to asume that a missing response was most likely a ‘no pain’ response.) If the respondent skipped the question on if they had pain but went on to fi ll out a severity level, they were classifi ed as having pain with that severity level (0.8%). If, however, the respondent skipped the ques-tion on if they had pain and also skipped the question about severity level, they were grouped with the ‘no pain’ group (9.8%).  When this group was analysed separately from the ‘no pain’ group in the logistic regressions, this group and the ‘no pain’ group had very similar odds ratios, supporting the decision to group them together. Th e small number of respondents (1.7%) who indicated they had no pain but also indicated a severity level were re-classifi ed as having pain with that severity level.  U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H3 1disposing, presenting and pain variables. For the positive outcome a total of 18 variables entered sequentially, and for the negative outcome a total of 27 variables entered. Aft er the fi rst few steps, each addition variable added only marginal improvement to the fi t or performance of the logistic regression as measured by the pseudo r2 value, so variables that entered in the fi rst four steps only are presented (between step 4 and step 5 the per cent increase in pseudo r2 was 0.9% for the positive outcome and 1.3% for the negative outcome).Because the results from the correlation matrix indicated that many of the experience variables are very highly cor-related, and because it’s useful to know the contribution of each experience variable separately from the others, fi ve fi nal models were created for each outcome variable. Th e fi rst four fi nal models were created for each outcome variable using survey logistic, with the reduced sets of predisposing and presenting variables and with one of the experience variables that had entered in the fi rst four steps. Th e fi ft h fi nal model for each outcome variable had the reduced sets of predisposing and presenting variables and all four of the experience variables that entered in the fi rst four steps (See Appendix C and D).Because pain level was expected to infl uence patient rating of experience, an interaction term was included between pain level and courtesy of ED staff  in the fi nal model for the negative outcome (See Appendix D).Data were analyzed using SAS 9.1.3 survey procedures.What is a pseudo r2 value?In linear regression models, r2 is a standard measure of fi t. It takes the values between 0 and 1, becomes larger as the model fi ts better and can be interpreted as the proportion of the total variability explained by the model. In logistic regression there is not an exact replica for r2, but there are several measures intended to mimic the r2 measure of fi t, oft en called a pseudo r2.* Th e Cox and Snell r2 used in this report is one such measure, and is a measure of the improve-ment of the full model over the intercept-only model. One drawback with the Cox and Snell pseudo r2 is that it cannot reach a maximum of 1, so that while the interpretation is not quite the same as an r2 from a linear regression, it can be interpreted as an approximate measure of the amount of variance in the dependent variable accounted for by the model, with a higher value indicating better model fi t.What is an odds ratio?An odds ratio is defi ned as the ratio of the odds of an event occurring in one group to the odds of it occurring in another group. It is perhaps best illustrated by an example. To take an example from the data, consider two groups of patients, those in pain and those not in pain, who are asked to rate overall quality of care. For those patients in pain, the odds of giving negative ratings of overall quality of care is estimated by: # of patients in pain who gave negative ratings   =  1,713   =   0.237 # of patients in pain who gave other ratings    7,236Th e odds of patients not in pain giving negative ratings of overall quality of care is estimated by: # of patients not in pain who gave negative ratings  =  589    =   0.117 # of patients not in pain who gave other ratings   5,035Th e relative chances of negative ratings in the two groups can be estimated by calculating the ratio of the pain and no pain odds, called an odds ratio: OR  = 0.237 = 2.03           0.117Th e odds ratio indicates that patients in pain have twice the odds of rating overall quality as negative, compared to those patients not in pain (reference group). If the odds ratio were less than one, it would indicate that the group under consideration has lower odds of giving negative ratings as compared to the reference group, while an odds ratio of one would indicate that both groups are equally likely to give negative ratings.What is an adjusted odds ratio?An adjusted odds ratio is an odds ratio that is statistically ad-justed (controlled) to account for contributions from other variables in the model.  * It should be noted that diff erent pseudo r2 measures can arrive at very diff erent values.I N  P U R S U I T  O F  Q U A L I T Y3 2APPENDIX  BSurvey results by predisposing characteristics, presenting characteristics and care experiencesAll returned surveysn=16,837Weighted* n=16,85080%+ complete data cohortn=15,619Weighted* n=15,758Predisposing characteristicsAge  0–12** 12.1% 12.5%  13–19 6.0% 6.2%  20–34 11.6% 12.1%  35–49 17.0% 17.2%  50–64 22.3% 22.5%  65–74 13.4% 13.3%  75+ 17.6% 16.1%Sex  Female 53.6% 53.3%  Male** 46.3% 46.6%  Unknown/missing 0.1% 0.1%Education  Public school 12.5% 12.8%  High school 30.8% 31.8%  College, trade or technical school 25.9% 26.7%  University undergraduate 11.0% 11.4%  Post university/graduate education** 8.5% 8.7%  Missing 11.3% 8.6%Self-reported ethnicity  Aboriginal 3.4% 3.5%  Asian 7.6% 7.9%  Caucasian** 78.0% 79.9%  Other 4.3% 4.5%  Missing 6.7% 4.2%Self-reported health status, in general  Poor 5.70% 5.8%  Fair 16.2% 16.4%  Good 30.9% 31.6%  Very good 28.1% 29.2%  Excellent** 14.9% 15.5%  Missing 4.2% 1.5%Do you have a regular family physician  Yes** 90.4% 92.9%  No 6.0% 6.1%  Missing 3.5% 1.0%Days in bed due to illness/injury, in past month  None** 52.0% 53.4%  1–3 days 18.9% 19.7%  4–10 days 15.1% 15.5%  More than ten days 8.9% 9.1%  Missing 5.0% 2.2%Patient in a hospital overnight or longer, in past 6 months  No** 75.1% 77.3%  Yes, only one time 13.7% 14.0%  Yes, more than one time 6.9% 7.0%  Missing 4.3% 1.7%Presenting characteristicsDay/time of visit  Weekday: 00:00–06:59 11.4% 11.4%  Weekday: 07:00–17:59 44.2% 44.0%  Weekday: 18:00–23:59 13.4% 13.5%  Weekend: 00:00–06:59 5.3% 5.3%  Weekend: 07:00–17:59 19.5% 19.7%  Weekend: 18:00–23:59** 6.1% 6.0%  Missing 0.1% 0.1%Reason for visit***  It clearly was an emergency 41.8% 43.6%  I was told to go by a health professional/    BC Nurse Line19.7% 20.4%  I didn’t know if my health condition was  an emergency or not**15.6% 16.2%  There were no other options/didn’t know  where else to go9.8% 10.2%  Other 4.7% 4.7%  Missing 8.3% 4.9%Self-reported seriousness of injury/illness  Extremely serious 11.0% 11.2%  Very serious 25.9% 26.9%  Moderately serious 36.1% 37.4%  Slightly serious 16.3% 16.8%  Not at all serious** 4.8% 4.8%  Missing 5.9% 3.0%Canadian Acuity Triage Scale (administrative data)  Level I – Resuscitation 0.2% 0.2%  Level II – Emergent 7.0% 7.0%  Level III – Urgent 26.0% 25.6%  Level IV – Less Urgent 24.3% 24.7%  Level V – Non Urgent** 5.3% 5.3%  Missing 37.2% 