Background Reducing delays for patients who are safe to be discharged is important for minimising complications, managing costs and improving quality. Barriers to discharge include placement, multispecialty coordination of care and ineffective communication. There are a few recent studies that describe barriers from the perspective of all members of the multidisciplinary team.
Study objective To identify the barriers to discharge for patients from our medicine service who had a discharge delay of over 24 hours.
Methodology We developed and implemented a biweekly survey that was reviewed with attending physicians on each of the five medicine services to identify patients with an unnecessary delay. Separately, we conducted interviews with staff members involved in the discharge process to identify common barriers they observed on the wards.
Results Over the study period from 28 October to 22 November 2013, out of 259 total discharges, 87 patients had a delay of over 24 hours (33.6%) and experienced a total of 181 barriers. The top barriers from the survey included patient readiness, prolonged wait times for procedures or results, consult recommendations and facility placement. A total of 20 interviews were conducted, from which the top barriers included communication both between staff members and with the patient, timely notification of discharge and lack of discharge standardisation.
Conclusions There are a number of frequent barriers to discharge encountered in our hospital that may be avoidable with planning, effective communication methods, more timely preparation and tools to standardise the discharge process.
- discharge planning
- HEALTH SERVICES ADMINISTRATION & MANAGEMENT
- INTERNAL MEDICINE
Statistics from Altmetric.com
Delayed discharge is a long-standing problem in hospitals. As health reform and an ageing population lead to increased patient volumes, reducing delays in discharge is a cost-effective way of increasing capacity.1 Despite the fact that average length of stay (LOS) has decreased over the past 20 years, there is evidence that the proportion of unnecessary patient-days has remained stable.2
Discharging patients when they are medically safe to leave is important for quality of care, cost control and health outcomes. Patients who stay longer than necessary are at risk of iatrogenic complications, including infections and deep vein thromboses.3 Each hospital day is a considerable cost, especially when compared with the cost of outpatient facilities and home care. Patients with delayed discharges fill beds that could otherwise be used to manage patients who have acute care needs, leading to increased emergency department waiting times and patient overflow.4 ,5 Prolonged LOS is additionally associated with lower patient satisfaction and staff burnout.6 Patients who leave too early and without appropriate follow-up care, however, can be at increased risk for readmission and poorer outcomes.7
A number of causes for unnecessary patient delays have been proposed. Finding an appropriate discharge facility is challenging, particularly for elderly patients without available family caregivers and for patients with serious mental health conditions.3 ,5 ,8 Delays in procedures, tests and consults are common, especially over the weekend.9 Patients cleared for discharge late in the day often have a delayed discharge as transport and acceptance at a facility are difficult. Poor communication both between members of the care team and between patients and providers can delay confirmation of discharge and impact patient readiness for discharge.3 ,10 ,11 Finally, hospitals face financial pressure to avoid readmission, as patients who return within 30 days are often not eligible for insurance coverage and hospitals can face financial penalties from Medicare.12
There are only a few studies that identified the major barriers to discharge for general medicine patients who have a prolonged LOS.13 These studies, however, were performed over 10 years ago and did not take into account the perspectives of multiple healthcare team members.
The objective of our study was to identify discharge barriers for all patients admitted to the medicine ward who stayed more than 24 hours after they were deemed medically safe to leave the hospital by the attending physicians. We also conducted interviews with all members of the multidisciplinary care team to incorporate different staff members' perspectives on these barriers.
This was a prospective study conducted over a 4-week period.
Stanford Hospital and Clinics is a tertiary care centre with 444 beds. The medicine ward (which excludes cardiology, pulmonary hypertension, haematology, oncology and post-transplant patients) is comprised of five separate medical teams. Each team is staffed by one attending physician, who attend for 1–2 week blocks, one resident, two interns and one to two medical students. Each team has a dedicated case manager (CM). There are three floors within the medicine service and one pharmacist covers each floor. There are multiple physical and occupational therapists and two discharge pharmacists who help manage complex prescriptions. There is one dedicated discharge planner. Staff roles are described further in table 2.
