Stakeholder views regarding patient discharge from intensive care : Suboptimal quality and opportunities for improvement

*Co-first authors, both of whom contributed equally to the work 1Department of Medicine, University of Calgary and Alberta Health Services – Calgary Zone; 2Institute for Public Health; 3Department of Community Health Sciences, University of Calgary; 4Department of Critical Care Medicine, University of Calgary and Alberta Health Services – Calgary Zone, Calgary, Alberta Correspondence: Dr Henry T Stelfox, Teaching Research & Wellness Building, University of Calgary, 3280 Hospital Drive Northwest, Calgary, Alberta T2N 4Z6. Telephone 403-944-2334, fax 403-283-9994, e-mail tstelfox@ucalgary.ca Transitions of patient care among providers are vulnerable periods in health care delivery (1). These transfers of accountability and responsibility for patient care require reliable and valid communication of patient data (eg, diagnoses, tests, treatments, goals of care) to ensure continuity of care (2-5). Patients deserve and expect seamless care across the health care continuum (6); however, evidence suggests that transitions of care are associated with medical errors (7), adverse events (8), increased health care costs (9) and poor patient satisfaction with care (10). The discharge of patients from the intensive care unit (ICU) to a medical or surgical hospital ward is one of the most challenging and high-risk transitions of care. Up to 18% of adverse events experienced by ICU patients are attributed to the discharge process (11). Medication errors are common during ICU discharge (12), and are associated with poor communication (13) and poor medication reconciliation (accurate and complete transfer of medication information at interfaces of care) (7,14,15). Discharge from the ICU should ensure that patients are discharged using a process that optimizes patient care and reduces the risk of adverse events including hospital mortality. However, there is no consensus on an ideal ICU discharge model (16), and there appears to be important variation in the morbidity and mortality of discharged patients (17-19). Given the challenges associated with ICU discharge and limited available literature describing current discharge practices, we conducted a survey in three countries as an essential first step toward identifying opportunities for improvement. We surveyed ICU administrators (unit manager, director) for an institutional perspective and members of professional care associations for an ICU provider perspective. Our objectives were to understand the organization of ICU discharge; explore stakeholder needs and expectations surrounding ICU discharge; and determine stakeholder perceptions of quality of care during ICU discharge and opportunities for improvement. ORIGINAL ARTICLE


Stakeholder views regarding patient discharge from intensive care: Suboptimal quality and opportunities for improvement
Pin Li MD MSc 1,2 *, Jamie M Boyd BA 3 *, William A Ghali MD MPH 1,2,3 , Henry T Stelfox MD PhD 1,2,3,4   *Co-first authors, both of whom contributed equally to the work 1 Department of Medicine, University of Calgary and Alberta Health Services -Calgary Zone; 2 Institute for Public Health; 3  T ransitions of patient care among providers are vulnerable periods in health care delivery (1).These transfers of accountability and responsibility for patient care require reliable and valid communication of patient data (eg, diagnoses, tests, treatments, goals of care) to ensure continuity of care (2)(3)(4)(5).Patients deserve and expect seamless care across the health care continuum (6); however, evidence suggests that transitions of care are associated with medical errors (7), adverse events (8), increased health care costs (9) and poor patient satisfaction with care (10).
The discharge of patients from the intensive care unit (ICU) to a medical or surgical hospital ward is one of the most challenging and high-risk transitions of care.Up to 18% of adverse events experienced by ICU patients are attributed to the discharge process (11).Medication errors are common during ICU discharge (12), and are associated with poor communication (13) and poor medication reconciliation (accurate and complete transfer of medication information at interfaces of care) (7,14,15).Discharge from the ICU should ensure that patients are discharged using a process that optimizes patient care and reduces the risk of adverse events including hospital mortality.However, there is no consensus on an ideal ICU discharge model (16), and there appears to be important variation in the morbidity and mortality of discharged patients (17)(18)(19).Given the challenges associated with ICU discharge and limited available literature describing current discharge practices, we conducted a survey in three countries as an essential first step toward identifying opportunities for improvement.We surveyed ICU administrators (unit manager, director) for an institutional perspective and members of professional care associations for an ICU provider perspective.Our objectives were to understand the organization of ICU discharge; explore stakeholder needs and expectations surrounding ICU discharge; and determine stakeholder perceptions of quality of care during ICU discharge and opportunities for improvement.