37.2%U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H3 3In pain during encounter  Severe pain 23.9% 24.7%  Mild/Moderate pain 33.8% 35.2%  No pain/missing** 40.7% 38.6%  Pain, did not indicate severity† 1.5% 1.4%Experiences with careWait for nurse  Right away** 29.0% 29.9%  15 minutes or less 33.5% 34.7%  More than 15 minutes 28.8% 29.6%  Don’t know 4.2% 4.1%  Missing 4.5% 1.8%Wait for doctor  I did not wait at all** 8.6% 8.8%  Less than ½ hour 37.6% 39.3%  Between ½ hour and 1 hour 24.4% 25.5%  1 to 2 hours 13.5% 14.0%  More than 2 hours 9.7% 9.8%  Missing 6.3% 2.7%Total time spent in ED  Less than 1 hour** 13.1% 13.4%  Between 1 and 3 hours 36.7% 38.1%  Between 3 and 6 hours 24.9% 26.1%  Between 6 and 12 hours 12.9% 13.4%  More than 12 hours 6.3% 6.5%  Missing 6.0% 2.5%Pain control  Staff defi nitely did everything they could 31.5% 32.9%  Staff somewhat did everything they could 15.3% 16.0%  Staff did not do everything they could 10.5% 10.8%  Had pain, but unknown staff response 2.3% 2.1%  No pain** 27.5% 28.4%  Missing 12.9% 9.8%Pain medication  Not enough pain meds 4.2% 4.4%  Right amount of pain meds 26.4% 27.6%  Too much pain meds 0.6% 0.6%  Had pain, but did not get meds 24.9% 25.9%  Had pain, unknown meds 2.0% 1.7%  No pain** 28.7% 29.7%  Missing 13.3% 10.1%Believed you/family member suffered personal injury or harm which resulted from a medical error or mistake  Yes 2.3% 2.3%  No ** 82.5% 86.0%  I don’t know 5.9% 6.0%  Missing 9.3% 5.7%Received all ED services needed  Yes, completely** 57.2% 59.7%  Yes, somewhat 28.9% 30.2%  No 9.0% 8.7%  Missing 4.9% 1.4%Courtesy of nurses  Poor 2.8% 2.9%  Fair 7.9% 8.2%  Good 21.9% 22.7%  Very good 35.0% 36.3%  Excellent** 26.5% 27.7%  Missing 5.8% 2.3%Courtesy of doctors  Poor 2.4% 2.4%  Fair 6.4% 6.7%  Good 20.0% 20.7%  Very good 34.0% 35.2%  Excellent** 31.7% 33.1%  Missing 5.4% 1.9%Courtesy of ED staff  Poor 2.8% 2.8%  Fair 8.9% 9.2%  Good 21.7% 22.5%  Very good 32.1% 33.4%  Excellent** 29.4% 30.8%  Missing 5.1% 1.2%How well doctors and nurses worked together  Poor 2.7% 2.8%  Fair 7.8% 8.1%  Good 23.5% 24.5%  Very good 33.0% 34.7%  Excellent** 26.0% 27.4%  Missing 7.0% 2.5%Was ED as clean as it should have been?  Yes, defi nitely** 57.6% 60.1%  Yes, somewhat 27.8% 29.0%  No 8.4% 8.6%  Missing 6.2% 2.2%Enough privacy during ED visit  Yes, always** 57.2% 59.8%  Yes, sometimes 19.8% 20.7%  No 9.3% 9.5%  Doesn’t apply 8.7% 9.0%  Missing 5.0% 0.9%I N  P U R S U I T  O F  Q U A L I T Y3 4Times when did not get the help needed  Yes, often 5.5% 5.5%  Yes, sometimes 17.0% 17.9%  No** 51.5% 53.6%  Did not need help 20.2% 21.1%  Missing 6.1% 2.0%Availability of ED nurses  Poor 6.6% 6.7%  Fair 14.9% 15.4%  Good 27.2% 28.2%  Very good 27.8% 28.9%  Excellent** 17.9% 18.6%  Missing 5.7% 2.3%Waited too long to see ED doctor  Yes, defi nitely 17.4% 17.7%  Yes, somewhat 27.3% 28.4%  No** 50.2% 52.1%  Missing 5.1% 1.8%Waited too long to get ED test(s) completed  Yes, defi nitely 6.3% 6.5%  Yes, somewhat 14.2% 14.7%  No** 44.1% 45.9%  Missing 35.4% 32.9%Wait too long for other doctor/specialist  Yes, defi nitely 4.7% 4.8%  Yes, somewhat 7.6% 7.9%  No** 19.3% 20.0%  No other doctor was needed 59.2% 62.1%  Missing 9.2% 5.2%Particular doctor in charge of your care in the ED  Yes** 69.6% 72.5%  Not sure 19.0% 19.3%  No 7.2% 7.1%  Missing 4.3% 1.1%Confi dence/trust in ED nurses  Yes, always** 65.9% 68.7%  Yes, sometimes 23.8% 24.5%  No 4.9% 4.8%  Missing 5.4% 1.9%Confi dence/trust in ED doctors  Yes, always** 67.3% 70.1%  Yes, sometimes 21.9% 22.7%  No 5.2% 5.2%  Missing 5.6% 1.9%ED nurse discussed fears/anxieties  Yes, completely** 25.0% 26.0%  Yes, somewhat 19.2% 19.9%  No 12.5% 12.7%  Did not have anxieties or fears 38.6% 40.1%  Missing 4.7% 1.2%ED doctor discussed fears/anxieties  Yes, completely** 36.3% 37.9%  Yes, somewhat 21.1% 21.9%  No 8.0% 8.1%  Did not have anxieties or fears 29.1% 30.4%  Missing 5.5% 1.7%ED got messages to family/friends  Yes** 12.2% 12.6%  No 8.7% 8.6%  I had no messages 74.5% 77.1%  Missing 4.7% 1.7%Had enough say about ED care  Yes, defi nitely** 56.1% 59.1%  Yes, somewhat 27.8% 29.1%  No 8.5% 8.7%  Missing 7.6% 3.1%Treated with dignity/respect by ED staff  Yes, always** 72.2% 75.4%  Yes, sometimes 17.1% 17.7%  No 5.8% 5.9%  Missing 5.0% 1.0%ED nurses talked as if patient wasn't there  Yes, often 3.9% 4.0%  Yes, sometimes 10.5% 10.8%  No** 80.2% 83.4%  Missing 5.3% 1.7%ED doctors talked as if patient wasn't there  Yes, often 3.8% 3.9%  Yes, sometimes 7.6% 7.7%  No** 82.8% 86.4%  Missing 5.7% 2.0%ED nurses answered questions understandably  Yes, always** 48.0% 49.8%  Yes, sometimes 20.9% 21.6%  No 4.6% 4.5%  Did not have any questions 22.3% 23.2%  Missing 4.2% 0.9%U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H3 5ED doctors answered questions understandably  Yes, always** 57.2% 59.8%  Yes, sometimes 20.1% 20.8%  No 5.0% 5.0%  Did not have any questions 12.2% 12.6%  Missing 5.5% 1.9%ED admission answered questions  Yes, completely** 51.1% 52.7%  Yes, somewhat 20.3% 20.7%  No 4.2% 4.3%  Did not have any questions 20.0% 20.6%  Missing 4.4% 1.7%ED explained causes for problem understandably  Yes, completely** 45.5% 47.9%  Yes, somewhat 23.7% 24.7%  No 8.7% 8.9%  Did not need explanation 16.4% 17.0%  Missing 5.8% 1.6%ED explained test results understandably  Yes, completely** 35.5% 37.2%  Yes, somewhat 14.5% 15.1%  No 11.5% 11.8%  Missing 38.5% 35.9%Explained reason for ED wait  Yes** 24.3% 25.1%  No 38.8% 40.1%  Did not have to wait 28.7% 29.8%  Missing 8.2% 5.0%ED explained reasons for tests understandably  Yes, completely** 38.2% 40.1%  Yes, somewhat 13.9% 14.3%  No 10.1% 10.4%  Missing 37.8% 35.2%ED explained danger signals to watch for  Yes, completely** 44.9% 47.2%  Yes, somewhat 22.0% 22.9%  No 24.7% 25.4%  Missing 8.4% 4.5%Knew who to call with questions when left ED  Yes** 59.0% 61.8%  No 17.4% 18.1%  Not sure 14.1% 14.7%  Missing 9.5% 5.4%ED explained how to take new medications  Yes, completely** 29.5% 31.0%  Yes, somewhat 6.9% 7.2%  No 7.2% 7.5%  Did not need explanation 42.4% 44.6%  Missing 13.9% 9.7%ED explained medication side effects  Yes, completely** 14.7% 15.5%  Yes, somewhat 7.3% 7.6%  No 17.4% 18.2%  Did not need explanation 46.3% 48.8%  Missing 14.3% 10.0%Appt for treatment made before left ED  Yes, with a new doctor or nurse** 10.5% 11.0%  Yes, with same doctor or nurse 6.0% 6.3%  No 29.5% 30.8%  Missing 53.9% 51.9%Outcome variablesOverall rating of care received  Poor 4.7% 4.6%  Fair 10.6% 11.0%  Good 22.8% 23.6%  Very Good 30.9% 32.3%  Excellent 26.0% 27.3%  Missing 5.1% 1.3%Positive patient ratings  Poor/Fair/Good/Very Good 68.9% 71.5%  Excellent‡ 26.0% 27.3%  Missing 5.1% 1.3%Negative patient ratings  Good/Very Good/Excellent 79.7% 83.2%  Poor/Fair‡ 15.2% 15.5%  Missing 5.1% 1.3%Facility-level groupsType of facility  Teaching Hospital 16.9% 16.9%  Community Hospital 62.5% 62.7%  Small Hospital 14.7% 14.7%  Urgent Care Centre 5.4% 5.3%  Outpost Hospital 0.5% 0.4%I N  P U R S U I T  O F  Q U A L I T Y3 6Health Authority  Interior 25.5% 25.4%  Fraser 19.7% 19.9%  Vancouver Coastal 19.5% 19.6%  Vancouver Island 20.5% 20.1%  Northern 11.4% 11.4%  PHSA 3.5% 3.6%OtherWho responded to survey  Patient 76.3% 78.4%  Someone else 18.5% 18.8%  Missing 5.2% 2.7%Appointment#  Yes 4.9% 5.0%  No 85.0% 88.2%  I do not know 1.4% 1.4%  Missing 8.7% 5.4%     * Weighted percents (for provincial-level report). Weights were calculated based on representativeness of full sample (N=55,613). (Representativeness of youth/non-youth within facilities (DTU/regular cases for St. Paul’s Hospital); of facility volumes within HA volumes; and of HA volumes within the province.)  ** Reference groups. *** In the survey, respondents could choose more than one reason for their visit.  For our analyses, respondents were as-signed to one reason only in a hierarchical order, as shown in the table.    † Because the experience of patients who were in severe pain, moderate pain, mild pain or no pain was expected to be diff erent, patients who indicated they had pain but did not indicate the severity were excluded from the fi nal cohort.    ‡ Dependent variables.     # Patients who came to the emergency department because they had a scheduled appointment (n=1,047) were excluded from the fi nal cohort in order to focus on a group of non-scheduled emergency patients.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H3 7APPENDIX  CResults of logistic regression statistical model to predict positive patient ratings of overall quality of careIn order to identify what underlies positive patient ratings of the overall quality of care in emergency departments, we used statistical methods to identify factors that are associated with the likelihood that a survey respondent would view overall quality as excellent (27 per cent of all people who completed the vast majority of survey ques-tions). Age, gender, ethnicity, education and general health status matter; but not that muchOlder or male patients are more likely to rate overall quality of care in emergency departments positively. In fact, older adults (50+ years) are roughly two times more likely to report positive experiences compared to patients that are less than 12 years of age. Interestingly, teenagers and young adults (13 to 34 years) are least likely to report positive ratings.*    Aboriginal or Asian patients are less likely than Cauca-sian patients to report positive ratings of overall quality of care. Asian patients are the least likely to off er positive patient ratings.**Patients who have high school education as their highest level of education are less likely than patients who have university and post university/graduate education to rate overall quality of care positively.***    * Th e adjusted odds ratio for teens and young adults was 0.80 and 0.88 (p<.0001), respectively. Th e reference group includes individuals less than 12 years. Odds ratios are adjusted for other predisposing characteristics.   ** Th e adjusted odds ratio for Aboriginal relative to Caucasian patients is 0.79 and the adjusted odds ratio for Asian patients is 0.32 (p<.0001). Odds ratios are adjusted for other predisposing characteristics.   *** Th e adjusted odds ratio for high school relative to post university/graduate education is 0.82 (p<.05). Odds ratios are adjusted for other predisposing characteristics.   † Th e adjusted odds ratio for poor health status is 0.35 (reference group is excellent health status) (p<.0001). Odds ratios are adjusted for other predisposing characteristics.   ‡ Th e pseudo r2 value of the logistic regression model was 0.06.   # Th e adjusted odds ratio for midnight until 7:00 am on a weekday is 0.87 and a weekend is 0.81 (reference group is week-end from 6:00 pm until midnight) (p<.01). Odds ratios are adjusted for predisposing and presenting characteristics.Th e lower the overall health status of patients the less likely they are to rate overall quality of care positively. Th is is true whether health is measured on a fi ve-point scale or by a count of the number of days spent in bed in the last month due to illness or injury.†As a collection, these predisposing characteristics had relatively little power to predict positive patient ratings of overall quality of care.‡ When these factors are consid-ered in tandem with presenting characteristics and with information on experiences in emergency departments, the only factor that remained important to predicting positive patient ratings was ethnicity. Th erefore, predis-posing characteristics such as age, gender, ethnicity and general health status infl uence positive patient ratings of overall quality of care but not to a great extent.Th ese fi ndings suggest that factors above and beyond the predisposition of patients infl uence their views of overall quality of care in emergency departments.Time of day, seriousness of illness, acuity and pain matter; but not that muchPatients who visit the emergency department between midnight and 7 a.m, are less likely to rate overall quality of care positively.#I N  P U R S U I T  O F  Q U A L I T Y3 8    * Th e adjusted odds ratio for extremely serious is 1.53 (reference group is ‘not at all serious’) (p<.0001). Odds ratios are adjusted for predisposing and presenting characteristics. ** Th e adjusted odds ratio for CTAS level 1 – Resuscitation is 1.95 (reference group is ‘non-urgent’) (p<.0001). Odds ratios are adjusted for predisposing and presenting characteristics.     *** Th e adjusted odds ratio for severe pain is 0.67 and the adjusted odds ratio for moderate/mild pain is 0.77 (reference group is no pain or no response to the survey question regarding pain) (p<.0001). Odds ratios are adjusted for predisposing and presenting characteristics.   † Th e pseudo r2 value for the logistic regression model was 0.07.   ‡ Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.07. Th e inclusion of information on staff  courtesy increased the pseudo r2 value to 0.51 and the full forward stepwise logistic regression model had an pseudo r2 value of 0.57 and included 18 types of survey items measuring experiences.   # Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.07. Th e inclusion of information on teamwork increased the pseudo r2 value to 0.42 and the full forward stepwise logistic regression model had a pseudo r2 value of 0.57 and included 18 types of survey items measuring experiences. ## Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.07. Th e inclusion of information on ‘got all services needed’ increased the pseudo r2 value to 0.23. Patients are more likely to rate overall quality of care posi-tively if they perceive that the illness that brought them to the emergency department was extremely serious.* Patients are also more likely to rate overall quality of care positively if they have high acuity (clinician-assessed).**Th e more pain patients experience, the less likely they are to rate overall quality of care positively. Th at is, patients who are in severe pain are less likely than those in moder-ate/mild pain to off er positive ratings. Patient who are in moderate/mild pain are, in turn, less likely than those in no pain to rate overall quality of care positively.***As a collection of factors, these presenting characteris-tics had relatively little power to predict positive patient ratings of overall quality of care.† When these factors are considered in tandem with predisposing characteris-tics and with information on experiences in emergency departments, the only factor that remains important to predicting positive overall ratings of quality of care was patients’ assessment of the seriousness of their illness or injury. Th erefore, presenting characteristics such as time of day, seriousness of illness (self-assessed), acuity (clinician-assessed) and pain underlie positive patient ratings of overall quality of care. However, they do not infl uence ratings to a great extent. Th is fi nding suggests that factors other than patients’ predisposition and their presentation at emergency departments infl uence their views of overall quality of care.  