In Norway, the suicide rate has not declined during the last 20 years, despite numerous action plans, better knowledge and improved treatment. Most people who take their own lives are not in psychiatric treatment, and societal conditions are a main reason why the suicide rate is not falling.
When Norway established a national programme for suicide prevention in 1993, it was the second country in the world to do so after Finland. The first action plan to combat suicide was added two years later. In 2008, guidelines for suicide prevention in mental health care were published, followed by the last action plan against suicide and self-harm to date. The Storting recently decided to prepare a new action plan.
Despite these efforts, the suicide rate has remained largely unchanged in Norway from 1995 to 2015 (Figure 1) (1–6). Over the same period there has been a significant decline in Denmark, while the trend in Norway and Sweden has generally been similar. In other words, for 20 years, we have had programmes aiming to reduce the suicide rate – apparently to no avail.
Lack of evaluation
Øivind Ekeberg had the main responsibility for preparing the national programme in 1993. When the action plan was launched in 1995, a number of proposed measures had been curtailed or cut. For example, it was recommended to establish regional resource centres for suicide prevention in the university cities with six to nine staff in each, but only one and a half full-time equivalents were established in each location.
For action plans to be useful, the measures require structured follow-up and evaluation, with participation by international experts. When the most recent action plan was launched in 2014, it failed to take previous plans into account with evaluations of measures to be expanded or reduced.
There has been some criticism of the suicide prevention measures because of their one-sided biomedical concept of knowledge with a special focus on depression (7–9). The critics claim that there is weak evidence for saying that at least 90 % of all suicide victims have had an underlying mental disorder (7). This view is debatable. We have met thousands of patients who have attempted suicide, and all of them suffered from problems that fulfilled the criteria for a psychiatric diagnosis and thus also for follow-up in the primary or secondary health services. Common disorders in suicidal behaviour include affective disorders, personality disorders and behavioural disorders caused by psychoactive substances. Many also suffer from more situationally dependent conditions, such as acute stress reaction, which lasts for some days, or adjustment disorder, which commonly resolves in six months.
Prescription of antidepressants showed a multifold increase in the Nordic countries in the years 1995–2005. However, this did not lead to a decline in the suicide rate.
Hagen and co-authors call for more attention to be paid to contextual issues (7). It is easy to agree that the context in which the patient lives is crucial and may trigger an acute stress reaction, adjustment disorder and other reactive conditions. When teaching university students as well as in the specialist curriculum, emphasis is placed on a biopsychosocial understanding of the patients’ condition (10). In our experience, most of those who use diagnoses do not take a one-dimensional approach, at least not as far as treatment is concerned. Those who follow up patients after a suicide attempt focus on cognitive, psychodynamic, family, relational and social problems, and to a lesser extent on biomedical or drug-related issues, even though these also may often form an integral part of the follow-up.
In Norway, the suicide rate doubled over the twenty-year period until 1988. Significant changes have also occurred in other countries. For example in South Korea, the rate increased from 11 to 30 per 100 000 per year over the period 1993–2009 (11), while the rate in Estonia fell from 30 to 18 from 2000 to 2015 (12). We have little knowledge of why such major shifts occur. Since there is no equivalent change in the prevalence of mental disorders or in the therapeutic methods, changed social conditions are likely to play a crucial role.
Treatment programmes have emphasised diagnosis and treatment of depression, and the prescription of antidepressants showed a multifold increase in the Nordic countries in the years 1995–2005. However, this did not lead to a decline in the suicide rate (Figure 1) (12). In Norway and Sweden, the most marked decline in the suicide rate occurred in the period 1990–95, i.e. before the launch of the action plan.
Procedures versus empathy
Increasing attention has been devoted to suicidal behaviour during psychiatric treatment. A study of the period 2009–11 found that nearly 60 % of all those who took their own lives in Norway were individuals who had had no contact with mental healthcare services over the preceding year (13). Among those who had, the contact is likely to have been sporadic for many months prior to the suicide. This concurs with a meta-analysis showing that 3.7 % of the suicides were committed by patients during hospitalisation in a psychiatric ward, 18 % by patients who had been in contact with an inpatient ward during the last year prior to their suicide and 26 % by patients who had been in contact with an inpatient ward or received treatment as outpatients (14).
A goal might be set that no suicides should occur while a patient is hospitalised in a psychiatric ward. However, there is a need to remain realistic and keep in mind that serious suicide risk is a main reason for hospitalisation, and in order to achieve this goal, the staff would need to implement an undesirably extensive monitoring regime.