Study questions
ICU administrators and providers were queried about their experiences and perceptions of ICU discharge.Stakeholders were asked three specific questions: 1.How patient discharge from ICU is performed (structure, process) at their institution?2. How is information provided to patients and families on discharge from ICU? 3. What is the quality of care during patient discharge from ICU and how can it be improved?

Study participants
Canadian ICUs were identified by reviewing Internet resources and government health sites, and contacting university medical schools and affiliated hospitals.'ICU' was defined as a distinct hospital specialty care unit staffed by specialized professionals where immediate and continuous life-sustaining treatment (minimum capabilities include invasive monitoring, vasoactive medications and mechanical ventilation) is provided to patients who have life-threatening conditions regardless of subspecialty (20).
Step-down units were excluded.Key informants (medical director, unit manager) were identified from a search of their hospitals' websites, contacts of the project team and local telephone directories.To obtain an ICU provider perspective, members of professional critical care associations in Canada, the United States (US) and the United Kingdom (UK) were surveyed using membership distribution lists from the Canadian Association of Critical Care Nursing (CACCN) (21), Society for Critical Care Medicine (SCCM) (22) and the Intensive Care Society (ICS) (23).These professional critical care associations were selected because their membership primarily comprises ICU providers working in health care systems with advanced critical care capacities, and the associations had protocols for posting links to electronic surveys on association newsletters and websites.

Survey design and testing
Items for the survey instrument were generated based on a structured literature review.To identify relevant articles, the Medline, EMBASE, CINAHL and the Cochrane Central Register of Controlled Trials databases were searched using a combination of the following terms: "critical care", "intensive care", "discharge plan", "transfer process", "patient discharge" and "patient transfer".The survey was designed to obtain basic information about ICU organizational characteristics (12 questions), current discharge practices (11 questions), communication with patients and families during discharge (three questions) and the quality of ICU discharge and opportunities for improvement (seven questions).The survey underwent an iterative development process (24).Members of the research team serially reviewed the instrument, eliminating redundant items and revising retained items.
The survey was then pilot tested with three ICU managers and providers from local ICU facilities to evaluate the clinical sensibility of the revised items and further modified based on their suggestions.The final survey was comprised of 31 close-ended questions and two openended questions, and was documented to take less than 10 min to complete (Appendixes 1 and 2).

Survey administration ICU administrators:
Survey distribution began February 23, 2012.The survey was mailed to the medical director or unit manager identified for each Canadian ICU.Return postage was included.If the initial survey was not returned, up to two follow-up mail surveys were sent, followed by an e-mail reminder (when available) in three-week intervals.ICU providers: Professional associations provided resources and protocols for distribution of surveys to members in Canada, the US and the UK.A descriptive paragraph and link to the electronic survey was included in three newsletter e-mails sent to members of CACCN (Canada) and the link was also included on the associations' website.
SCCM (US) members received the link in two newsletter e-mails.ICS members (UK) were able to access the link through the associations' website.Survey distribution began in February 2012 and closed July 31, 2012.

Statistical analysis
The strategy for the primary analysis was to describe ICU discharge practices by answering each of the three specific study questions.Survey responses were summarized using nominal (proportions) and ordinal (median and interquartile range [IQR]) measures.Formal hypothesis testing was not performed.Using the same descriptive measures used for the primary analysis, four prespecified subgroup analyses were performed according to ICU education status (teaching versus nonteaching), location (urban versus rural), unit size (number of funded beds) and perceived quality of discharge (low versus medium versus high).Summary statistics were calculated using SPSS version 17.0 (IBM Corporation, USA) (25).Ethics approval was obtained from the Conjoint Health Research Ethics Board at the University of Calgary, Calgary, Alberta.

Survey response
Between February 2012 and July 2012, the survey was sent to 140 ICU administrators, of which 118 (representing 114 ICUs) (81%) responded.During the same time period, a total of 737 ICU providers from critical care professional associations in Canada (n=311), the US (n=276), the UK (n=90) and other countries (n=60) responded (denominator unknown becasuse the survey link was posted on association newsletters and websites and association memberships were not disclosed).