Teamwork and receipt of comprehensive care matter; but courtesy of emergency department staff matters mostTh e factor most strongly associated with positive rat-ings of overall quality of care is to what degree patients feel emergency department staff  are courteous. In fact, patient ratings regarding staff  courtesy have a very strong infl uence. When this factor is considered in tandem with patients’ predisposing and presenting characteristics, no other experience in the emergency room matters as much.‡Another experience that underlies positive patient ratings of overall quality of care is ratings of the degree to which doctors and nurses work together. In fact, patient ratings of teamwork infl uence their views of overall quality of care above and beyond the degree to which they see staff  as being courteous.#Two other factors infl uence positive patient ratings of overall quality of care: the degree to which patients report receiving all the services they needed## (i.e. compre-U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H3 9Bivariate Predisposing Reduced Set (r2=0.0557)Predisposing, Presenting and Pain (r2=0.0702)Predisposing, Presenting, Pain and Step 1 Experience variable (r2=0.5062)Predisposing, Presenting, Pain and Step 2 Experience  variable (r2=0.0.4198)Predisposing, Presenting, Pain and Step 3 Experience  variable (r2=0.2286)Predisposing, Presenting, Pain and Step 4 Experience  variable (r2=0.3363)Predisposing, Presenting, Pain and Steps 1, 2, 3 and 4 Experience  variable (r2=0.5480)Variable Wt N % Odds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valuePredisposingAge  <.0001 <.0001 <.0001 0.0047 <.0001 <.0001 <.0001 0.08060 to 12* 1,878 12.7  13 to 19 947 6.4 0.79 0.80 0.85 1.02 0.92 1.01 1.00 1.1920 to 34 1,839 12.5 0.78 0.88 0.94 1.11 1.09 1.12 0.99 1.3235 to 49 2,556 17.3 1.09 1.38 1.46 1.48 1.58 1.44 1.43 1.6150 to 64 3,260 22.1 1.49 2.03 2.05 1.60 1.77 1.78 1.83 1.6265 to 74 1,924 13.0 1.55 2.27 2.21 1.74 1.74 1.76 1.90 1.6275 and older 2,346 15.9 1.20 1.88 1.79 1.55 1.79 1.53 1.65 1.66Sex  <.0001 <.0001 0.0004 0.1565 0.0814 0.0033 0.2029 0.4770Missing 8 0.1 6.71 7.33 7.96 11.84 2.10 4.57 4.00 3.48Female 7,874 53.4 0.84  0.87 0.89 0.95 0.89 0.92 0.95 0.98Male* 6,869 46.6     Education  0.0010 0.0452 0.0425 0.4789 0.2604 0.0302 0.0301 0.1426Missing 1,247 8.5 0.76 0.87 0.85 0.85 0.84 0.78 0.86 0.69Public school 1,859 12.6 0.69  0.80 0.80 0.79 0.79 0.71 0.71 0.67High school 4,667 31.6 0.75 0.82 0.82 0.79 0.76 0.76 0.75 0.66College, trade or technical school3,968 26.9 0.74 0.77 0.76 0.76 0.78 0.76 0.74 0.72University undergraduate1,716 11.6 0.89 0.93 0.92 0.93 0.91 0.89 0.94 0.90Post university/ graduate education*1,294 8.8  Self-reported ethnicity  <.0001 <.0001 <.0001 0.0358 <.0001 <.0001 <.0001 0.0463Missing 623 4.2 0.84 0.91 0.89 1.23 1.09 1.06 0.90 1.35Aboriginal 494 3.4 0.64  0.79 0.78 0.98 0.82 1.01 0.89 1.08Asian 1,194 8.1 0.29  0.32 0.31 0.59 0.41 0.36 0.33 0.53Caucasian* 11,780 79.9  Other 659 4.5 0.70 0.78 0.77 1.45 0.70 0.78 0.77 1.31hensive care) and their perception of the availability of nurses.* Th ese factors are important but not as important as staff  courtesy and teamwork.**In summary, patient experiences in the emergency department infl uence their views of quality of care more so than does their predisposition and their presentation at emergency departments.  Specifi cally, patient experi-ences with staff  courtesy and with teamwork are the two principal drivers of their positive ratings of overall quality of care in emergency departments, but receiving compre-hensive care and the availability of nurses matters also.      * Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.07. Th e inclusion of information on ‘availability of nurses’ increased the pseudo r2 value to 0.34.  ** Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.07. Th e forward stepwise logistic regression model initially included staff  courtesy (pseudo r2 value of 0.51), then teamwork (pseudo r2 value of 0.53), then ‘got all services needed’ (pseudo r2 value of 0.541) and then ‘availability of nurses’ (pseudo r2 value of 0.548).Results of logistic regression statistical models for positive patient ratings of quality of care (27% of respondents). Cohort: 80% complete Emergency cohort (excludes appointments and pain level=unknown severity). Weighted n=14,572.I N  P U R S U I T  O F  Q U A L I T Y4 0Self-reported general health status  <.0001 <.0001 <.0001 0.0079 <.0001 <.0001 <.0001 0.0802Missing 221 1.5 0.76 0.61 0.56 0.78 0.70 0.55 0.65 0.68Poor 848 5.7 0.51 0.35 0.33 0.61 0.55 0.38 0.48 0.77Fair 2,381 16.1 0.46 0.32 0.31 0.58 0.48 0.38 0.42 0.65Good 4,619 31.3 0.47 0.35 0.34 0.62 0.50 0.36 0.41 0.65Very Good 4,356 29.5 0.59 0.50 0.49 0.72 0.60 0.48 0.51 0.76Excellent* 2,325 15.8  Days in bed due to illness/injury, in past month<.0001 0.1031 0.2851 0.2593 0.4845 0.6270 0.2347 0.3268Missing 305 2.1 0.76 0.78 0.78 0.65 0.69 0.79 0.71 0.63None* 7,937 53.8One–three days 2,918 19.8 0.76 0.91 0.95 1.01 0.91 1.02 0.95 1.05Four–ten days 2,279 15.4 0.70 0.87 0.91 0.89 0.91 1.03 0.84 0.87More than ten days 1,313 8.9 0.76  0.86 0.85 0.79 0.90 1.11 0.97 1.01PresentingDay/time of visit 0.0002 0.0081 0.4645 0.9878 0.1995 0.6036 0.6647Missing 13 0.1 2.76 2.28 1.26 1.24 1.96 1.22 0.77Weekday, 00:00–06:591,701 11.5 0.86 0.87 1.10 1.10 0.94 0.87 1.23Weekday, 07:00–17:596,348 43.0 1.03 0.97 1.08 1.08 1.03 1.06 1.21Weekday, 18:00–23:592,046 13.9 1.00 1.00 1.17 1.12 1.02 0.99 1.15Weekend, 00:00–06:59806 5.5 0.80 0.81 1.37 1.11 0.89 1.06 1.40Weekend, 07:00–17:592,922 19.8 0.97 0.93 0.93 1.04 0.98 0.97 1.05Weekend, 18:00–23:59*914 6.2Reason for visit 0.0037 0.0553 0.1877 0.5975 0.3806 0.2932 0.6605Missing 745 5.0 0.99 0.86 1.28 0.82 0.81 0.74 0.98It clearly was an emergency6,547 44.4 1.01 1.01 1.15 1.01 0.99 0.99 1.09I was told to go by a health professional/ BCNurse Line2,794 18.9 0.83 0.83 0.88 0.97 0.88 0.90 0.93I didn't know if my health condition was an emergency or not*2,469 16.7   There were no other options/didn't know where else to go1,562 10.6 0.80 0.88 0.93 0.88 0.92 0.88 0.89Other 635 4.3 0.92 0.96 1.09 0.92 0.90 0.84 0.82Self-rated seriousness <.0001 <.0001 0.0113 0.0316 <.0001 0.0314 0.0629Missing 410 2.8 1.21 1.31 1.71 1.46 1.60 1.30 1.66Extremely serious 1,621 11.0 1.32 1.53 1.35 1.38 1.75 1.49 1.39Very serious 3,959 26.8 0.92 1.08 1.01 0.99 1.27 1.10 1.06Moderately serious 5,577 37.8 0.88 0.99 1.11 1.05 1.08 1.10 1.13Slightly serious 2,505 17.0 1.03 1.13 1.43 1.20 1.12 1.25 1.41Not at all serious* 679 4.6U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H4 1Canadian Triage Acuity Score <.0001 <.0001 0.0123 0.0065 0.0009 0.0073 0.2539Missing 5,451 37.0 1.02 1.09 0.97 1.10 1.12 1.19 0.97Level 1 – Resuscitation28 0.2 2.13 1.95 1.89 1.92 2.27 1.37 2.01Level 2 – Emergent 1,052 7.1 1.44 1.33 1.03 1.49 1.39 1.60 1.14Level 3 – Urgent 3,789 25.7 0.81 0.85 0.78 1.00 1.00 1.03 0.92Level 4 – Less urgent3,686 25.0 0.77 0.81 0.74 0.87 0.90 1.03 0.78Level 5 – Non-urgent*744 5.1Pain levelIn pain during encounter<.0001 <.0001 0.0572 0.0007 0.0609 0.0054 0.3347Severe pain 3,738 25.3 0.68 0.67 0.79 0.74 0.90 0.77 0.84Moderate/mild pain 5,324 36.1 0.73 0.77 0.85 0.79 0.87 0.86 0.90No pain/missing* 5,689 