Re: Psychiatric Diagnosis and Suicide

Abstract

The paper of Bertolote, Fleischmann, De Leo, and Wasserman (Crisis, 2004, 25:147–155) is a potentially influential review of psychiatric diagnosis and suicide, as it purports to present data that diminish the importance of the treatment of mental disorders, particularly depression, in suicide prevention. However, it has a number of flaws that should be acknowledged before necessarily accepting their conclusions. From the methodological point of view one can not simply add the numbers of suicides from widely varying studies to arrive at valid data in terms of general population or psychiatric inpatient suicides. For example, subjects taken from a series of persons who have jumped from bridges or who have died by suicide in underground railways are hardly likely to be representative of a broader general sample of suicides in the community. Furthermore, the three largest studies from the UK, Denmark, and Israel, which comprised over 75% of what the authors referred to as general population suicides, were actually suicides of persons who had had psychiatric contact. Indeed, not only were many of those persons psychiatric inpatients at the time of death, but the authors of the UK study reported that their large sample comprised less than 25% of the total UK suicides. It is probable that psychotic illnesses were more likely to be represented in such psychiatric samples, therefore biasing the results. Bertolote et al. have also confined their analysis to traditional psychological autopsy-type data and have ignored more recently available population attributablerisk research, such as that from Denmark (Qin, Agerbo, & Mortensen, 2003), which allows for the relative importance of various contributing factors to be estimated. That research demonstrated clearly that psychiatric illness sufficiently severe to lead to hospitalization is of overwhelming importance in suicide at the broad population level, and that psychosocial issues such as being on a pension, having a low income, or being unemployed are of far lesser importance. The authors have also stated that they were interested in looking at the geographical origin of the published research, but those two studies from non-Western countries comprised only 216 subjects, or 1.4% of the total, and, therefore, one must be particularly cautious in considering that the overall results would apply to other nonWestern countries. The authors have suggested that strategies should not be “focusing exclusively on the identification and treatment of depression,” but I am not aware of any broad international programs that have been so focused, although there have been regional programs that have done so and that show promise in suicide reduction. The authors have also not distinguished the approaches to suicide prevention that may be appropriate in one country as opposed to another. The focus on mental illness, and in particular depression, is probably the most appropriate approach in developed countries, as many of the psychosocial stressors that may be more amenable to social manipulation have already been addressed. However, for countries with very high suicide rates, clearly addressing psychosocial factors will hold more promise in terms of reducing suicide. This has been illustrated in a model that integrates the role of biological and psychosocial factors contributing to suicide (Goldney, 2003), where it was postulated that there is a base rate of suicide caused by intractable, biologically based, mental disorders, albeit with a contribution from psycho-social stressors, but the relative importance of biological factors diminishes as suicide rates increase. There is a need to delineate more clearly the relative importance of psychiatric illness and psychosocial factors, not only in individual countries, but also in specific regions within countries where there may be unusually high suicide rates. The lumping together of disparate