Respondent characteristics
The characteristics of the respondents and their ICU facilities are summarized in Table 1.Respondents to the ICU administrator survey self-identified primarily as managers (eg, unit manager, patient care manager) who worked at teaching hospitals and in ICUs that served medical and surgical patient populations, with a nurse-to-patient ratio of 1:1 or 1:2.Respondents to the ICU provider survey self-identified primarily as nurses or physicians who similarly primarily worked at teaching hospitals and in ICUs that served surgical and medical patient populations.The majority of ICU administrators and ICU providers reported that their facility discharged ≥10 patients per week.

How ICU discharge is performed
Table 2 summarizes the reported ICU discharge structures and processes.Both ICU administrators and ICU providers reported that ICU physicians were the primary responsible physician for decisions regarding patient discharge including assessing patient safety for discharge (94% and 96%, respectively); ensuring the receiving service has all pertinent patient information (92% and 87%); and deciding the appropriate service to receive the discharged patient (81% and 87%).One-half (52%) of ICU administrators and ICU providers (51%) reported the use of a standardized ICU discharge protocol at their facility.Formal communication (using a systematic approach such as sequentially proceeding by systems or problems, in contrast to informal, unstructured communication) between nurses (84% and 88%) was reported to be more common than between physicians (60% and 55%) during patient discharge.Telephone was reported as the most frequently used mode for discharge communication (70% and 74%).

How information is provided to patients and families on discharge from the ICU
Table 3 summarizes respondents' impressions of the modes of communication used with patients and family members for ICU discharge.Verbal communication of discharge information was perceived to be the most common mode of provider-patient communication by ICU administrators (99%) and ICU providers (95%).A minority reported providing patients and families with written communication before discharge (30% and 20%).Very few ICU providers (3%) reported that no information is provided to patients and families at discharge.

Quality of care during patient discharge from the ICU and opportunities for improvement
More than one-half (62%) of ICU administrators and ICU providers (51%) reported the use of quality indicators in their ICU to measure the quality of ICU discharge.Readmission rates (<24 h, <48 h, <72 h, time not specified) were the most commonly reported quality indicator (41% and 39%), with the time of discharge (13% and 5%), length of stay (8% and 3%), post-ICU discharge follow-up (5% and 7%) and hospital mortality (3% and 5%) less frequently identified.The median percentage of patients reported to be readmitted per month by ICU administrators was 3 (IQR 2 to 5) and by ICU providers was 5 (IQR 4 to 10). Figure 1 summarizes respondents' ratings of ICU discharge quality.The distribution of ratings was similar for administrators and providers.Table 4 outlines potential areas for improvement to current ICU discharge practices, with the areas of information provided to patients and families (71% of ICU administrators and 59% ICU providers), collaboration among hospital units (65% and 66%), patient or family participation during discharge (67% and 57%), continuity of care (64% and 67%) and completeness of information (62% and 64%) being among the most frequently reported.When asked about an electronic discharge tool, the majority of respondents indicated that it could improve patient care (94% and 85%).Suggested elements for a tool to facilitate patient discharge from the ICU included a summary of ICU stay (33% and 22%), medications and medication reconciliation (41% and 35%), and treatment and care plans (42% and 54%).

Subgroup analyses
Subgroup analyses of respondents according to ICU teaching status, location and number of funded beds produced similar results.ICU discharge practices, communication with patients and families, and suggested opportunities for improvement varied according to respondents' ratings of ICU discharge quality.Both ICU administrators and providers, who rated the quality of ICU discharge as high or very high compared with those who rated it as low or very low, were more likely to report that ICU discharge at their facility involved formalized nurse-to-nurse communication, formalized physician-to-physician communication, medication reconciliation, the use of standardized ICU discharge protocols and checklists, and that written communication of information was provided to patients and families (Figure 2).Conversely, ICU administrators and providers who rated the quality of ICU discharge as low or very low were more likely to identify opportunities for improvement in the ICU discharge process.