38.6ExperienceCourtesy of ED staff (Step 1) <.0001 <.0001 <.0001Missing 176 1.2 64.18 55.51 10.01Poor* 423 2.9   Fair 1,354 9.2 0.49 0.50 0.59Good 3,348 22.7 0.74 0.70 0.38Very Good 4,937 33.5 14.08 12.26 1.98Excellent 4,513 30.6 813.02 723.03 44.35How well doctors and nurses worked together (Step 2)<.0001 <.0001 <.0001Missing 357 2.4 45.09 36.43 5.31Poor* 402 2.7    Fair 1,209 8.2 1.46 1.50 1.25Good 3,654 24.8 4.65 4.39 3.08Very Good 5,131 34.8 54.31 46.00 6.22Excellent 3,998 27.1 970.58 832.88 23.05Got all services needed (Step 3) <.0001 <.0001  <.0001Missing 198 1.3 4.73 4.62  1.54Yes, completely 8,753 59.3 20.15 20.48 4.09Yes, somewhat 4,509 30.6 1.20 1.33  0.86No* 1,290 8.7     Availability of nurses (Step 4) <.0001 <.0001 <.0001Missing 323 2.2 15.38 14.24 6.29Poor* 999 6.8    Fair 2,288 15.5 2.33 2.25 3.05Good 4,179 28.3 6.18 5.60 3.07Very Good 4,239 28.7 26.80 23.34 5.01Excellent 2,723 18.5 224.57 197.80 11.46    * Reference category.I N  P U R S U I T  O F  Q U A L I T Y4 2APPENDIX  DResults of logistic regression statistical model to predict negative patient ratings of overall quality of careIn order to identify what underlies negative patient rat-ings of the overall quality of care in emergency depart-ments, we used statistical methods to identify factors that are associated with the likelihood that a survey respondent would view overall quality as fair or poor (16 per cent of all people who completed the vast majority of survey questions). Age, gender, ethnicity and general health status matter; but not that muchTeenagers, young adults and/or females are more likely to rate overall quality of care in emergency departments negatively. Older adults are the least likely to report nega-tive overall patient ratings.*Aboriginal or Asian patients are twice as likely as Cauca-sian patients to rate overall quality of care negatively.** Th e lower the general health status of patients the more likely they are to rate overall quality of care negatively. Th is is true whether health is measured on a fi ve-point scale or by a count of the number of days spent in bed in the last month due to illness or injury.***As a collection, these predisposing characteristics had relatively little power to predict negative patient ratings of overall quality of care.†  When these factors are consid-ered in tandem with presenting characteristics and with information on experiences in emergency departments, only age remained important in predicting negative patient ratings. Th erefore, predisposing characteristics such as age, gender, ethnicity and general health status infl uence patient ratings of overall quality of care but not to a great extent. Th ese fi ndings suggest that factors above and beyond the predisposition of patients infl uence their views of overall quality of care in emergency departments.     Time of day, reason for the visit, seriousness of illness and acuity matter; but not that muchPatients who visit the emergency department between midnight and 7 a.m. are more likely to rate overall quality of care negatively.‡  Th ey are also more likely to report negative ratings if they feel they ‘had no other options/didn’t know where else to go’ or were ‘told to go [to the emergency department] by a health professional or BC Nurse Line’.#    * Th e adjusted odds ratio for teens and young adults was 1.63 and 1.58 (p<.0001), respectively. Th e adjusted odds for older adults 65 to 74 and 75 or older was 0.38 and 0.40 (p<.0001), respectively. Th e reference group includes individuals less than 12 years. Th e adjusted odds for females was 1.20 (p<.005). Odds ratios are adjusted for other predisposing characteristics.    ** Th e adjusted odds ratio for Aboriginal patients is 1.93 and the adjusted odds ratio for Asian patients is 2.02 (refer-ence group includes Caucasians) (p<.0001). Odds ratios are adjusted for other predisposing characteristics.  *** Th e adjusted odds ratio for poor health status is 2.71 (reference group is excellent health status) (p<.0001). Th e adjusted odds ratio for more than 10 days in bed is 1.72 (reference group is no days in bed with illness or injury) (p<.0001). Odds ratios are adjusted for other predisposing characteristics.   † Th e pseudo r2 value for the logistic regression was 0.06.   ‡ Th e adjusted odds ratio for midnight until 7:00 am on a weekday is 1.36 and weekend is 1.47 (reference group is weekend from 6:00 pm until midnight) (p<.05). Odds ratios are adjusted for predisposing and presenting charac-teristics.   # Th e adjusted odds ratio for ‘no other options’ is 1.51 and for ‘was told to go by a health professional’ was 1.54 (refer-ence group is didn’t know if health condition was an emergency) (p<.0001). Odds ratios are adjusted for predispos-ing and presenting characteristics.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H4 3    * Th e adjusted odds ratio for extremely serious is 1.39, very serious is 1.32, moderately serious is 1.19 and slightly serious is 0.95 (reference group is ‘not at all serious’) (p<.05). Odds ratios are adjusted for predisposing and present-ing characteristics.  ** Th e adjusted odds ratio for Level 1 – Resuscitation is 0.64 and for Level 2 – Emergent is 0.56 (reference group is Level 5 – Non-Urgent) (p<.0001). Odds ratios are adjusted for predisposing and presenting characteristics. ***   Th e adjusted odds ratio for severe pain is 2.10 and the adjusted odds ratio for moderate/mild pain is 1.39 (refer-ence group is no pain or no response to the survey question regarding pain) (p<.0001). Odds ratios are adjusted for predisposing and presenting characteristics.     †  Th e pseudo r2 value for the logistic regression was 0.08.   ‡ Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.08. Th e inclusion of information on staff  courtesy increased the pseudo r2 value to 0.35 and the full forward stepwise logistic regression model had a pseudo r2 value of 0.46 and included 27 types of survey items measuring experiences.   # The pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.08. The forward stepwise logistic regression model initially included staff courtesy (pseudo r2 value of 0.35), then got all services needed (pseudo r2 value of 0.39), then ‘how well doctors and nurses worked together’ (pseudo r2 value of 0.41) and then ‘waited to long to see an emergency department doctor’ (pseudo r2 value of 0.43). The full forward stepwise logistic regression model had a pseudo r2 value of 0.46 and included 27 types of survey items measuring experiences. ##    Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.08. Th e inclusion of information on ‘got all services needed’ increased the pseudo r2 value to 0.27.Th e more serious patients perceive the illness that brought them to emergency to be, the more likely they are to rate overall quality of care negatively.*  However, they are less likely to report negative ratings if they are assessed as having a higher acuity by a clinician.**Th e more pain patients experience, the more likely they are to rate overall quality of care negatively. Patients in severe pain are twice as likely to off er negative ratings and patients who are in moderate/mild pain are, in turn, 1.4 times as likely as those in no pain to report negative overall ratings of quality of care.***As a collection of factors these presenting characteristics had relatively little power to predict negative patient ratings of overall quality of care.