Introduction by the Column Editor: In an article published in this journal in 2009, these authors presented initial findings from an evaluation of a system in the Netherlands for supervising suicides. Several such models have been established in the European region, all with the aim of improving services. The effectiveness of this process depends on trust between clinicians, managers, and the inspection system. Understandably, anxieties are easily aroused, and supervision’s effects are hard to predict. In this column, the authors offer further insights into experiences in the Netherlands. Of interest is the influence of clinicians’ culture-specific attitudes toward suicide, which may have an impact on expectations and actions.
Many people who die by suicide have received mental health care. To reduce suicide rates, several European countries have implemented reporting systems for suicides by mental health care recipients. The aim and operation of these systems vary. In the United Kingdom, information is gathered anonymously (1). Other countries, such as the Netherlands, Norway, and Sweden, require mental health services to report details about all suicides to supervisory organizations to identify structural problems in care and improve quality. Institutes or practitioners can be sanctioned, although this rarely occurs. These systems assume a relationship between suicide and the quality of care.
In the Netherlands, suicides must be reported by the responsible clinician and the medical director of the mental health service to the Health Care Inspectorate, an independent organization under the Ministry of Health, Welfare, and Sport. The inspectorate considers every such suicide as a potential “incident”—defined by the inspectorate as an unintentional or unexpected event related to the care provided that leads to the death or serious injury of a patient. Annually, about 650 such suicides are reported, representing about 42% of annual suicides in the Netherlands. A previous report indicated that about 40% of these suicide notifications were followed up with questions or remarks by the inspectorate (2). No cases of malpractice have been detected by the inspectorate in the past three decades.
For this study, we interviewed clinicians and mental health directors who reported a suicide and inspectors about their views of the usefulness of this procedure. Thirty-four of 38 mental health services participated, for a response rate of 89% (3). Data were gathered through in-depth interviews, conducted from November 2007 to July 2008. Services were 28 large mental health institutions, two addiction services, two general hospital psychiatric wards, and two private practices. Thirty-one clinicians were interviewed: 15 psychiatrists, one physician, nine mental health nurses, and six psychologists. Fifteen inspectors who dealt with suicide notifications were interviewed; four had backgrounds in psychiatry, two in psychology, six in mental health nursing, and three in other fields.
Among the 31 clinicians, 84% thought the supervision procedure was valid and useful. Most reported that the central function of the notification system was supervision of the quality of care. Evaluation of the suicide with colleagues was seen as valuable. However, for 18 clinicians, the notification procedure added stress to the difficulty of dealing with the suicide. Some were anxious about having made mistakes and feared criticism by the inspectorate. Nearly all of the 31 clinicians were ambivalent about the procedure’s usefulness in contributing to suicide prevention or major changes in daily practices. None thought that the suicide could have been prevented, and they considered the care to have been optimal. However, the suicide generally evoked feelings of guilt and failure.
All clinicians generally agreed that providing the highest standard of care is very important. However, about a third were skeptical about whether the effectiveness of suicide prevention in mental health services could be further improved, because suicides occur even with the highest-quality care. Suicides often occur unexpectedly and are hard to predict. Clinicians must take certain risks in treating suicidal patients. Because many patients who die by suicide are severely ill, suicide should be seen, according to the respondents, as a complication of treatment, comparable to complications of surgery for severely ill patients in general hospitals. Nonetheless about two-thirds of the clinicians thought it possible to improve suicide prevention in mental health care services.
All 28 medical directors agreed on the importance of critically reflecting on the quality of care provided to a person who later completes suicide and examining whether policy improvements are needed. The value of internal investigation by clinicians involved or by an internal suicide review committee was considered to be more important than external investigation by the inspectorate. However, all but one of the directors valued the external and independent role of the inspectorate in supervising the internal evaluation and quality of care. Most (85%) reported that the procedure underlines the importance of suicide prevention and keeps both medical directors and clinicians alert. In this context, most mental health institutions (85%) either had developed or were developing guidelines for the treatment of suicidal patients.
Many directors argued that suicide usually does not occur as a result of inadequate quality of care. The fact that every suicide should be reported seemed to imply that all suicides are preventable. Questions posed by inspectors were formulated in such a way that clinicians felt criticized or persecuted. In addition, disagreements with the inspectorate about critical aspects of treating suicidal patients frequently occurred; 56% of the directors argued that no valid risk assessment standards exist that would enable clinicians to reliably predict a suicide. Another point of disagreement was that some directors (30%) did not want to work in a restrictive manner in regard to suicidal patients, although in the directors’ view, the inspectorate was stimulating this. Most directors (93%) did not believe that the inspectorate had a clear influence on policy content.
All of the 15 inspectors agreed that it is difficult to judge suicide notifications uniformly. The way inspectors handled notifications differed considerably; some followed up with either questions or remarks in about 25% of the cases, and some responded to nearly every notification. The self-critical evaluations were considered to be a vital aspect of a notification by all inspectors. They saw the services’ attention to suicide prevention as an important advantage of the procedure, and most services had developed suicide prevention guidelines. But some inspectors felt that more improvement was possible in clinicians’ awareness and in undertaking action to prevent suicides. They felt that other advantages of the procedure were the attention given to suicide risk assessment and improved coordination between outpatient treatment and the clinical services of mental health care institutions.