Hagen and co-authors ask the pertinent question of whether efforts to prevent suicide in mental healthcare today have become more a matter of implementing procedures than of facilitating the best possible care for the patients (7). The use of suicide risk assessments is especially controversial (7, 15). We believe that they can be useful in predicting suicides in the short term (16, 17). If they were not, it should not be reasonable to have a law that accepts serious suicide risk as a supplementary criterion for coercive hospitalisation. A methodological problem in this context is that it appears to be wrong to assume that a patient will take his or her own life, if an intervention prevents it. Moreover, it is hardly necessary to be able to predict who will kill themselves in order to prevent suicides. We have seen a reduction in the number of road traffic fatalities without any attempts to predict who will die in traffic accidents.
What should we do?
There is no reasonable relationship between clinical research on suicidal behaviour on the one hand and the scope of the problem on the other. In addition, most of the research tends to focus on risk factors and epidemiology, and less on the effect of treatment. In particular, we need more knowledge about what kinds of interventions are effective.
There is also a dearth of studies on the characteristics of people at risk of suicide who are referred to mental health care, what happens during their treatment and the follow-up provided to the patients when their treatment is concluded. We know little of whether the patients have been helped to solve their problems and the empathy and care that were provided to them.
If we assume that 600 suicides occur in Norway each year and that 5 % of them occur during hospitalisation in a psychiatric ward, this will mean that approximately 30 suicides are committed by patients during hospitalisation. In 2017, more than 150 000 patients were treated in mental healthcare institutions, and more than 25 000 had been admitted to inpatient wards (18). This gives a suicide rate of 120 per 100 000 hospitalised patients per year, which is approximately ten times higher than the suicide rate in the general population. Acute suicide risk is a common cause of hospitalisation, and if we assume that this applies to approximately one-half of all hospitalisations (19) this will mean that approximately 0.2 % of all those who were admitted because of acute suicide risk took their own life during their hospitalisation. It is clearly to be deplored that 30 patients die during hospitalisation, although we should not forget that during the same period approximately 13 000 were helped to overcome their suicidal crisis. This example shows that in any case, better suicide prevention during hospitalisation is likely to have no more than a marginal effect on the suicide rate. Improving the follow-up after suicide attempts and identifying those who are approaching a suicidal crisis would be likely to have a better effect. This would provide for better opportunities to help the patient overcome the problems that caused the crisis.
There is no reasonable relationship between clinical research on suicidal behaviour on the one hand and the scope of the problem on the other
Many are critical of the requirement to assess suicide risk as stipulated by the guidelines. They believe that people who are committed to a psychiatric inpatient ward as a result of suicidal behaviour need to meet empathic health workers who are able to establish good contact and seek to understand, recognise the patient’s thoughts and feelings and involve him or her in the treatment process (7). Others criticise the systematic use of suicide risk assessment because they believe that this approach may divert attention away from exploring the patient’s experience towards a clinical examination, where the patient is a supplier of information to the expert who assesses risks and measures (15). They also believe that such an examination may intensify the sense of loss of control, since the expert thereby assumes the responsibility for the patient’s life. In our opinion, there is no contradiction involved in this (16, 17). It is fully possible to undertake an assessment of suicide risk and implement preventive measures while remaining attentive and empathic and showing interest in the patient’s life and history. In many situations, however, the procedures may become too extensive, both in cases where it is quickly made clear that the patient is not at acute risk of suicide and when repeated assessments have been made without showing any changes in the patient’s clinical condition or life situation. Drawing a parallel with cardiology, there is no contradiction between listening to the patient’s experiences and symptoms on the one hand and assessing whether an acute myocardial infarction has occurred on the other. Nor is it necessary to perform a coronary angiography each time the patient comes in for an examination, unless a new clinical situation has arisen.
Most suicides occur among people who are not in contact with the health services. Identifying suicidal persons who have not come in for treatment should thus be a main goal for further measures.
The second main goal is associated with measures within the health services. Here, there is least to be gained in relation to the admissions to psychiatric wards, where only a minor proportion of the suicides take place. The primary health services could probably improve their ability to identify persons who are at risk, in order to refer them to other services. There is a potential for improvement also in the specialist health services. The treatment objective should be not only to solve the acute suicide crisis, but also improve the patient’s mental health.
We may have a lot to learn from prevention of cardiovascular diseases. Although the measures aimed at the most ill patients have steadily improved, the greatest reduction in cardiovascular morbidity and mortality has come as a result of lifestyle changes. Suicide prevention efforts will probably also be more effective if we succeed in improving the mental health of the population.