The medical team begins rounds every day at 9:00. Rounds are usually conducted within the team room and discussions begin with patients who are being discharged that day. Rounds pause at 10:00 while the trainees attend morning report and the attending physicians attend and run multidisciplinary rounds, where discharge plans for patients are discussed with the CMs, nurses, discharge pharmacists and social workers (SWs). Following multidisciplinary rounds, the attending physicians often visit patients to be discharged that morning to confirm that they are ready. Rounds then restart at 11:00 when the trainees return from morning report and usually finish before noon. The attending physicians and trainees spend the rest of the afternoon seeing all the patients on the wards.
Stanford has a comprehensive discharge planning process. Patients have a whiteboard in their rooms with a space for estimated date of discharge (EDD) which providers will update as the patient's condition changes. Family meetings and disposition planning occur throughout a patient's stay in the hospital. There is an effort at Stanford Hospital to discharge patients by noon based on evidence that early discharges help reduce patient capacity bottlenecks.14 Each member of the care team has a specific role in the discharge process that is described in table 2.
Our study sought to evaluate barriers to discharge for adult patients admitted to the medicine ward at Stanford Hospital.
The first was to quantify and characterise delays from the attending physician's perspective over a 4-week period.
The second was to conduct interviews with all team members involved in the discharge process to determine what barriers to discharge they experience and to seek recommendations for reducing these barriers.
Survey development and data collection
We conducted a literature review to determine barriers to discharge faced by hospitals. Using our findings, we developed a survey consisting of possible discharge barriers. The barriers were divided into six categories described in online supplementary figure S1. The survey was then piloted with two hospitalists with expertise in quality improvement and discharge planning and their feedback was incorporated into the final survey. In the survey, we defined a ‘delayed discharge’ as one where a patient was ready to leave the hospital from a medical standpoint, as assessed by the attending physician, but had stayed more than 24 hours for at least one reason outlined on the survey. We excluded patients whose delay was less than 24 hours or who were delayed for reasons regarding their actual medical condition. We included patients who were readmitted within 7 days.
We met with the attending physicians on each of the five medicine teams biweekly from the period of 28 October to 22 November to identify patients currently on their service who experienced a discharge delay. For each patient who met the barrier criteria, we recorded the number of days they were delayed and the barrier(s) that caused the delay. Patients who experienced delays but had not yet been discharged by the end of the study period were included, and their delay was recorded as the number of days from the time of medical readiness to the last day of the study period. For patients who experienced more than one type of barrier, we asked the attending physicians to quantify the number of days the patient was delayed due to each barrier. In a secure REDCap Database, we recorded the patient's medical record number, the date of assessment, any barriers experienced and the number of days of delay. We de-identified the data when aggregating for analysis.
Staff interviews: interview guide development, data collection and data analysis
Following the gathering and analysis of survey data, staff member interviews were conducted with two objectives. The first was to identify additional role-specific barriers that may not be captured by survey data. The second was to seek recommendations from staff on ways to eliminate the barriers identified in the survey data results and interviews. The research team, in conjunction with two hospitalists with expertise in quality improvement, developed a list of staff members to interview (table 2) and pertinent questions for each staff member. The team decided to interview between three and five members of each role based on the availability of staff members during the study time period. There were nine questions that were asked of all staff members and between two to four additional role-specific questions. One to two members of the research team conducted the structured interviews. All questions listed on the staff-specific interview guide were asked, and follow-up questions were asked when the research team member thought elaboration was necessary. At the end of each interview, staff were asked to propose their own recommendations for eliminating both barriers they had specifically discussed as well as barriers revealed through the survey data. A sample of an interview guide is shown in the online supplementary index 2.
All free-response questions were analysed using grounded theory and thematic analysis. A group of five research team members conducted the data analysis. Units of analysis were quotes and anecdotes from the interviews, which were then tagged with the staff member type. The constant comparative method was used to generate themes based on groupings of similar quotes and anecdotes across all staff members. All disagreements in coded themes were resolved through group discussion until a consensus was reached.
Our study protocol was reviewed by the institutional review board at Stanford and was determined not to meet the federal definition of human subjects research.
Over the study period, there were 259 patients discharged from the medicine ward. Eighty-seven of these patients (33.6%) had at least one issue causing a delay of over 24 hours. These 87 patients are characterised in table 1. The mean delay was 3.09 days (median 2 days) with a range of 1–42 days. Each patient with a delayed discharge experienced an average of two barriers, and the total number of separate barriers experienced across all 87 patients totalled 181. The total number of quantifiable delayed days across all 87 patients was 281 days.
