Teaching a DISCHARGE Framework to Hospital Teams to Improve Transitions of Care

Background: Residents typically learn about managing transitions of care as part of the informal curriculum in an ad-hoc, reactive manner. Learning may be enhanced by using a framework to proactively practice addressing key domains for a patient soon-to-be-discharged from the hospital. We developed such an evidence-based framework, DISCHARGE , as a cognitive aid. Using this framework, we implemented and evaluated a workshop designed for hospital teams to learn addressing key components of discharging a patient. Methods: All members of 8 Internal Medicine teams across 4 rotations were invited to attend an hour-long workshop ranging from September 2017 - February 2018. Participants completed a retrospective, pre-post survey on their perceived change in discharge-related behaviors. We evaluated the perceived effectiveness of the workshop with a retrospective pre-post questionnaire. We used Wilcoxon signed-rank tests for pairwise comparisons to access perceived changes in discharge behaviors. Results: A total of 90 of 140 team members (64%) attended the workshop and 79 of the 90 (87%) completed the questionnaire. The session was effective in increasing the likelihood of addressing patient-centered behaviors at discharge (mean 1.4 improvement on a 5-point scale, P<0.001, R>0.5). In addition, senior residents and attendings projected they were more likely to discuss the importance of discharge planning with the team early in a rotation. Interns noted they were more comfortable asking the team for help in carrying out a discharge plan (p<0.001, R>0.5). Conclusions: Teaching teams a cognitive aid to practice managing hospital discharges may increase the likelihood of addressing important domains for their patients. Incorporating the team allows for explicit alignment for priorities and communication. Further study is needed to document how such learning is translated into discharge practice.


Introduction
The transition from hospital to home renders patients vulnerable. Hospitalized patients are often discharged with poor understanding of the discharge plan which can lead to preventable morbidity and readmissions (1)(2)(3). National medical societies and health services interventions have identified several domains to be addressed prior to discharge (4)(5)(6). Additionally, the Accreditation Council for Graduate Medical Education highlights transitions of care as a key milestone in resident education (7,8). However, this emphasis has not fully translated into formal medical education, particularly for hospital discharges (9)(10)(11). Much is left to informal curricula that vary based on particular resident or attending style, team workload, and care organization (12). Uncertainty and assumptions about other team members responsibilities at discharge add to the problem (13).
Published transitions of care curricula often focus on recognizing and preventing failed transitions (10,11). Frameworks can serve as a cognitive aid for learning new skills and even improve patient outcomes, (14)(15)(16) yet discharge educational interventions often lack a framework that can enable consistent, proactive practice prior to discharge. Moreover, prior curricula focus on individual healthcare team members without emphasis on teamwork (10,11).
To equip teams with a cognitive approach to discharging patients, we created an evidence-based framework based on prior literature,(13,17-25) summarized as the acronym, "DISCHARGE:" Drugs, Identifying barriers, Self-management of diseases, Communication with primary care physicians (PCP) and caregivers, Home services, Appointments, Red flags signs, Go (Activity), and Educate (Teach) back. Here, we describe our use of this educational intervention to improve management of hospital discharges and evaluate its impact. 4

Setting and Participants
From September 2017 -February 2018, an hour-long discharge planning workshop was given at a safety-net hospital 4 times to rotating internal medicine residents, medical students, and their attending physicians during an inpatient rotation. Of the 140 total members invited during the four sessions, 90 (64%) attended the workshop.

Intervention
The workshop was designed using the following instructional strategies (Figure 1): casebased discussion, large group discussion, and team-based simulation. First, actual postdischarge home visit cases were used to guide the learners in the use of the DISCHARGE framework. Teams were prompted to identify gaps in transition with these post-discharge cases. As gaps were uncovered, those gaps were then organized and revealed the acronym, DISCHARGE.
Interventions for each domain were described to demonstrate potential patient impact.
Participants then brainstormed how to address the DISCHARGE domains in the context of their current caseload and workflow.
Finally, teams were challenged to role-play and analyze one of their own team's soon-tobe-discharged patients using the DISCHARGE framework. The team acted out dividing up responsibilities and role-played patient education.

