Medical schools are required to provide end-of-life instruction for purposes of accreditation. However, the quality, quantity, methodology, and timing of formal and informal instruction are highly variable. Although recent national surveys report trending improvement in end-of-life instruction, a substantial portion of fourth-year medical students still feel unprepared for death and dying discussions with patients and families (1, 2). In 2006, 80% of medical school graduates felt their training in death and dying was at least adequate, versus 71% in 1999 (1). Greater improvements were reported in knowledge about pain-management (21%) and palliative-care principles (15%) during this time period. However, a more probing national survey revealed that although students are gaining confidence in pain-management and discussions about end-of-life decisions with patients, a substantial portion still struggle with the psychological impact of working with dying patients (2).

A report from the “National Consensus Conference on Medical Education for Care Near the End of Life” suggests that training institutions use death-and-dying teaching opportunities during the preclinical years (3). Recommendations for empathic communication tasks to be accomplished during the preclinical curriculum include 1) understanding psychological aspects of suffering, loss, and bereavement; and 2) developing the basic communication skills of listening to the impact that illness has on life; exploring hopelessness, helplessness, and fear; and discussing bereavement with the patient and family. However, most research on end-of-life education focuses on the experiences of medical students during the mandatory clinical clerkships or during elective rotations (4, 5).

The purpose of this study was to gather pilot data describing the learning potential for a behavioral and palliative-medicine collaboration introducing first-year medical students to the death-and-dying doctoring experience through small-group conversations between students and family members of recently-deceased loved ones. Our focus for the activity was to improve student comfort and knowledge with end-of-life issues, and we hypothesized that the majority of students would self-report benefit from participating in this activity.

This study was a retrospective analysis of first-year medical student small-group activity evaluations for the Death and Dying component of a required course titled Human Behavior. The University of Texas Health Science Center at San Antonio (UTHSCSA) Institutional Review Board (IRB) approved this protocol with “exempt” status to review anonymous activity evaluations without written consent.

The Death-and-Dying Human Behavior Curriculum

The Human Behavior course at UTHSCSA blends lecture and required small-group activities to introduce basic psychological principles to first-year medical students. Small-group activities pair about 15 students with two psychology/psychiatry faculty members for in-depth discussions covering mental defense mechanisms, personality styles, diversity, human sexuality, and death and dying. For the 2008–2009 academic year, the course directors collaborated with palliative medicine to develop an interdisciplinary death-and-dying curriculum. This consisted of 2 hours of lecture delivered by a palliative-medicine specialist and a geriatric psychiatrist and a 2-hour small-group activity. For the small-group activity, Palliative Medicine team members recruited family members of recently-deceased loved ones to discuss their medical/hospice experiences with the first-year medical students. As part of their preparation for group participation, all volunteering family members were informed by the palliative-medicine team that the purpose of the activity was to improve students’ knowledge and comfort with end-of-life issues. They were asked to share their experiences with the students and were given examples of questions likely to be asked. Family members volunteered to form 14 small groups, each consisting of 16 students, 2 psychology/psychiatry small-group leaders, 1 palliative-team member, and 1 family member. Faculty facilitated a 1-hour 15-minute conversation with family members, where they shared their experiences and answered questions. Afterward, family members were escorted to a separate room by the palliative-medicine team member to discuss their experience and provide feedback in an open, safe, debriefing environment. Students remained in the room and discussed the experience with their small-group leaders.

Small-Group Activity Evaluations

Anonymous small-group activity evaluations were distributed to all students (N=222) immediately after the group-processing component. Students were required to submit an evaluation before leaving the room, which ensured a 100% response rate. Each evaluation used a 5-point Likert-scale (1: strongly agree, 2: agree, 3: neutral, 4: disagree, and 5: strongly disagree) to assess students’ perspectives of faculty and family member comfort level with end-of-life issues, pre- and post-activity student comfort level with end-of-life issues, post-activity improvement in knowledge, appropriateness of the activity, usefulness of the group-processing session, activity organization, and overall importance of the activity. All information, along with written comments, was extracted from each evaluation and entered into a database for analysis.

All students participating in the activity submitted an evaluation (N=222); 170 students (77%) reported increased comfort levels with discussions about end-of-life issues as a result of the activity; 98 students (44%) reported having felt comfortable with discussions related to death and dying at baseline or before the small-group activity. Within this subset, 79 (81%) stated that the small-group activity enhanced their previous comfort level. Of 51 students, 43 (84%) who reported discomfort with end-of-life issues before the activity, reported benefit; 74 students expressed uncertainty about their baseline comfort level with end-of-life issues (scored their pre-activity comfort level as “3”), of which 48 (65%) claimed benefit from the experience. A majority (85%; 189 students) endorsed improved knowledge as a result of participating in the activity. Students reporting gains in comfort and knowledge with end-of-life issues were more likely to perceive higher facilitator and family comfort levels with end-of-life discussions.

Before conclusions can be drawn, several limitations should be acknowledged. First, our primary outcome measures of improved comfort and knowledge are subjective self-reports, based on required student evaluations designed to give the course instructors feedback for course improvement. We did not assess for reliability or validity before implementation, and there is the possibility of a response bias in a socially-desirable direction. Second, we did not compare pre- and post-activity evaluations or use a comparison or control group to analyze our results. It is possible that this activity has no more benefit than role-playing, as has been done in the past, or attending just the preceding lectures that discussed communication skills with suffering loved ones. Third, this activity focused on the dying experience of chronically ill patients and did not address unexpected deaths from trauma, or sudden causes, such as heart attack or stroke. Finally, our only follow-up to-date occurred immediately after the intervention. We do not yet know whether this activity will have a lasting impact.

Despite these limitations, this study suggests that an interdisciplinary death-and-dying preclinical activity utilizing the perspectives of family members of recently-deceased hospice patients may decrease distress and improve end-of-life knowledge at an early point in medical students’ careers; 77% of students reported increased comfort levels with end-of-life discussions, and 85% reported increased knowledge of end-of-life issues. As one student wrote in the evaluation, “Really meaningful experience. [It] caused me to think about some things that I hadn't before. Listening to an actual survivor relate their experiences made a far more significant impact than a discussion in the absence of someone like our family member. Thanks for an unexpectedly poignant opportunity.”

Future studies should focus on assessing the real impact of this activity and determining whether it has a lasting effect through the clinical years. More rigorous analyses, using validated assessment instruments, should compare pre- and post-activity evaluations, compare intervention and control groups, and examine repeated measures to determine the longitudinal utility of this activity. Suggestions offered by students included incorporating role-playing scenarios to practice giving “bad news,” providing students with pre-activity background information about the family members’ experiences, and assigning more family members to each session, for a more diverse experience.

Clearly, activities such as these are not enough. Patients, patient-advocates, and medical students demand and are deserving of more formal death-and-dying instruction in the curriculum. Electives, although often highly effective and novel, are not sufficient; nor may it be wise to delay “face-to-face” training until the clinical years, given the angst often expressed by students with end-of-life discussions. However, as suggested by Barnard et al., the best approach may not be to add an entirely new curriculum to overburdened faculty and staff (3). Rather, it is more prudent to identify opportunities within each institution’s existing framework. This collaborative endeavor between behavioral and palliative medicine identified a potential opportunity within the context of Human Behavior. Time will determine the long-term impact of this “poignant” experience for our first-year medical students.

The authors thank Rosalie Salas and Mary Garzafor coordinating this important learning opportunity.

This study was funded by a grant from a UT Innovations in Health Professions Education Grant and a Geriatric Academic Career Award from the Department of Health and Human Services (SSR).