†  When these factors are considered in tandem with patients’ predisposing characteristics, as well as with information on experi-ences in emergency departments, only experiences with pain remained important to predicting negative patient ratings. Th erefore, presenting characteristics such as time of day, reason for visit, seriousness of illness and acuity underlie positive patient ratings of overall quality of care but not to a great extent. Th ese fi ndings suggest that factors above and beyond patients’ predisposition and their presentation at emer-gency departments infl uence their views of overall quality of care. Th eir level of pain is very important, as outlined in this report.Comprehensive services and wait times matter; but courtesy of emergency department staff matters mostTh e factor most strongly associated with negative patient ratings of overall quality of care is the degree to which emergency department staff  is considered to be courte-ous. In fact, patient ratings of staff  courtesy have a strong infl uence. When this factor is considered in tandem with predisposing and presenting characteristics, there are no other experiences in the emergency department that mat-ter as much in predicting a negative rating.‡However, other factors infl uence patient ratings of a nega-tive experience beyond the degree to which staff  are cour-teous.#  Th ese include the degree to which patients report they received all the services they needed,## perceived I N  P U R S U I T  O F  Q U A L I T Y4 4that doctors and nurses worked together,* and/or waited too long to see an emergency department doctor.**In summary, patients’ experiences are more impor-tant than their predisposition and their presentation at emergency departments to their views of overall quality of care. Patient experiences with staff  courtesy principally underlie negative patient ratings of overall quality of care in emergency departments, but comprehensive care, teamwork and waiting too long for a doctor also matter.   Staff courtesy towards patients that experience pain is pivotalTh e more pain patients experience, the more likely they are to rate overall quality of care negatively.***  Patients in severe pain are twice as likely to off er negative overall ratings and patients who are in moderate/mild pain are, in turn, 1.4 times as likely as those in no pain to report negative overall ratings of quality of care. Importantly, patients who said they were in severe or moderate/mild pain represent the majority (74%) of all patients that off er negative ratings of overall quality of care.    * Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, includ-ing pain, was 0.08. Th e inclusion of information on ‘how well doctors and nurses worked together’ increased the pseudo r2 value to 0.32.  ** Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.08. Th e inclusion of information on ‘waited to long to see an emergency department doctor’ increased the pseudo r2 value to 0.23.***   Th e adjusted odds ratio for severe pain is 2.10 and the adjusted odds ratio for moderate/mild pain is 1.39 (refer-ence group is no pain or no response to the survey question regarding pain) (p<.0001). Odds ratios are adjusted for predisposing and presenting characteristics.  However, patients in pain don’t always report negative ratings—it very much depends on how they rate emer-gency department staff  courtesy. When patients in mild, moderate or severe pain rate staff  courtesy as excellent, very good or good, they are no more likely than patients not in pain to rate overall quality of care negatively. But when patients in mild, moderate or severe pain rate staff  courtesy as fair they are more likely to rate care negatively than patients without pain who rate staff  courtesy as excellent. What’s more, when patients in mild, moderate or severe pain rate staff  courtesy as poor they are many more times more likely to rate overall quality of care negatively than patients who have no pain and who rate staff  courtesy as excellent.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H4 5Results of logistic regression statistical models for negative patient ratings of quality of care (16% of respondents). Cohort: 80% complete Emergency cohort (excludes appointments and pain level=unknown severity). Weighted n=14,572.Bivariate Predisposing Reduced Set (r2=0.0560)Predisposing, Presenting and Pain (r2=0.0770)Predisposing, Presenting, Pain and Step 1 Experience variable (r2=0.3497)Predisposing, Presenting, Pain and Step 2 Experience variable (r2=0.2687)Predisposing, Presenting, Pain and Step 3 Experience variable (r2=0.3222)Predisposing, Presenting, Pain and Step 4 Experience variable (r2=0.2269)Predisposing, Presenting, Pain and Steps 1, 2, 3 and 4 Experience variables (r2=0.4260)Predisposing, Presenting, Pain and Experience up to Step 4 PLUS interaction (r2=0.4274)Variable Wt N % Odds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valuePredisposingAge <.0001 <.0001 <.0001 0.0004 <.0001 <.0001 <.0001 0.0518 0.04720 to 12* 1,878 12.7  13 to 19 947 6.4 1.69 1.63 1.43 1.17 1.23 1.24 1.33 1.06 1.0820 to 34 1,839 12.5 1.74 1.58 1.37 1.15 1.13 1.00 1.46 1.02 1.0135 to 49 2,556 17.3 1.34 1.11 0.92 0.90 0.91 0.87 1.10 0.86 0.8550 to 64 3,260 22.1 0.79 0.62 0.54 0.78 0.57 0.69 0.73 0.85 0.8565 to 74 1,924 13.0 0.50 0.38 0.34 0.59 0.42 0.48 0.47 0.66 0.6675 and older 2,346 15.9 0.59 0.40 0.38 0.57 0.42 0.40 0.49 0.58 0.57Sex  <.0001 0.0013 0.003 0.9032 0.0541 0.2493 0.0371 0.8412 0.8404Missing 8 0.1 0.16 0.10 0.08 0.62 0.24 0.71 0.32 2.07 2.14Female 7,874 53.4 1.32  1.20 1.18 1.01 1.18 1.14 1.18 1.04 1.03Male* 6,869 46.6     Self-reported ethnicity  <.0001 <.0001 <.0001 0.1555 <.0001 0.0001 <.0001 0.4887 0.4851Missing 623 4.2 1.70 1.49 1.51 1.19 1.16 1.23 1.48 1.08 1.09Aboriginal 494 3.4 2.60  1.93 1.92 1.39 1.74 1.79 1.97 1.43 1.43Asian 1,194 8.1 2.37  2.02 2.03 1.24 1.84 1.45 1.30 1.09 1.09Caucasian* 11,780 79.9 1.70  Other 659 4.5 1.66 1.38 1.38 1.11 1.47 1.58 1.07 1.17 1.16Self-reported general health status  <.0001 <.0001 <.0001 0.0072 <.0001 0.0039 <.0001 0.2055 0.1776Missing 221 1.5 1.53 1.86 1.80 1.15 2.16 1.35 1.94 1.22 1.26Poor 848 5.7 1.80 2.71 2.64 1.30 2.22 1.67 2.26 1.43 1.41Fair 2,381 16.1 1.36 2.10 2.18 0.97 1.74 1.38 1.94 1.04 1.04Good 4,619 31.3 1.06 1.41 1.42 0.75 1.17 1.00 1.34 0.84 0.84Very Good 4,356 29.5 0.89 1.00 1.01 0.75 0.94 0.91 0.95 0.89 0.86Excellent* 2,325 15.8  Days in bed due to illness/injury, in past month<.0001 <.0001 0.0405 0.1887 0.5051 0.5534 0.1130 0.8753 0.8865Missing 305 2.1 1.75 1.53 1.38 1.32 1.49 1.31 1.39 1.40 1.43None* 7,937 53.8   One–three days 2,918 19.8 1.48 1.18 1.09 1.09 0.97 1.08 1.04 1.07 1.07Four–ten days 2,279 15.4 1.88 1.43 1.18 1.11 1.00 1.16 1.19 1.00 1.01More than ten days 1,313 8.9 2.16  1.72 1.35 1.45 1.07 1.20 1.33 1.11 1.11I N  P U R S U I T  O F  Q U A L I T Y4 6PresentingDay/time of visit <.0001 0.0133 0.2288 0.1030 0.3721 0.0310 0.1298 0.1556Missing 13 0.1 0.27 0.32 1.31 0.60 1.02 0.56 2.45 2.39Weekday, 00:00–06:591,701 11.5 1.43 1.36 1.32 1.35 1.37 1.26 1.37 1.36Weekday, 07:00–17:596,348 43.0 1.01 1.11 1.45 1.13 1.30 1.06 1.50 1.50Weekday, 18:00–23:592,046 13.9 1.01 1.02 1.09 1.06 1.14 0.99 1.08 1.08Weekend, 00:00–06:59806 5.5 1.57 1.47 1.44 1.56 1.63 1.62 1.77 1.75Weekend, 07:00–17:592,922 19.8 1.00 1.08 1.22 1.03 1.20 1.01 1.19 1.21Weekend, 18:00–23:59*914 6.2Reason for visit <.0001 <.0001 0.0092 0.0098 0.0460 0.0279 0.3593 0.3833Missing 745 5.0 1.05 