The mean LOS for patients admitted to the general medicine service at Stanford (4.45 days, FY2012) is comparable with the national average for all hospitals (4.5 days, 2012). The readmission rate for patients admitted to the general medicine ward at Stanford was 15.78% in FY2012, slightly lower than the national average of 17.5% that year.15
The top barriers and the rate at which each barrier was experienced from the attending physician's perspective through the survey are depicted in figure 1.
A total of 20 interviews were conducted and the breakdown of interviews by type of staff member is shown in table 2. There were six major themes that emerged. These are described in table 3, along with the frequency with which the barrier was described in interviews, the staff members who described the barrier and specific quotes and anecdotes.
Discussion and recommendations
This paper is the first to our knowledge to characterise the exact reasons of discharge delay from the perspective of the attending physicians and to conduct focused interviews with all members of the healthcare team. We found there were additional barriers described in the interviews that were not found through the survey results. One explanation for this may be that the role of the attending physicians in the discharge process is different from those of other healthcare team members. The variation in perspective is a unique aspect of our study that should be further explored. In our discussion, we elaborate on the top barriers that emerged from both methodologies and identify areas for improving the discharge process.
Common reasons for lack of patient readiness included patients who did not feel comfortable leaving, families who felt ill-prepared to care for them and a lack of a support system at home. Patient readiness has been reported as a barrier in previous literature; one large patient survey showed that nearly one-third of patients did not feel ready to leave the hospital at time of discharge.16 ,17
CMs, nurses and SWs felt that attending physicians needed upfront conversations with patients about expected discharge in order to impact readiness. They discussed that routine use of an EDD could keep everyone on the team updated. Studies have shown that the use of a whiteboard to convey updated information, names of healthcare staff and EDD increases patient satisfaction and effective communication between patients and providers.11 ,18
Wait times for procedures, tests, results and consult recommendations
Previous studies have shown that up to 25% of all delays were due to lack of services on the weekends, including procedures, staff members, tests and results.8 ,13 ,19–21 The most common procedures patients waited for in our study were gastrointestinal endoscopies, imaging and biopsies. Nurses and attending physicians reported that patients were often scheduled for procedures while in the hospital because it was presumed they would happen much faster than if referred to the outpatient setting. Many patients also faced delays in waiting to see a consult service, SW or physical therapist. One CM noted that the multispecialty nature of the centre contributed to frequent delays, as there are a number of different teams that must ‘sign-off’ before he or she can be considered ‘medically ready’ for discharge. Wait times for these services have been linked previously to a prolonged LOS and to negative impacts patient outcomes.19
Increased staffing and capacity for procedures over the weekend9 and early identification and scheduling of non-urgent procedures that can be safely done in an outpatient setting could reduce wait times in the hospital.22 We found discrepancies between the medicine team's assessment of patient readiness for discharge and the consult service assessment. As discussed by CMs, proactive coordination and improved communication could reduce this barrier.
Postdischarge facility placement
Availability of an accepting facility, complexity in care management and patient and family preferences were all major contributors to delayed discharges. Preferred facilities did not always have openings. It often took multiple days to confirm insurance coverage at an appropriate skilled nursing facility (SNF), especially for patients who lived further away from the hospital. Most SNFs needed orders at least 2 hours before the patient left the hospital, which was challenging for patients identified for discharge in the afternoon. Some SNFs did not have the capability to manage complex patient conditions; for example, one facility had difficulties accommodating a patient with class III obesity. Finally, patients and their families often took time to agree on a facility, a finding previously reported in the literature.23
CMs noted that when orders were placed before noon, it was easier to coordinate with the facility, family, transport and pharmacy to plan for SNF placement by the deadline of 17:00. CMs and patients both suggested that families should be provided with potential SNFs early in the admission to allow time for discussion and agreement. Some CMs had extensive lists of SNFs in various geographic areas, including the available services, which helped with timely placement. Finally, CMs suggested that coordinating more formal contracts between hospitals and nearby SNFs may be a long-term solution.