Outcomes
We evaluated the perceived effectiveness of the workshop with a retrospective pre-post questionnaire that participants voluntarily completed immediately after the workshop (Appendix 1). On a 5-point scale (1 = rarely to 5 = almost always), participants answered how likely they were to perform the following behaviors before and after the session: review medications, educate on disease self-management, review discharge plan with caregivers, communicate with the PCP, and use teach-back with patients. In order to keep the questionnaire brief, we asked about these five domains because we believed they were the most actionable communication changes. We measured perceived change in team communication and the value placed by the team on discharges by assessing the likelihood of interns asking team members for help discharging patients and the likelihood of residents or attendings speaking about the importance of discharge early in the rotation. An open-ended question prompted to "Share your reflection on the workshop experience" for more insight into the instruction. Our Institutional Review Board approved the protocol as exempt from review.

Analysis
We used Wilcoxon signed-rank tests for pairwise comparisons of the retrospective pre and post workshop perceived behaviors. In a secondary analysis, the results were stratified by role (medical students, interns, residents, and attendings). Effect sizes for the Wilcoxon signed rank test were calculated using the R statistic (27). Content analysis was used to assess free-response reflections (28). improvement, 95% CI 0.9,1.9). The retrospective pre to post intervention improvement in all five measures remained significant among subgroups of students, interns, residents, and attendings, with students and interns demonstrating larger improvements than the attendings (P < 0.001). The effect size estimate (R) ranges from 0.57 to 0.60 across these 7 comparisons, suggesting at least a modest effect.

Results
Concerning team communication, senior residents and attendings projected that they would be more likely to discuss the importance of discharge planning with the team early in a rotation in the future. Interns noted they felt more comfortable asking the team for help in carrying out a discharge plan after the session (p < 0.001; R > 0.5).
Open-ended responses emphasized the usefulness of this cognitive aid to patient care and appreciation of practicing as a team. For example, responses highlighted the utility of this framework in uncovering blind spots: "We role-played with a seemingly straightforward patient of ours, but after applying the DISCHARGE mnemonic, there were actually a lot more things to consider where her care could potentially fall through the cracks." Additionally, respondents appreciated the opportunity to align priorities: "Best thing was giving us the time to talk about what each of us should be vs. are doing on discharge. It has always been distressing to put in DC instructions in the computer and hoping it will work out."

Discussion
Applying our comprehensive DISCHARGE approach to patient cases significantly increased team members' self-reported likelihood of conducting safe discharge practices on inpatient teams in a safety-net hospital medical service. To our knowledge, this is the first educational intervention utilizing a shared mental model of key discharge domains, 7 allowing teams the guided opportunity to strategize discharge care for their own patients by practicing teamwork and communication (10,11). Further efforts to standardize learning about discharge care can decrease variation among team members and help improve discharge care and patient outcomes after hospitalizations.
Limitations include application at a single institution, a lack of nurses in the workshop, and no assessment of actual discharge behaviors or patient outcomes. Further research should include longer-term patient follow-up, scaling up to other hospitals and specialties, and including the entire interdisciplinary team in this deliberate practice.

Conclusions
The gap between the competency that trainees require for effectively managing the hospital discharge process and the skills that are explicitly taught can be bridged using an evidence based conceptual framework, DISCHARGE and a brief, active, skill-building workshop. Proactively teaching cognitive aids with deliberate practice can initiate learners in developing tools to practice high-quality transitional care, a needed step towards ensuring the delivery of consistent and comprehensive discharge care to all patients. Abbreviations "DISCHARGE:" Drugs, Identifying barriers, Self-management of diseases, Communication with primary care physicians (PCP) and caregivers, Home services, Appointments, Red flags signs, Go (Activity), and Educate (Teach) back Declarations Ethics approval and consent to participate: Our Institutional Review Board approved the protocol as exempt from review. The consent for participation was considered to have been given as each participant agreed to fill in and submit an anonymous questionnaire. Survey.

Consent for publication: Not applicable
Availability of data and materials: Not Applicable 8 Competing interests: The authors declare no competing interests.
Funding: There was no funding for this work.
Authors' contributions: SPT conceptualized, executed the workshop, analyzed the data and wrote the manuscript. ML helped in creating the DISHCARGE framework and provided substantial feedback on the manuscript. MDS helped analyze the data and provided sustainable feedback on the manuscript. All authors read and approved the final. Manuscript.
12 Figure 1 Instructional Design of Discharge Workshop