0.93 0.81 1.05 0.93 0.99 0.88 0.87It clearly was an emergency6,547 44.4 1.57 1.19 1.05 1.17 1.14 1.02 0.94 0.94I was told to go by a health professional/ BCNurse Line2,794 18.9 1.68 1.54 1.48 1.44 1.36 1.27 1.22 1.22I didn't know if my health condition was an emergency or not*2,469 16.7  There were no other options/didn't know where else to go1,562 10.6 1.81 1.51 1.23 1.55 1.51 1.38 1.25 1.22Other 635 4.3 1.23 1.04 0.80 1.10 1.10 0.96 1.02 1.02Self-rated seriousness <.0001 0.0322 0.0296 0.7918 0.0043 0.0376 0.1691 0.1698Missing 410 2.8 1.59 1.37 1.39 1.13 1.34 1.21 1.11 1.09Extremely serious 1,621 11.0 1.82 1.39 1.59 1.25 1.68 1.43 1.51 1.53Very serious 3,959 26.8 1.67 1.32 1.70 1.14 1.63 1.16 1.43 1.45Moderately serious 5,577 37.8 1.38 1.19 1.33 1.08 1.27 1.10 1.14 1.16Slightly serious 2,505 17.0 1.03 0.95 1.07 1.00 0.99 0.86 0.96 0.97Not at all serious* 679 4.6Canadian Triage Acuity Score <.0001 0.0001 0.0794 0.0359 0.0415 0.9155 0.1853 0.2271Missing 5,451 37.0 0.94 0.87 0.86 0.86 0.90 1.00 0.92 0.93Level 1 – Resuscitation28 0.2 0.58 0.64 0.67 0.39 0.37 0.82 0.14 0.14Level 2 – Emergent 1,052 7.1 0.54 0.56 0.76 0.55 0.71 0.80 0.88 0.91Level 3 – Urgent 3,789 25.7 1.22 1.10 1.08 0.89 1.13 0.98 0.90 0.92Level 4 – Less urgent3,686 25.0 1.16 1.09 1.13 0.99 1.09 0.98 1.05 1.07Level 5 – Non-urgent*744 5.1Pain levelIn pain during encounter <.0001 <.0001 0.0102 0.0034 0.0006 <.0001 0.6358 0.0922Severe pain 3,738 25.3 2.69 2.10 1.40 1.40 1.52 1.65 0.94 See inter-actionModerate/mild pain 5,324 36.1 1.60 1.39 1.25 1.12 1.23 1.23 1.06No pain/missing* 5,689 38.6U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H4 7ExperienceCourtesy of ED staff (Step 1) <.0001 <.0001 <.0001 <.0001Missing 176 1.2 31.36 31.70 8.10 See inter-actionPoor 423 2.9 >999.9 1275.83 71.85Fair 1,354 9.2 395.13 356.15 31.85Good 3,348 22.7 26.06 25.04 4.04Very Good 4,937 33.5 3.53 3.53 1.48Excellent* 4,513 30.6Got all services needed (Step 2) <.0001 <.0001 <.0001 <.0001Missing 198 1.3 10.51 11.82 3.31 3.41Yes, completely* 8,753 59.3Yes, somewhat 4,509 30.6 15.65 13.06 3.33 3.37No 1,290 8.7 108.03 91.09 13.52 13.54How well doctors and nurses worked together (Step 3)<.0001 <.0001 <.0001 <.0001Missing 357 2.4 53.88 62.90  5.08 5.12Poor 402 2.7 >999.9 943.22  19.19 19.15Fair 1,209 8.2 269.307 227.73  9.79 9.85Good 3,654 24.8 29.27 26.99  2.88 2.91Very Good 5,131 34.8 2.87 2.87  1.09 1.10Excellent* 3,998 27.1Waited too long to see ED doctor (Step 4)<.0001 <.0001 <.0001 <.0001Missing 254 1.7 3.87 3.90 1.77 1.89Yes, defi nitely 2,659 18.0 35.30 30.10 7.22 7.37Yes, somewhat 4,210 28.5 7.07 6.32 2.66 2.66No* 7,628 51.7InteractionPain level * Courtesy of ED staff 0.0310Severe pain, missing courtesy 32.74Severe pain, poor courtesy 207.27Severe pain, fair courtesy 44.75Severe pain, good courtesy 7.72Severe pain, very good courtesy 4.26Severe pain, excellent courtesy 1.68Moderate/mild pain, missing courtesy 9.37Moderate/mild pain, poor courtesy 92.83Moderate/mild pain, fair courtesy 75.54Moderate/mild pain, good courtesy 8.66Moderate/mild pain, very good courtesy 3.01Moderate/mild pain, excellent courtesy 3.56No pain, missing courtesy 13.40No pain, poor courtesy 175.72No pain, fair courtesy 87.82No pain, good courtesy 8.51No pain, very good courtesy 2.27No pain, excellent courtesy 1.00    * Reference category.I N  P U R S U I T  O F  Q U A L I T Y4 8APPENDIX  EResults of logistic regression statistical model to predict patient views on waiting too long to see a doctorAft er accounting for patients’ predisposition and charac-teristics of their presentation to emergency departments, the issue most highly associated with their view that the wait to see a doctor was defi nitely too long was—not surprisingly—the time they waited to see a doctor.  Other     * Th e pseudo r2 value of the logistic regression model with predisposing and presenting characteristics, including pain, was 0.06. Th e forward stepwise logistic regression model initially included wait time to see doctor (pseudo r2 value of 0.25), then availability of nurses (pseudo r2 value of 0.30), then not getting help when needed (pseudo r2 value of 0.31) and then ‘waited to long for test results’ (pseudo r2 value of 0.33). Th e full forward stepwise logistic regression model had a pseudo r2 value of 0.37 and included 24 types of survey items.types of experiences that underlie patient views on wait-ing too long are, in rank order, the availability of nurses, not getting help when needed, and waiting too long for test results.*Bivariate Predisposing Reduced Set (r2=0.0309)Predisposing, Presenting and Pain (r2=0.0570)Predisposing, Presenting, Pain and Step 1 Experience variable (r2=0.2508)Predisposing, Presenting, Pain and Step 2 Experience variable (r2=0.1818)Predisposing, Presenting, Pain and Step 3 Experience variable (r2=0.1589)Predisposing, Presenting, Pain and Step 4 Experience variables (r2=0.1367)Predisposing, Presenting, Pain and Steps 1, 2, 3 and 4 Experience variables (r2=0.3264)Variable Wt N % Odds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valueOdds ratioOverall p valuePredisposingAge <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.00010 to 12* 1,878 12.7  13 to 19 947 6.4 1.38 1.43 1.35 1.11 1.18 1.28 1.21 1.0020 to 34 1,839 12.5 1.22 1.18 1.10 0.85 0.95 0.95 1.13 0.8035 to 49 2,556 17.3 0.99 0.92 0.84 0.70 0.82 0.87 0.89 0.7550 to 64 3,260 22.1 0.68 0.63 0.59 0.51 0.67 0.67 0.63 0.5965 to 74 1,924 13.0 0.50 0.47 0.46 0.38 0.60 0.59 0.52 0.5675 and older 2,346 15.9 0.58 0.52 0.52 0.37 0.62 0.62 0.53 0.46Sex  0.0160 0.0481 0.0631 0.3467 0.4990 0.3154 0.1348 0.8063Missing 8 0.1 0.12 0.10 0.09 0.28 0.29 0.18 0.10 0.54Female 7,874 53.4 1.12  1.08 1.06 1.06 0.97 0.99 1.04 0.96Male* 6,869 46.6     Education  0.0160 0.0481 0.0631 0.3467 0.4990 0.3154 0.1348 0.8063Missing 8 0.1 0.12 0.10 0.09 0.28 0.29 0.18 0.10 0.54Public school 7,874 53.4 1.12  1.08 1.06 1.06 0.97 0.99 1.04 0.96High school 6,869 46.6     College, trade or technical school3,968 26.9 0.74 0.77 0.76 0.76 0.78 0.76 0.74 0.72University undergraduate1,716 11.6 0.89 0.93 0.92 0.93 0.91 0.89 0.94 0.90Post university/ graduate education*1,294 8.8  Results of logistic regression statistical model to predict patient views on defi nitely waiting too long to see a doctor  (18% of respondents). Cohort: 80% complete Emergency cohort (excludes appointments and pain level=unknown severity). Weighted n=14,497.