Patient suicide lurks as a frightening possibility in the professional life of every psychiatrist. It has been estimated that patient suicide has been experienced by as many as half of psychiatrists in practice (1) and by one-third of psychiatry residents at some point during their training (2). Not surprisingly, surveys have established that mental health practitioners in general view suicide as their leading source of work-related stress (3), leading some to view it as an "occupational hazard (4)." As common as suicide might be, patient suicide is far from a trivial event in the life of the treating clinician: the loss of a patient to suicide appears to affect practitioners on a level comparable to that of a death in their family (4).
However stressful a patient's suicide might be to an experienced practitioner, it stands to reason that the experience would have an even greater impact on clinicians in training. The combination of clinical inexperience and lack of perspective for dealing with such events places trainees at greater emotional risk should a patient die by suicide on their "watch." Indeed, research supports this prediction, suggesting even greater stress among trainees, relative to experienced clinicians, when a patient commits suicide (5). Sacks et al. (6) described a variety of ways that a patient's suicide adversely affects not only the resident but also the resident's peers and the program's administration staff. Residents may be especially at risk during the first 2 years of training, which typically place residents on inpatient and emergency services, where they are most likely to be confronted with severely disturbed, suicidal patients. It is during these early rotations that they are least prepared, in knowledge, skills, and experience, to deal with such patients or the emotional impact of a death by suicide.
Clinical supervisors and training directors encounter this issue on a number of levels. In addition to training issues, emotional, legal, and administrative ramifications result from a patient suicide in the context of a training program. In the wake of this unpredictable event, the needs of both the resident and the patient's surviving family members must be attended to. Implications for both the individual resident's training and the clinical and training program in general must be addressed. Administrative reporting must be performed and quality review conducted, with legal considerations in clear view. Moreover, in addition to a training director's concern about the emotional impact of a patient's suicide on a trainee, training directors, already concerned about recruitment difficulties, often worry about a resident's doubts about career choice. It is not uncommon to hear of residents (or medical students) who "swore off" psychiatry as a specialty after experiencing a patient's suicide (7).
A need clearly exists for a thoughtful, systematic approach to the issue of patient suicide in the context of a training program. An adverse impact resulting from patient suicide can be minimized through two lines of action. The first, of course, is to reduce the chances that those suicides will occur in the first place by training residents to detect suicide risk and intervene appropriately. Although prior studies have indicated a need for increased emphasis on this front in psychology and some other training arenas (8), our literature search revealed no surveys of these training activities in psychiatry residency programs. Therefore, the first goal of this study was to learn about the presence and nature of training in assessment and treatment of suicidal patients in psychiatry residency programs.
The second line of action is active "postvention" with the resident after a patient suicide has occurred. The concept of suicide postvention was developed by Shneidman (9) to address the needs of the deceased patient's family and other survivors. It can be succinctly defined as "those appropriate and helpful acts that come after the dire event itself (p. 165)." When used as a component of training, however, postvention presents an added set of complications. For not only might the affected resident question his/her clinical competence and the director question the adequacy of the training or supervision, but both also quickly become aware of legal considerations. For example, one author has urged supervisors to exercise caution in supervisory sessions following a patient's suicide, stemming from concern that whatever is said might be used against the trainee, supervisor, and/or training institution in subsequent legal proceedings (10).
Thus, when postvention is undertaken in training settings, several priorities present themselves. These include 1) taking care of the needs of trainees, 2) practicing sound quality assurance, and 3) taking care to engage in "neither a whitewash nor a witch-hunt" (2, p. 429). Although our literature search revealed a number of works on the impact of suicide on trainees and how to assist them (5—7), we found no references to surveys of actual postvention practices in training programs. This was the second goal of our survey.
