The experience of patient suicide among Irish psychiatrists

One hundred and nine consultant psychiatrists completed an anonymous questionnaire on the suicide of patients in their care. A patient suicide was reported by 82%,of whom 47% had experienced their first patient suicide within five years of entering psychiatric practice. A formal review of the suicide had occurred In 24%of cases. Thestudy examines the perceived effects of their first patient suicide on the professional practice of the respondents, and measures the personal stress related to It, using the Impact of EventScale. Wediscussthe implications of thiscommon experience for psychiatric training programmes. The suicide of a patient can have a traumatic effect on relatives and clinicians alike. In addition to loss, there frequently follows a period of prolonged questioning-a search for meaning. We surveyed 119 psychiatrists to see how they would act with a patient who has taken a potentially lethal overdose, has no mental Illness, and Is refusing treatment. There was substantial disagreement. There may be a risk of action under civil law whether the psychiatrist decides to treat the patient without hisor her consent or not. The Law Commission are examining whether mental disorder, as defined in the Mental Health Act 1983, should be used as a test of Incapacity to give consent to medical treatment. Such legislation would help doctors but may encourage a wide Interpretation of the definition of mental disorder.

The suicide of a patient can have a traumatic effect on relatives and clinicians alike. In addition to loss, there frequently follows a period of prolonged questioning-a search for meaning. Families feel stigmatised by suicide, and may be unwilling to discuss it, a position compounded by social attitudes of blame (Ness & Pfeffer, 1990). Psychiatrists share some of these experiences when their patient commits suicide (Litman, 1965). They may experience anger and sadness, with consequent blame, guilt and denial, associated with obsessive reviewing and Impaired performance. Fears of blame or legal action may lead to avoidance of contact with the family and of the funeral. Selfdoubt regarding clinical performance may result in conservatism or avoidance of suicidal patients.
The reported frequency of patient suicide as experienced by psychiatrists varies from 22-57% (Kahne, 1968;O'Reilly et ai 1990 The aims of this study were: (a) to ascertain the proportion of Irish psychiatrists who have experienced the suicide of one or more of their patients (b) to determine the frequency of formal reviews and funeral attendance following patient suicide (c) to estimate the level of stress following the first patient suicide (immediately and currently), using the Impact of Event Scale (Horowitz et cd, 1979), and (d) to estimate the effect of the first patient suicide on respondents' clinical practice.

The study
An anonymous questionnaire was sent to 178 psychiatrists working in Ireland with a reminder two months later. This included questions on gender, years in practice, type and location of practice and therapeutic orientation, in addition to experience of patient suicide during treatment. In the event of a patient suicide having occurred, the respondents completed questions regarding their years in psychiatric practice at the time of their first suicide, whether the patient was an in-patient, and whether there was a formal review following the event. They indicated how they felt the first patient suicide had affected their practice of psychiatry on a four point scale (from no impact to major impact) in the areas of likelihood to admit patients to hospital, record-keeping, legal aspects of practice, use of peer consultation and focus on suicide. They recorded whether they had attended the funeral and if so whether this had been helpful personally and to the patient's relatives. They then completed the Impact of Event Scale (Horowitz et al 1979) for the two weeks following the first patient suicide, and for the two weeks prior to completing the survey. This scale is a 15-item, self-report, forced choice questionnaire which measures subjective stress related to a specific event. It divides into two sub-scales which measure intrusive thoughts and dreams related to the event and avoidance responses.

Findings
The sample One hundred and nine psychiatrists completed the survey (61.2%). A wide range of subspeciality interests and therapeutic orientations was represented, and the mean duration of psychiatric practice of the respondents was 18.7 years.

The experience of patient suicide
Eighty-two per cent reported that a patient had committed suicide during treatment, and of those reporting one suicide, 81% had experienced more than one. The first suicide occurred within five years of starting psychiatry for 47%, and while the patient was an in-patient for 31%. A formal review followed the event for 24%. Fifteen per cent attended the funeral, of whom 55% found it personally helpful, and 62% considered it helpful to the patient's relatives. There was no difference in gender or years of psychiatric practice between those who had experienced patient suicide and those who had not. Eightyeight per cent of general adult psychiatrists, compared with 50% of child and adolescent psychiatrists and 57% of mental handicap psychiatrists, reported a patient suicide.