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H4 9Self-reported ethnicity  <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001Missing 623 4.2 1.30 1.15 1.17 1.09 1.08 1.13 1.14 1.08Aboriginal 494 3.4 1.56  1.31 1.32 1.35 0.99 1.23 1.23 1.09Asian 1,194 8.1 2.37  2.07 2.13 1.88 2.02 1.84 1.90 1.74Caucasian* 11,780 79.9   Other 659 4.5 1.70 1.45 1.49 1.70 1.78 1.43 1.37 1.83Self-reported general health status  0.1933 0.0002 <.0001 0.0145 0.5834 0.0830 0.0006 0.6520Missing 221 1.5 0.91 1.09 1.03 1.04 0.79 0.88 0.96 0.81Poor 848 5.7 1.22 1.76 1.78 1.53 1.10 1.40 1.62 1.05Fair 2,381 16.1 1.02 1.46 1.54 1.50 1.05 1.33 1.51 1.17Good 4,619 31.3 0.91 1.17 1.18 1.16 0.92 1.10 1.16 0.98Very Good 4,356 29.5 0.93 1.03 1.05 1.02 0.92 1.02 1.01 0.94Excellent* 2,325 15.8  Days in bed due to illness/injury, in past month<.0001 0.0376 0.8703 0.9136 0.9985 0.8779 0.9361 0.4372Missing 305 2.1 1.28 1.28 1.16 1.04 1.05 0.99 1.20 0.94None* 7,937 53.8    One–three days 2,918 19.8 1.28 1.10 1.03 1.04 1.01 0.93 0.98 0.96Four–ten days 2,279 15.4 1.37 1.18 1.00 0.96 0.99 0.92 0.99 0.92More than ten days 1,313 8.9 1.46  1.33 1.10 0.92 0.99 0.93 0.99 0.75PresentingDay/time of visit 0.0195 0.0731 0.2153 0.1124 0.0157 0.0480 0.0268Missing 13 0.1 0.23 0.31 0.37 0.74 0.29 0.26 0.48Weekday, 00:00–06:591,701 11.5 1.28 1.30 1.04 1.29 1.42 1.20 1.13Weekday, 07:00–17:596,348 43.0 1.01 1.14 1.01 1.22 1.27 1.05 1.12Weekday, 18:00–23:592,046 13.9 1.05 1.09 1.07 1.15 1.17 1.02 1.10Weekend, 00:00–06:59806 5.5 0.94 0.90 0.68 0.82 0.89 0.76 0.56Weekend, 07:00–17:592,922 19.8 1.02 1.13 0.99 1.16 1.25 1.08 1.10Weekend, 18:00–23:59*914 6.2Reason for visit <.0001 <.0001 0.0080 0.0040 0.0004 0.0010 0.0388Missing 745 5.0 0.89 0.94 0.94 1.02 0.70 0.80 0.79It clearly was an emergency6,547 44.4 1.59 1.35 1.32 1.33 1.17 1.31 1.17I was told to go by a health professional/ BCNurse Line2,794 18.9 1.70 1.61 1.44 1.54 1.46 1.50 1.31I didn't know if my health condition was an emergency or not*2,469 16.7  There were no other options/didn't know where else to go1,562 10.6 1.61 1.41 1.52 1.42 1.39 1.34 1.46Other 635 4.3 1.33 1.20 1.24 1.32 1.22 1.23 1.32Self-rated seriousness 0.0015 0.2660 0.5589 0.4117 0.3987 0.3453 0.5773Missing 410 2.8 1.29 1.26 1.22 1.19 1.11 1.16 1.04Extremely serious 1,621 11.0 1.36 1.12 1.01 1.13 0.83 1.11 0.83Very serious 3,959 26.8 1.50 1.25 1.10 1.23 1.04 1.21 0.97Moderately serious 5,577 37.8 1.25 1.10 1.01 1.05 0.98 1.05 0.90Slightly serious 2,505 17.0 1.07 1.00 0.89 1.01 0.91 0.97 0.80Not at all serious* 679 4.6I N  P U R S U I T  O F  Q U A L I T Y5 0Canadian Triage Acuity Score <.0001 <.0001 0.0001 <.0001 <.0001 <.0001 0.0004Missing 5,451 37.0 0.77 0.76 0.81 0.73 0.79 0.70 0.74Level 1 – Resuscitation28 0.2 0.57 0.64 1.36 0.79 0.59 0.79 0.94Level 2 – Emergent 1,052 7.1 0.41 0.43 0.45 0.44 0.41 0.39 0.39Level 3 – Urgent 3,789 25.7 1.24 1.18 0.88 1.12 1.17 1.06 0.83Level 4 – Less urgent3,686 25.0 1.26 1.18 1.05 1.10 1.19 1.08 0.95Level 5 – Non-urgent*744 5.1Pain levelIn pain during encounter <.0001 <.0001 <.0001 <.0001 0.0002 <.0001 0.1505Severe pain 3,738 25.3 2.08 1.85 1.60 1.51 1.44 1.60 1.18Moderate/mild pain 5,324 36.1 1.44 1.31 1.16 1.15 1.16 1.18 0.98No pain/missing* 5,689 38.6ExperienceWait for doctor (Step 1) <.0001 <.0001  <.0001Missing 383 2.6 16.32 16.36  9.89Did not wait at all* 1,229 8.3  Less than 1/2 hour 5,821 39.5 1.16 1.09  1.07Between 1/2 hour and 1 hour3,774 25.6 3.72 3.16  2.141 to 2 hours 2,082 14.1 12.96 11.17  6.48More than 2 hours 1,462 9.9 59.56 56.23  28.77Availability of nurses (Step 2) <.0001 <.0001  <.0001Missing 323 2.2 3.77 3.83  1.63Poor 999 6.8 36.05 30.77  6.12Fair 2,288 15.5 9.29 8.33  2.76Good 4,179 28.3 3.51 3.27  1.74Very Good 4,239 28.7 1.25 1.20  0.89Excellent* 2,723 18.5   Did not get needed help (Step 3) <.0001 <.0001  <.0001Missing 285 1.9 3.91 4.25  2.73Yes, often 796 5.4 17.79 16.18  4.72Yes, sometimes 2,642 17.9 6.73 5.98  2.53No* 7,946 53.9  Did not need help 3,083 20.9 1.66 1.68  1.40Waited too long for test results (Step 4) <.0001 <.0001 <.0001Missing 116 0.8 4.89 4.97 1.86Yes, defi nitely 958 6.5 17.99 15.79 5.06Yes, somewhat 2,177 14.8 3.36 2.97 1.43No* 6,782 46.0Did not get any tests 4,718 32.0 2.29 2.22 1.67    * Reference category.U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H5 1APPENDIX  FRelationship between wait times and patient views on waiting too long Patient views on waiting too long to see a doctor in an emergency departmentPatient self-reported wait time in an emergency department in 2007TotalDid not wait < 1/2 hour ½ to 1 hour 1 to 2 hours > 2 hoursYes, defi nitely Frequency (weighted cases) 47.441 260.69 493.76 712.41 1017.7 2532Percent (%) 0.33 1.84 3.48 5.02 7.18 17.86Row percent (%) 1.87 10.30 19.50 28.14 40.19Column percent (%) 3.94 4.53 13.23 34.69 70.95Yes, somewhatFrequency (weighted cases) 67.866 913.02 1750.3 1038.7 361.1 4131Percent (%) 0.48 6.44 12.34 7.32 2.55 29.13Row percent (%) 1.64 22.10 42.37 25.14 8.74Column percent (%) 5.63 15.87 46.89 50.57 25.17NoFrequency (weighted cases) 1089.8 4580.8 1488.6 302.74 55.678 7517.6Percent (%) 7.69 32.30 10.50 2.13 0.39 53.01Row percent (%) 14.50 60.93 19.80 4.03 0.74Column percent (%) 90.43 79.60 39.88 14.74 3.88Total Frequency (weighted cases) 1205.09 5754.53 3732.61 2053.82 1434.45 14180.5Percent (%) 8.50 40.58 26.32 14.48 10.12 100.00Frequency missing=570I N  P U R S U I T  O F  Q U A L I T Y5 2APPENDIX  GResults of statistical analyses to account for differences in patient characteristics and their presentation at emergency departments on differences in patient ratings of overall quality of careWe found variation across types of hospitals and health regions about patient views on overall quality and ratings on the factors that matter to them. Simultaneously, we found that patients’ characteristics and their presentation at emergency departments had some infl uence on patient ratings about overall quality of care. Since patients from diff erent types of hospitals and health regions diff er in these characteristics, we used statistical methods to risk-adjust performance metrics. Th is was done to determine the degree to which variation in these characteristics underlies diff erences between hospitals and regions on patient views on overall quality of care in emergency departments. Importantly, we found that the rank order of highest and lowest performance across health regions remained un-changed even aft er sophisticated analysis was conducted to account for diff erences in the characteristics of patients and their presentation at emergency departments. Th is is true for other measures profi led in this report. Th e crude and adjusted mean or average patient rating of overall quality of care* for each health authority are slightly dif-ferent but the relative rank of highest and lowest remains unchanged aft er risk-adjustment. Th us, accounting for diff erences between health regions in patient and present-ing characteristics do not mitigate regional diff erences or alter relative rankings.       * Means were calculated by assigning a value of 1 to patient ratings of poor, 2 to patient ratings of fair, 3 to patient ratings of good, 4 to patient ratings of very good and 5 to patient ratings of excellent. Th us, patient ratings of overall quality of care have the potential to range from one to fi ve.   ** Adjusted for age group, gender, education level, self-reported ethnicity, self-reported general health status, days in bed due to injury/illness in past month, day and time of emergency department visit, reason for emergency department visit, self-rated seriousness, Canadian Triage Acuity Score and pain level.Mean and adjusted mean patient ratings of overall quality of care, by health region in 2007Overall patient rating of overall quality of careMean Adjusted mean**Interior Health 3.81 3.85Fraser Health 3.43 3.49Vancouver Coastal Health 3.69 3.79Vancouver Island Health 3.72 3.75Northern Health 3.72 3.75U B C  C E N T R E  F O R  H E A LT H  S E R V I C E S  A N D  P O L I C Y  R E S E A R C H5 3I N  P U R S U I T  O F  Q U A L I T Y5 4UBC Centre for Health Services and Policy ResearchThe University of British Columbia201-2206 East MallVancouver, B.C. Canada  V6T 1Z3Tel:  604.822.4969Fax:  604.822.5690Email: enquire@chspr.ubc.cawww.chspr.ubc.caAdvancing world-class health services and policy research, training and data resources on issues that matter to Canadians

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0048321/manifest

Comment

Related Items