After exploring various issues concerning the postmortem review of patient suicides with legal and administrative consultants, we constructed a two-page questionnaire that uses a checklist format. The questionnaire was kept brief by design to maximize the response rate; it required about 5 minutes to complete. The following information was requested:
Names and addresses were obtained from 1994—1995 Directory of the American Association of Directors of Psychiatric Residency Training (AADPRT), which represents more than 90% of all psychiatry residency programs in the United States (Meinsler, L., personal communication, 1996). A cover letter explained the purposes of the survey but did not convey any expectations regarding results. Respondents were guaranteed confidentiality and were given the option to omit identifying program information if they wished. Returns were monitored through the use of a numbering system, which permitted a reminder card to be sent to nonrespondents 2 weeks after the initial mailing.
All listed AADPRT members were included on the mailing list, including directors of both general residency and child psychiatry fellowship programs. To maximize program representation, assistant directors also were included. Returns were carefully monitored, and assistant director responses were used only in cases in which directors did not respond. In five instances, the questionnaires were returned by both the director and assistant director of the same program; in each case, the assistant director's questionnaire was excluded to eliminate duplication. For the purposes of this survey, general residency and child fellowship programs from the same institution were regarded as separate programs, since these programs represent separate sets of trainees and almost always are run by different directors. In addition, general and child programs typically have separate sets of policies and procedures and undergo independent accreditation processes.
Packets containing a questionnaire, cover letter, and return envelope were mailed to a total of 296 programs. The packets were mailed in October 1994. One hundred seventy-one questionnaires were returned. Five questionnaires were eliminated as duplicate responses, resulting in a final tally of 166 programs, a response rate of 56.1%. Nearly two-thirds (62.7%) of returned questionnaires were from general psychiatry residencies, with the remainder (37.3%) from child psychiatry fellowship programs.
Responses were received from every AADPRT region, including 16 from Region I (New England), 22 from Region II (New York State), 14 from Region III (Mid-Atlantic), 39 from Region IV (Midwest), 44 from Region V (Southeast), 15 from Region VI (California), and 16 from Region VII (Western States). Return rates varied considerably by region, with the differences reaching statistical significance (χ2=14.79, df=6, P=0.022, N=166). Best represented was Region VII, with an 80% return rate; least represented was Region III, with a 36% response rate.
Two-thirds (67.3%) of the responding programs were located in university medical centers, with another fifth (21.6%) in hospitals or other medical centers. The remainder were classified as Veterans Affairs/military settings, state hospitals, and other psychiatric centers. The average general psychiatry program reported 28.2 residency positions, the average child fellowship 4.6 positions. Average program age was 27.0 years. More than 90% of the programs reported full accreditation status.
Preliminary statistical analyses revealed no significant differences in the survey responses between the general psychiatry and child fellowship programs, so the data sets were combined in all further analyses. As shown in T1, the vast majority of the responding programs (94%) reported some form of didactic training in the treatment of suicidal patients, with relatively little variation among types of training sites. This figure contrasts sharply with the 56% figure obtained in Bongar and Harmatz's study of graduate programs in clinical psychology (8). However, their survey considered only formal graduate course work, in contrast to our study, which examined only programs in which training was delivered in the context of supervised clinical work.
Although responses to the general question regarding suicide-related training were quite uniform, considerable variability emerged when we asked about the specific forms of that training. As shown in F1, course offerings drop off considerably as training intensity, formality, and narrowness of focus on suicide increase. This is particularly notable in the case of skill-development workshops devoted to suicide, which are offered by only about one-fourth of the responding programs. By far, the greatest numbers of programs offer suicide-related training in contexts less specifically devoted to the topic of suicide, such as therapy supervision, seminars on general topics, and case conferences.
In addition to inquiring about didactic training activities, we also asked directors to indicate what clinical rotations provided residents with supervised experience in working with suicidal individuals. As expected, the most prominent among these by far was emergency room rotations, reported by 86.5% of the responding programs. (Standard inpatient and outpatient rotations were not listed on the questionnaire because these rotations are universal and supervised experience with suicidal patients and can be assumed to be present.) About one-third of programs (35.7%) offer rotations in other kinds of crisis-intervention settings, and slightly fewer (29.8%) through outpatient depression clinics. A few programs (10.5%) included experience working on telephone crisis-counseling hotlines.