Is supervision of suicides by mental health care recipients useful? One approach is to compare the Dutch system with systems in other countries. To our knowledge, the only published non-Dutch study of this topic is a study in Norway by Rønneberg and Walby (4), who concluded that 19% of suicides by mental health patients were not reported according to requirements and that few institutes seemed to improve the quality of care after suicides. In contrast, there is a long tradition of compliance in the Netherlands, where notification has been required since 1984. Comparing the Dutch system with the National Confidential Inquiry in the United Kingdom is more complex (1), because the aims and methods differ considerably. Advantages of the U.K. system are its anonymous, blame-free nature and the extensive data and knowledge that it has generated.
Our findings highlight medical directors’ and clinicians’ ambivalence about the usefulness of the procedure. The inspectorate’s role can be seen as motivating mental health services to critically self-reflect suicides. Directors and clinicians were unanimously positive in their assessment of the care provided before a suicide. Critically examining such care can lead to improvement in some cases. The fact that the inspectorate supervises this process underlines its importance. Ambivalence about the procedure’s usefulness centered on the issue of guilt implied by the incident-driven work of the inspectorate. A patient’s suicide can have a substantial traumatic impact on a treating clinician (5). Thus the stress evoked by the procedure is an unbeneficial effect. It may be of interest for the inspectorate to determine whether clinicians have been adequately supported by colleagues after a suicide.
In light of the guilt felt by and impact on clinicians, it was perhaps not surprising that none of them felt that the suicide could have been prevented. When clinicians were asked whether, in general, suicide can be prevented, they often answered differently. The general impression of the interviewers was that the absence of consensus on a Dutch standard practice guideline for assessment and treatment of suicidal patients played a significant role in clinicians’ views. The Dutch clinicians had no point of reference with which to compare their treatment policies. Until 2011, they had to rely on foreign guidelines, which do not always align with Dutch service delivery organization. Many reported the use of no-suicide contracts, and many had insufficient knowledge about risk assessment. Also, the pessimistic attitude about improving suicide prevention reported by about a third of the clinicians raises concerns about their willingness to implement policy changes and improvements. Perhaps the 2012 implementation of the new Dutch practice guidelines for assessment and treatment of suicidal behavior will lead to further improvements, especially through better training (6).
In many countries, suicide is regarded as resulting from substandard clinical care. This view is clearly not shared in the Netherlands, which may raise questions about whether Dutch clinicians’ reported inability to prevent some suicides is realistic or whether it reflects attitudes of clinicians toward autonomy or the right of self-determination of severely ill psychiatric patients. Our findings lead us to conclude that Dutch clinicians do their best to provide the highest-quality care in order to prevent suicides. But it is also true that they do not want to prevent suicide at any cost, and they emphasize patients’ autonomy, perhaps more so than in other countries.
The inspectorate’s requirement of a report for every suicide of a mental health care user implies a potential relationship between substandard care and suicide. Research in this area is scarce. Desai and colleagues (7) concluded in an exploratory study that suicide rates are probably not a useful indicator of the quality of care. Research on the relationship between suicide rates and the characteristics of mental health care has yielded conflicting results. Some studies have found a link between lower suicide rates and well-developed community-based mental health services, as opposed to inpatient-oriented services (8). Conversely, some ecological studies have found no association between suicide rates and mental health funding, service provision, or national mental health policies (9). However, a nationwide U.K. study by While and colleagues (10) showed that greater implementation by mental health services of National Confidential Inquiry recommendations was associated with lower suicide rates. Recommendations that were important in this context were provision of 24-hour crisis care, local policies about patients with dual diagnoses, and a multidisciplinary review after a suicide. These findings support the emphasis that the Dutch inspectorate puts on evaluation after a suicide. Also, it would be advisable for the inspectorate to supervise the implementation of suicide prevention measures—both in general, such as provision of adequate 24-hour crisis care, and in terms of specific improvements within a service.
In the Netherlands, the inspectorate changed the notification procedure in May 2011, so that most suicides can be reported in a blame-free manner within the mental health service. Only suicides of patients in involuntary treatment or for which the medical director determines an association with substandard care must be reported to the inspectorate. Services still must evaluate each suicide internally and provide an annual report to the inspectorate of an overview of suicides and resulting policy adjustments. It is hoped that the new procedure will reduce feelings of guilt. However, data about suicides and lessons after evaluations will no longer be collected systematically on a national level, even though such procedures may be viewed as critical for generating recommendations to improve suicide prevention policies. The new procedure will be evaluated.
This study was funded by the Health Care Inspectorate in the Netherlands.
Dr. Robben is an appointed advisor to the Dutch Health Care Inspectorate, which funded the research. The other authors report no competing interests.