Patient Death Debriefing Sessions to Support Residents' Emotional Reactions to Patient Deaths

Abstract

Background

There is no standard way to help residents deal with the emotional impact of patient deaths. Most available curricula are time and resource intensive.

Objective

We introduced “Patient Death Debriefing Sessions” into an inpatient medical oncology rotation at Memorial Sloan Kettering Cancer Center to provide a structured yet practical way to address residents' emotional reactions following the death of a patient. A questionnaire was used to evaluate the impact of these sessions.

Methods

Patient Death Debriefing Sessions consist of a brief (~10 minutes), real-time (within 24–48 hours), consistent (following each death), attending physician–led debriefing that focuses on internal medicine residents' emotional reactions following patient deaths. Sessions were guided by a pocketcard tool and did not require faculty training. Residents completing a 4-week medical oncology rotation were surveyed before and after their rotation. Prerotation and postrotation mean differences were evaluated based on the number of sessions they participated in (0 to ≥ 3) using analyses of variance.

Results

Ninety-one of 92 participants spanning all training levels completed questionnaires (99% response rate). Of these, 79 (87%) encountered a patient death and were included in the analyses. Overall, residents found debriefing sessions helpful, educational, and appreciated attending physician leadership. The number of debriefing sessions positively influenced residents' perception of received support.

Conclusions

This high-yield, novel pilot curriculum supported residents' emotional reactions to patient deaths and may foster communication with team members, including supervising attending physicians. This program is easily implemented and could be adapted for use in other clinical settings.

What was known and gap

Residents may not be prepared to deal with the emotional impact of patient deaths, but oftentimes instituting dedicated curricula is not feasible given time constraints.

What is new

A succinct bedside debriefing by faculty, using a pocketcard.

Limitations

Small sample size, self-reporting, and a questionnaire without established validity evidence.

Bottom line

This debriefing supported residents' emotional reactions to patient deaths, is feasible, and may be adaptable to other clinical settings.

Introduction

Proficiency in end-of-life care is essential yet remains challenging for physicians. The Accreditation Council for Graduate Medical Education and the American Board of Internal Medicine recommend that residency programs include core curricula on end-of-life topics.^1,2 Communication workshops,^3–5 didactic sessions,⁶ and palliative care rotations⁷ have been implemented with variable success. However, when surveyed, residents often feel unprepared to provide care for dying patients.⁸

Additionally, caring for dying patients has an emotional impact on physicians,^9,10 and if unaddressed, can lead to burnout and potentially compromise patient care.^9,11–13 Implementation of monthly, chief resident–led “death rounds” during variable rotations improved overall comfort with discussing end-of-life issues.^14–16 The literature^17–19 suggests that real-time supportive discussions and “teachable moments” may be most effective in addressing the emotional impact of patient deaths on physicians. To our knowledge, these methods have not been formally integrated into a residency curriculum.

We introduced “Patient Death Debriefing Sessions” (PDDS), which are real-time, pragmatic, attending physician–led sessions designed to address the emotional impact of patients' deaths on residents during an oncology rotation. We hypothesized that PDDS would be educational and would improve residents' comfort in discussing their emotional reactions to patient deaths.

Methods

Setting and Participants

PDDS were initiated in February 2012 as a pilot project at Memorial Sloan Kettering Cancer Center (MSKCC) in New York City. Each oncology service had 4 residents with an attending physician for 2 to 4 weeks at a time. The residents were comprised of full-time postgraduate year (PGY)–1s and rotating residents from 5 affiliate institutions. Only rotating residents who rotated for 4 weeks were included in the study. Subinterns were also included in the study, as they were given the same responsibilities as PGY-1s.

Intervention

PDDS focused on the emotional reactions of residents following each patient death. Within 24 to 48 hours of a death, PDDS were held for 10 minutes or more by the inpatient service attending physician in a confidential setting for the patient care team to take advantage of emotional preparedness. While no formal faculty training or standardization was required, a pocketcard (see figure) was created through consensus by the authors and expert faculty, and was distributed as a guide for attending physicians and residents. If no patient deaths occurred, PDDS were encouraged every 1 to 2 weeks on a patient at the end-of-life to maintain regularity. All participating attending physicians and residents were educated about the expectations prior to the rollout of PDDS. Chief residents sent weekly reminders and were available to answer any questions. The priority was to create a high-yield, easily integrated program that required minimal faculty and resident preparation and oversight.

FIGURE.

Pocketcard

Questionnaire

Residents rotating from March to June 2012 completed questionnaires at the beginning (pretest) and end (posttest) of their 4-week rotation. The questionnaire was adapted from a previously used evaluation tool with permission.¹⁰ Participants provided demographic information and training experience. Using a 5-point Likert scale ranging from 1 (not at all) to 5 (extremely), we asked 3 “emotional reaction” questions related to residents' perception of how they deal with patients' deaths, and 2 “attending support” questions to assess their comfort level in discussing patient deaths with attending physicians. We also asked 6 “debriefing session” questions (3 Likert scale, 3 open ended) on the educational value of PDDS. The open-ended questions were included as another opportunity for feedback.