DISCUSSION
Our study provides the first description of ICU discharge practices from the perspective of Canadian ICU administrators and ICU providers in Canada, the US and the UK.The findings reveal complexity in the ICU discharge process, considerable practice variation, perception of only medium quality and several potential opportunities for improvement.Approximately one-half of respondents reported the use of a protocol or checklist during the discharge process.Most respondents rated current ICU discharge practices to be of medium quality, while only one-half reported the use of quality indicators to monitor discharge quality.Opportunities identified for improvement included the information provided to patients and families, collaboration between hospital units and continuity of care.An evidence-informed electronic discharge tool was reported as one instrument that could be used to address these opportunities for improvement.
There are opportunities to improve the quality of patient discharge from the ICU.The limited existing literature describes variations in the models used for ICU discharge (16), and the associated risks of morbidity and mortality for discharged patients (18,19,26).Heidegger et al (18) surveyed Swiss ICUs and reported that discharge practices were heterogeneous and primarily directed by individual ICU providers with limited use of written guidelines.The medical directors of these Swiss ICUs reported different discharge decisions when presented with the same clinical scenarios of patients with chronic obstructive lung disease, altered level of consciousness, trauma, sepsis and myocardial infarction being considered for discharge.Communication between providers has been documented to be infrequent, incomplete and of poor quality (2,27), with Li et al (27) reporting from an observational study that ICU and ward physicians verbally communicated in only 15% to 25% of ICU discharges.To compound matters, provider-patient communication during ICU discharge is frequently suboptimal, with the majority of patients and families indicating that more information is needed before they leave the ICU (10,27).These variations and inconsistencies in discharge practices are on a backdrop of wide-ranging rates of readmission and mortality in patients following ICU discharge (28).
Implementation of evidence-informed discharge practices is the simplest and most effective way to standardize and improve the quality of ICU discharge (29).What is the optimal structure for ICU discharge?Although the current literature does not tell us, it is likely to be multidimensional.Our study adds to the literature in three ways.First, it highlights ongoing important practice variation in ICU discharge processes.Most notably, it appears that discharge decision making is left to individual clinicians, with approximately one-half of all Canadian ICUs using a protocol to guide the discharge process.These results are similar to those reported by Heidegger et al (18) in their survey of Swiss ICUs in 2002, and suggests that challenges to ICU discharge may be similar across different health care systems and largely unchanged by recent innovations in technology and patient care.Second, our study highlights potential opportunities for improving ICU discharge, most notably the information provided to patients and families, collaboration among hospital units and continuity of care.While there is a large body of literature involving end-of-shift (4) or end-of-service (30) patient handoffs, there is relatively little known about patient discharge between units (31).For example, in our survey, the use of formal structured communication was reported less commonly between physicians than nurses, suggesting simple concrete opportunities for improvement.Third, the respondents to our survey described their needs for implementing best practices to facilitate the discharge process.They specifically indicated that these should include a summary of the patient stay in the ICU, a reconciled list of medications and a prescribed treatment plan.Similar opportunities exist to provide patients and families with information and facilitate their engagement during the transition process.

Figure 2) Intensive care unit discharge processes according to administrator and provider perceptions of the quality of current discharge practices. MD Medical doctor; RN Registered nurse
There were limitations to the present study.The unknown denominator for the provider responses raises the question of whether the perceptions of nonrespondents were similar to respondents.While we cannot answer this, the large number of provider respondents and the good response rate from administrators, suggests that our findings are likely to represent the perspective of many ICU administrators and providers.To ensure feasibility, we restricted the study participants to ICU administrators in Canadian facilities and providers from selected professional organizations.The respondents were primarily comprised of ICU managers, nurses and physicians.Although administrator and provider responses were similar, it is possible that other stakeholders (eg, medical directors, pharmacists, physical therapists), as well as administrators and providers from other countries and professional organizations, may have different perspectives.The strengths and weaknesses of survey methodology complement other research methods.For example, survey participants reported higher proportions of verbal communication between providers during handoff than documented in observational studies.This may represent recall or social desirability biases for which survey methodology is susceptible.Conversely, our survey solicited descriptions of current ICU discharge practices and stakeholder needs from a broader and more diverse group of administrators and providers than could practically be accomplished through direct observation or other research methods.

