Suicide prevention: towards an evidence-based policy

Suicide prevention and adequate management of those with suicidal behaviour should be one of the major drivers in designing clinical services and guiding mental health policy decisions at all levels. In order for these decisions and this guidance to be effective, they need to be based on relevant and robust evidence covering both the magnitude of the problem at the local level where the decisions aim to have an impact, and on the effect that potential interventions may have on suicide and suicidal behaviour at the clinical and the population level. This issue of Epidemiology and Psychiatric Sciences includes two valuable contributions to the theoretical framework that needs to be used for an evidence-based policy on this area.

The contribution of Blasco-Fontecilla and collaborators (Blasco-Fontecilla et al., 2018) presents and argues the advantages and potential applicability of an indicator based on attempted suicides at the population level. The most widely used indicator for monitoring suicide at the national level is the annual age-standardised suicide rate per 100 000. It is employed to rank suicide mortality within countries by causes of death but also for cross-country comparisons, to identify trends over the years and to establish global, regional and/or national targets (WHO, 2014; NCD Countdown, 2030 collaborators, 2018). However, the global mortality figures provide limited information for planning, selecting, and evaluating targeted preventive interventions, particularly at the regional and local levels. For a proper analysis of the problem we need incorporate indicators related to risk factors that could be the specific target of interventions, and that could be sensible to change as a consequence of those interventions. Attempted suicide (AS) is the most relevant risk factor for completed suicide, and it is widely used as an outcome in longitudinal studies that focused on suicidal behaviour on both clinical and population samples (Hegerl et al., 2009; Arana et al., 2010). Thus the relevance of the proposal of Blasco-Fontecilla and his colleagues of a ratio of ASs/completed suicides as an indicator to be used in guiding policy at the population level. They list three potential measures of AS to be included in the numerator: annual AS rates from population surveys, annual AS rates from clinical registries and lifetime AS from population surveys. Because of the lack of worldwide information for the first two choices, they select the third option and calculate the proposed ratio in several countries to illustrate its potential use. There are limitations on their approach that are appropriately discussed in their editorial. As previously recognised, the information about past history of suicidal attempts collected in epidemiological surveys is by its nature retrospective and self-reported, and is thus subject to recall bias, underreporting and denial (Miret et al., 2014). In addition, it is not an adequate method for identifying trends in different age ranges or of properly evaluating preventive interventions. Their contribution highlights the need to advance towards a systematic collection, and reporting of information on the annual rate of suicidal attempts for refining our analysis capacity of suicidal behaviour at the population level for both monitoring and evaluation of interventions and policies.

As in other areas of medicine, the basis for the prevention of suicide is the identification of risk factors that are relevant to the context, and their alleviation by implementing appropriate interventions (WHO, 2014). However, suicidal behaviour is the result of a complex array of contributing factors that are in many circumstances interconnected, as illustrated in this issue of Epidemiology and Psychiatric Services by Hegerl and Heinz (Hegerl and Heinz, 2018). They critically discussed two different theoretical explanatory models concerning the causal relationships between psychosocial factors, depression and suicides. Ulrich Hegerl and his team have widely and successfully implemented suicide prevention programmes in several countries focusing on improving the care of people with depression (Székely et al., 2013), and thus favour an explanatory model that attributes a strong casual role to depression and other psychiatric disorders in explaining suicidal behaviour. They confront their favoured model with an alternative one that assumes that social determinants and life adversities are the major causes of suicidal behaviour. Their arguments clearly reflect the daily experience of clinicians attending to patients with suicidal ideation and behaviour. However, it needs to be highlighted that the dramatic decline of global suicide mortality observed in the last few decades (NCD Countdown, 2030 collaborators, 2018) can only be explained if we consider big social changes affecting urbanisation, social stability, social freedom and policies that have an impact on the consumption of alcohol, the improvement on the lives of the elderly, and/or access to the means to kill oneself, among others. It is clear that no single approach can solve the complex challenge that is the reduction of suicide mortality, and that there is still room for simultaneously testing and implementing measures that may have an impact at different modifiable factors.