This study was granted exemption from review by MSKCC's Institutional Review Board.

Data Analysis

Participants were divided into 4 groups based on the number that received “0,” “1,” “2,” or “3 or more” self-reported debriefing sessions. For the “emotional reaction” and “attending support” questions, posttest and pretest (post-pre) means were calculated and differences between groups were evaluated. For the Likert scale “debriefing session” questions, differences in posttest means were compared. For the open-ended “debriefing session” questions, the authors (J.E. and E.S.) separately reviewed and consolidated responses. Analyses of variance were used to identify differences between groups (SPSS version 18, IBM Corp). Statistical significance was set at P < .05.

Results

A total of 91 of 92 participants completed both pretests and posttests (99% response rate). Twelve residents did not encounter a patient death and were excluded, leaving 79 who experienced 0 to 5 PDDS (table 1). Participants ranged in age from 25 to 43 years, were slightly more often men, and represented all levels of training. A majority had prior formal coursework in end-of-life care, and had cared for a patient who died (table 2).

TABLE 1.

Participants Stratified by Self-Reported Debriefings and Patient Deaths

TABLE 2.

Participant Characteristics

For the “emotional reactions” and “attending support” questions, the post-pre means significantly differed between the groups for 1 of the “emotional reactions” questions (F value = 3.41, P < .05), and no other significant difference was found (table 3). For the 3 Likert scale “debriefing session” questions, the posttest means significantly varied between the groups (“helpful” F value = 4.01; “learn” F value = 4.42; “important” F value = 3.01; all P < .05), with group “3 or more” having the highest posttest means (“helpful” mean = 4.23; “learn” mean = 3.94; “important” mean = 4.29; table 3). table 4 highlights representative comments to the open-ended “debriefing session” questions.

TABLE 3.

Differences in Posttest and Pretest Means and Posttest Means for Emotional Reactions, Attending Physician Support, and Likert Scale Debriefing Session Questions

TABLE 4.

Representative Responses to Open-Ended Debriefing Session Questions

Discussion

This pilot study suggests that PDDS are feasible methods to address residents' emotional reactions to death in a real-time, minimally time-consuming, consistent manner. Studies have shown that caring for dying patients has a strong emotional impact on residents, and if not addressed, may have negative consequences.^9,11–13,20 PDDS improved residents' perception of support and coping, which we believe is a critical first step to effectively dealing with emotional reactions. The results also suggest that residents' responses may improve the most in the group with the highest number of debriefing sessions. Debriefing frequently and consistently on every patient death may foster a more open forum that normalizes sharing one's emotions, which we believe is an important, potentially culture-changing aspect of the program.

Residents often do not discuss reactions to patient deaths with attending physicians or find them helpful in this regard.^10,20 PDDS provide an opportunity and expectation to improve this communication gap. Overall, the residents found the sessions helpful and educational, and appreciated attending physician leadership.

Previous attempts to address residents' emotional reactions to death have been successfully implemented as monthly large group sessions^14–16 and multidisciplinary formal wrap-ups involving lengthy facilitator preparation.²¹ In contrast, PDDS are novel real-world interventions that can be easily integrated into a complex work environment with competing educational demands, making them sustainable. The pocketcard facilitates focused sessions without the need for extensive training. In fact, over the last 3 years, PDDS have been ongoing and have been integrated into the culture at MSKCC.

Our pilot study has several limitations. First, our data are self-reported, and the questionnaire was not tested in this population or setting and has little supporting validity evidence. Responses may have been subject to response and recall bias. While our study was designed to evaluate the impact of the differing number of PDDS on residents' perceptions of emotional support, our small sample size limits the ability to detect statistical differences. It is possible that the educational value of PDDS may simply have been due to increased face time with the team, or the pocketcard itself. While not a formal control group, the “zero debriefings” group was the only group to have lower posttest means, suggesting that participating in any debriefing sessions may be beneficial. Since the program intentionally was not standardized or observed, the residents' actual experiences with PDDS were not measured. Though the open-ended feedback questions were informative, a formal qualitative study may better assess the impact of such a program. Residents were followed for 1 month, and it is possible that the impact of PDDS may extend beyond this time.

Future larger-scale studies can evaluate the importance of session frequency, gender, or contextual factors, as well as the impact of this type of curriculum, on residents' long-term coping mechanisms, attending physicians' emotional reactions, and other venues.

Conclusion

We describe Patient Death Debriefing Sessions as novel and practical approaches to address the curricular gap in discussing residents' emotional reactions to patient deaths. Our pilot study demonstrated that this program may help residents cope better and feel more supported after patient deaths. With a manageable level of preparation using a pocketcard for guidance, this program effectively balanced time constraints on a busy medical oncology service and could be adapted to other educational settings.