As also shown in T1, in relatively few programs does training specifically prepare residents for the possibility of a patient's suicide. Less than half overall (46.9%) reported that training activities instruct residents on what to do if a patient should commit suicide. Training in suicide postvention procedures was significantly more likely to occur in some training settings than in others (χ2=14.36, df=5, P=0.013, N=171), with the programs in the hospital/medical center category most likely to offer such preparation. Far fewer programs reported that postvention procedures were specified in their policies and procedures manual. The mean frequency was 18.6%, with differences among the settings not reaching statistical significance (χ2=2.85, df=5, NS, N=171).
F2 shows what actions are required or recommended by training programs following the suicide of a trainee's patient. Most programs (71.3%) require the supervisor to be notified, about half (53.8%) require that the clinical director be notified, and slightly more than one-third (39.8%) require notification of the training director. Roughly half of the programs (54.4%) require that a formal incident report be filed. Smaller but significant numbers of programs recommend rather than require these actions. (It should be noted here that absence of an affirmative response on these items signifies only that certain procedures are not specified orally or in writing and should not be construed to suggest that programs recommend against any specific course of action.)
Given the potential conflict between liability management and supportive interventions with affected trainees (i.e., concern that debriefing statements might be scrutinized in some future legal proceeding and viewed as incriminating), we were especially interested to see what programs would report in terms of postvention. These results indicate that programs are more inclined to recommend than to require various supportive interventions. About one-third of programs require one or more supervision sessions to assess and process emotional impact, whereas slightly more (41.5%) recommend such sessions. Counseling or therapy is recommended by somewhat less than half of the programs (42.1%) and required by a few (4.7%).
The final issue addressed by the questionnaire was that of the review agenda following a patient's suicide, that is, determining what can be learned from the incident and whether any corrective measures need to be implemented. A postmortem meeting to "critique" the clinical intervention is required or recommended by almost three-quarters (71.3%) of the programs. About half of responding programs (51.4%) require a formal review by a quality assurance (QA) or morbidity and mortality (M&M) panel. It is interesting to note that somewhat more programs require a critique or QA/M&M review than require supportive supervision or counseling.
Results of this survey can be considered adequately representative of suicide training and postvention practices in U.S. psychiatry residency and child psychiatry fellowship programs. However, it must be noted that a 56% response rate, while typical for surveys of this nature, represents only about half of the existing programs. It also must be noted that these results may not adequately represent programs from the East, due largely to the underrepresentation of Mid-Atlantic programs and relative overrepresentation of programs in the West.
In commenting on training issues pertaining to patient suicide, Bongar and Harmatz (8) stated that, "It is…essential to understand that the management of suicidal patients may be more of a high-risk professional endeavor than most other clinical situations, and may well require extensive clinical and didactic training (p. 238)." However, despite the apparent frequency of patient suicide, there is concern that "we may not be training clinicians adequately to manage suicidal patients (11, p. 254)." Results of the present study are mixed in this regard. While the great majority of residency programs do provide some form of training and individual supervision in assessment and treatment of suicidal patients, such training often may be relatively superficial in nature. Whereas more than 90% of the programs report that the topic is covered in clinical supervision or seminars and journal clubs, only about one-fourth of the programs report training in the form of skill-building workshops devoted specifically to suicide assessment and intervention. It is possible that traditional vehicles for teaching information and skills necessary for competent diagnosis and treatment of psychiatric patients are incommensurate to the challenge (and risks) of treating suicidal patients. For, unlike most situations in clinical psychiatry, errors in dealing with suicidal patients are potentially irreversible.