Effects of patient suicide on professional practice
The percentage of psychiatrists who reported that their first patient suicide had brought a definite or major impact on their likelihood to admit patients to hospital was 15%, on detailed record keeping, 31%; on increased use of peer consultation, 25%; on increased attention to the legal aspects of practice, 25%; and on increased focus on suicide, 58%.

Comment
The 82% of Irish consultant psychiatrists who have had a patient commit suicide is higher than reported from studies in North America, where the highest figure was 57% (O'Reilly et al, 1990). The difference could be accounted for by a higher known to unknown ratio of suicides in Irish practice, since the population is small and less mobile. Henn (1978) found that the known to unknown ratio among psychiatric residents in Los Angeles for patient suicide was 1:12, but this related to contact within the previous year, and not to current treatment. It might also be accounted for by differences in the structure of psychiatric services between Ireland and North America, with the latter including a larger proportion of psychiatrists working in private psychotherapeutic practice, with less severely ill patients.
Brown (1987a&b) has highlighted the particular difficulties for psychiatric trainees when a patient in their care commits suicide. Although protected to some extent by the 'training umbrella', their personal sense of failure may be greater because of lack of expertise, and the sudden confrontation with their limitations in achieving "deeply held narcissistic aspirations". In this study, 47% of respondents who had experienced patient suicide did so during training, but their initial scores on the Impact of Event Scale were not significantly higher than those for whom patient suicide occurred later in clinical practice. Chemtob et al (1988) did find that younger, less experienced clinicians were more affected by a patient's suicide, and in their discussion, emphasise the need for establishing formal mechanisms to deal with its impact. A formal review followed the patient suicide for only 24% of the respondents in our study, but there was no significant difference in their Impact of Event scores compared to those who did not have such a review. However, this could be related to the particular circumstances surrounding the suicides where a review occured, or to the small numbers (20 and 64 respectively) which may have been insufficient to detect a difference. The personal stress related to their patient's suicide, as measured by the Impact of Event Scale, while significant, was lower than that recorded by American psychiatrists (Chemtob et al, 1988), where 53% of the sample had initial scores comparable to a clinical group. In our study, clinical levels of intrusion were reported by 36% and avoidance by 17.5%. However, only 15% attended their patient's funeral, and this, in a culture where funeral attendance is the accepted norm in everyday life, might indicate a higher level of avoidance than reported. The differences between the American and Irish psychiatrists may be cultural, or may relate to the more widespread practice of psychoanalysis in America, with the formation of long-term intense transference and countertransference relationships, resulting in a greater sense of loss and failure with suicide.
Apart from 'focus on suicide', the majority of respondents did not report that patient suicide had brought a definite or major impact on their professional practice. However, 15% did consider themselves more likely to admit patients to hospital -a change in practice with serious consequences in the context of reducing in-patient bed numbers.
Brown (1987a&b) has suggested that training schemes introduce a "comprehensive programmatic response" to patient suicide. This should include an anticipatory phase, in which trainees, in addition to learning about the epidemiology and dynamics of suicide, are encouraged to imagine their own reaction in the event of a patient committing suicide. He also suggests the introduction of policies of Joint care with a senior psychiatrist for extremely suicidal patients, while ensuring clear clinical responsibility. When suicide occurs, the trainee should be initially offered support and an opportunity to ventilate his or her feelings with a senior colleague, and also supported in meeting the patient's relatives. After a period of some weeks, the "psychological autopsy" should allow an opportunity to examine the possible reasons for the suicide, to evaluate the quality of the care given, and the effects on the patient's family and therapist. Brown also suggests two follow-up phases, the first to assess the individual trainee's progress with resolution, and the second to prepare trainees for independent clinical practice. We could usefully implement these recommendations, given that patient suicide is an almost universal event for Irish (and most likely British) psychiatrists, frequently occurs during training, and has significant personal and professional consequences. The programme could also be usefully extended to other mental health professionals in the context of evolving community services.