CONCLUSION
The discharge of patients from the ICU to a hospital ward is one of the most challenging and high-risk transitions of care.Our data indicate that the ICU discharge process is multidimensional, directed by local practice patterns and associated with important practice variation.Stakeholders perceive the quality of care provided during patient discharge to be only medium.Opportunities for improvement exist with priorities including the information provided to patients and families, collaboration among hospital units and continuity of care.Development and implementation of evidence-informed ICU discharge practices may help standardize key elements of the discharge process and help health care providers deliver the highest-quality care during these transitions.ACKNOWLEDGEMENTS: PL, JB, WG and HS contributed to the study's conception and design.JB collected and cleaned the data.PL, JB, WG and HS performed and interpreted the analyses and assisted in the successive revisions of the final manuscript.All four authors read and approved the final manuscript.The project is supported by a Seed Grant from the University of Calgary.Dr Stelfox is supported by a New Investigator Award from the Canadian Institutes of Health Research and a Population Health Investigator Award from Alberta Innovates Health Solutions.Dr Ghali is funded by a Senior Health Scholar Award from Alberta Innovates Health Solutions.Funding sources had no role in the design of the protocol and the authors are unaware of any conflicts of interest.The authors thank Elizabeth Fradgely for helping with instrument development.
FINANCIAL SUPPORT: This project was supported by a Seed Grant from the University of Calgary, Calgary, Alberta.

CONFLICTS OF INTEREST:
The authors have none to declare.

APPENDIX 1
International Survey of ICU Discharges -ICU Administrators Dear Intensive Care Unit Stakeholder, Attached is a short questionnaire designed to learn how intensive care units discharge patients.You are being contacted as you have extensive experience in intensive care.The survey takes less than ten minutes to complete, and your input is instrumental to this research project.
The primary aim of this research project is to gain detailed information on current ICU discharge practices, and to explore any standardized protocol that exists within Canada, United States, United Kingdom, Australia and New Zealand.You are being contacted to provide essential information about your ICU.
Discharge is broadly defined as the transfer of a patient out of an ICU.This includes discharge to home, or transition/transfer to another unit.
Discharge planning is one component of effective resource management, and is a pivotal point in a patient's transition from the ICU into other hospital facilities or into the community.Incorporating the Institute of Medicine's (IOM) aims of quality health care, the ICU discharge process should be safe, effective, efficient, and timely.However, there is little consensus on how to optimize patient safety and efficiency during ICU discharge.
Improving the discharge process could have substantial benefit in reducing readmission rates, decreasing adverse events, improving communication between hospital service providers, and reducing the use of costly ICU resources.
Your participation is voluntary and the results will be kept anonymous.The survey has full ethics approval and your participation implies consent.If you are not the appropriate person to complete this, please forward to the correct person.
If you prefer to complete this survey online, please go to: https://www.surveymonkey.com/s/ICUDischargeIf there are any concerns or suggestions, please do not hesitate to contact: Jamie Boyd (jamboyd@ucalgary.ca)Thank you for your participation.

Background Hospital and ICU Characteristics
This section is focused on the characteristics of your ICU and hospital.

Physician responsible for deciding the time a patient is discharged †
‡ Data represent respondents who reported using this mode of communication very frequently.MD Medical doctor; RN Registered nurse

TAble 4 Respondents identifying opportunities for improvement in the intensive care unit (ICU) discharge process Areas for improvement †
Is there a dedicated team or individual that transitions and monitors patients during discharge from the ICU to other hospital services?No Yes (please provide name of team or type of individual) __________________________________________________ 12. Please rank the following health services by the proportion of ICU patients discharged to that respective receiving health service (1=most discharges, 6=least discharges).____ Primary care, or family doctor ____ Surgical specialty ____ Medical specialty (including psychiatry, or internal medicine) ____ Discharged to another hospital ____ Discharged to home ____ Other Tools and Communication for ICU Discharge This page contains questions focused on the current ICU discharge process within your facility, and what tools are used to facilitate a seamless discharge.13.Is there a standardized ICU checklist or protocol completed before a patient is discharged to another unit or health service?Please select the frequency at each mode of communication is used during the discharge process.Swift score).Please enter the patient assessment score if used._____________________________________ 18. Are there any other tools used to facilitate a patient's discharge from ICU services?__________________________________________________ 19.Is there an electronic system currently used for patient care within the ICU? Yes No 20.Does this electronic system facilitate ICU patient discharge?Yes No