Communication challenges within healthcare teams and between teams and patients
Delays in rounding, finalising of prescription plans and the placement of orders all contributed to a lack of effective communication between members of a patient's multidisciplinary care team. Nurses and CMs were most likely to report that communication difficulties impeded timely discharge, a finding that has been reported previously.24 Staff members felt they were unable to initiate conversations around discharge before the attending physicians discussed the topic. Communicating with larger families was noted as particularly challenging and time consuming.
One CM said that the implementation of secure texting on her team helped ease the communication barrier, a finding also described in the literature.25 Staff members recommended that doctors update the discharge date written on the whiteboard, elaborate on the need for follow-up care and describe treatment care goals (eg, for cases when hospice care is the only reasonable option) to family members on a daily basis.
Timeliness of notification of patient discharge
Most interviewees felt that timing of discharge was often delayed because they were not notified early enough. It was common for patients to stay an extra day because discharge paperwork, preauthorisation of medications, sign-off by therapists and consult services, transport or acceptance of facility were not coordinated. The education conference at 10:00 attended by residents and interns interfered with timely completion of rounds in the morning and as a consequence, CMs did not find out about confirmed discharges until the early afternoon.
Attending physicians can enter medications that require preauthorisation the day before the EDD to help pharmacists complete the process. Patients ready to be discharged can be seen early by the team during morning rounds to give CMs the go-ahead to start the discharge process. When possible, residents should inform CMs of discharge the evening before or early in the morning to increase the likelihood that the patient gets discharged that same day.
In the interview data, discharge was often described as a chaotic process because of specific items that must be completed prior to discharge, but were often left to the last minute. These included prior authorisation for some medications, confirmation of a follow-up appointment and completion of discharge summary, required for facility discharges. A number of interviewees discussed that their own role in the discharge process was often unclear to residents and other staff members, and that understanding the role each member played could improve the process.
CMs recommended annual resident training to discuss the role of each staff member in the discharge process, review essential items that must be completed before discharge and discuss ways to improve timeliness. In the literature, institutions have used checklists, whiteboards and forms within electronic medical records to help formalise and standardise the discharge process.26 All of these methods have been shown to improve the discharge process from the perspective of healthcare staff.11 ,17 ,18 One study showed that having a standardised discharge planning process incorporated into nursing activity resulted in fewer patient-reported problems in complying with discharge instructions and fewer patient-reported unmet needs.27
This was a single-centre study conducted in a tertiary care teaching hospital over a short time frame. Our results may not be generalisable to other settings. The attending physician's survey was not formally tested for reliability given the study's short time frame. In addition, methods used to identify barriers and unnecessary delays relied on one attending physician's assessment. Furthermore, while we characterised delays over 24 hours, the study does not take into account shorter delays.
Our study was the first to our knowledge to study barriers to discharge through a mixed-methods approach and to gain insights and recommendations from all staff members involved in the discharge process. We found that there are a number of common barriers to timely discharge of patients that amounted to a significant number of unnecessary hospital days. Staff members had relevant recommendations to reduce the frequency of these barriers, which were also supported by the literature.
More than 30% of all patients were delayed for more than 24 hours for a reason other than their clinical condition.
Patient and family readiness to leave, wait times for tests, results or consults, facility placement, communication challenges and lack of discharge standardisation were the major causes of delay.
There are solutions derived from the literature and staff interviews that could help reduce these barriers.
Current research questions
Does discussion of estimated day of discharge and discharge options with patient and family at the time of admission reduce discharge delays due to patient and family readiness?
Do scheduled, formal training programmes for residents on the institution's discharge process and an afternoon check-in the day before discharge reduce the proportion of unnecessary delays?
How often are the tests, results and consults that delay patients' discharge more than 24 hours essential to their care as an inpatient?
The authors would like to acknowledge Taulant Bacaj, PhD, Ida Bezabeh, BS, William Chen, BS, Simal Ozen Irmak, PhD and Christina Loh, PhD for their assistance in data collection and interpretation. No material in this manuscript has been reprinted or adapted from previously published literature.
Contributors Study concept and design: MVR, DS, LS. Data collection: MVR. Data analysis and interpretation: MVR. Drafting of the manuscript: MVR, DS and LS. Study supervision: DS and LS.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.