In his call for the study of suicide to be recognized as a core curriculum item in psychiatry residency training, Lomax (12) noted that "the knowledge base related to suicide should be covered in a formal didactic setting, such as seminars or lectures. Obviously, knowledge will grow out of the supervisory process as well, but residency programs cannot afford to delegate such critical material solely to the poorly regulated medium of supervision (p. 58)." The present data provide evidence that seminars and lectures are, indeed, being offered. However, we would suggest that lecture information is only a first step in learning to work with suicidal patients and that the ideal mechanism for developing the skills to apply such knowledge is the skill-building workshop. For it is here that trainees have the opportunity to practice skills and receive immediate corrective feedback. Results of this survey indicate that workshops in suicidology are currently an atypical training modality relative to the more traditional supervision and seminar-lecture formats. This presents an opportunity for training directors to enhance training and possibly reduce the likelihood of suicide by residents' patients.
Of course, regardless of the intensity of preventive efforts, patient suicides in training programs can be expected and should be planned for. Patient suicide has been described as "a common, if not universal, part of the psychiatric residency (13)." However, just as (and perhaps partially because) practicing professionals often have no mechanism in place for helping themselves and colleagues to cope with a patient's suicide (7), this survey revealed great variability in such procedures in residency training programs. Chances are roughly 50/50 that any given training program will provide residents with advance guidance for a possible patient suicide. It is even less likely that a postvention protocol will be outlined in the program's policies and procedures manual. An ad hoc approach appears to be more the rule. Several respondents commented on their survey forms that, because they rarely experienced patient suicides, they approached them as they would any other low-probability crisis.
Training directors might wish to consider three sets of needs when reviewing suicide postvention procedures. In the area of administration, program accountability requires systematic collection of data on suicides occurring in the context of a training program. It is crucial that statistical information on suicidal behavior by patients of residents be available to inform decision making about such matters as policies and procedures and staffing patterns. In this survey, administrative procedures such as notification of the supervisor or filing an incident report were reported in only one-half to three-fourths of the responding programs.
Educational and training needs, on both individual and programmatic levels, constitute a second area for focus and consideration. Chief among these are procedures for reviewing the case following the suicide of a resident's patient. Such review is needed to allow the trainee to review his/her intervention and to allow the program to determine whether procedural changes (such as training or supervision requirements) need to be implemented. Despite concerns about the legal implications of discussing any errors that might have been made (10), most of the programs participating in this survey require or recommend a critique meeting and/or review by a QA or M&M panel. This fact suggests a widely recognized utility and need from both training and administrative standpoints. It is possible that litigation fears (i.e., that information from these discussions might be used against the resident or training institution) are balanced by corresponding concerns about program liability if such reviews are not conducted. It is also worth noting that, in many states, such reviews are legally protected as privileged communication. It is less clear whether individual supervision enjoys the same protection.
The final, and often most immediately pressing, concern in the aftermath of a patient suicide is the emotional needs of the affected resident. As noted before, studies have consistently documented the toll that a patient's suicide has on therapists, from trainees to seasoned clinicians. Several authors have described ways to assist a trainee or fellow clinician following a patient's suicide. Sacks et al. (6) conceptualized the postvention period as a particularly critical one for the resident's development and outlined a number of ways that the program could provide needed emotional support. They also recommended the "psychological autopsy," despite its possible emotional difficulty, as an effective way for both the trainee and the institution to gain insight into the suicide and evaluate the therapeutic intervention.
Brown (2) outlined a model program response to patient suicide, designed to follow the stages of the resident's response. He recommended that the case review occur only after the resident has completed the initial "working-through period" and is in a better position to consider objectively the facts of the case. In this regard, it is worth noting that the present data show that administrative and clinical review procedures are more likely to be required or recommended than are supportive measures. It is possible that increased emphasis on the emotional reactions of residents affected by patient suicide would benefit not only the resident but also patients and the training program as well.
This work was supported by a grant from the Charleston Area Medical Center Foundation. Versions of this paper were presented at the 28th Annual Conference of the American Association of Suicidology, May 1995, Phoenix, AZ, and at a meeting of the American Association of Directors of Psychiatric Residency Training, January 1996, San Francisco, CA.
The authors thank the anonymous reviewers for their comments on an earlier version of this paper.