Delicate Considerations and Blunt Instruments

In addition to this month’s In Review,^1–3 the current issue of The Canadian Journal of Psychiatry contains 2 articles on suicide. Dr Angela Onkay Ho⁴ draws on historical, philosophical, and legal arguments to explore the question of whether suicide can be rational. Dr Anne E Rhodes and colleagues⁵ present data on time trends for suicide-related behaviours in Ontario. At first glance, the 2 papers seem to have little in common: 1 of them pursues a narrative exploration and the other an epidemiologic analysis. However, similarities lurk below the surface. Can suicide occur in the absence of, or be uninfluenced by, a psychiatric disorder? Using terminology familiar to epidemiologists, this is a question of whether psychiatric disturbances are a necessary cause of suicide. It would be very surprising if the answer was yes. With the exception of infectious diseases, very few contemporary health problems have necessary causes. Rather, almost all are determined by multiple component causes that interact with each other to form complex causal mechanisms. Mental disorders, in themselves, are neither necessary nor sufficient causes of suicide, but are examples of component causes. Primary prevention addresses component causes.

The epidemiologic logic of causal reasoning aligns well with that of clinical practice, although a different terminology predominates. The language of clinical practice uses familiar terms, such as “predisposing” and “precipitating” factors, “diathesis” and “vulnerability,” rather than the component cause terminology typically favoured by epidemiologists. However, both chafe against the more blunt reasoning associated with practice guidelines, and medicolegal anxieties. Regulatory decisions are also blunt instruments that can fall prey to simplified etiological assumptions. Discouraging the use of antidepressants in adolescents will reduce the risk of suicidal thoughts and actions, if antidepressants are a component cause of these outcomes, but the same regulatory warning may increase exposure to another component cause: untreated depression. The quantification of real-world risk, as provided by Dr Rhodes and colleagues’ study,⁵ is therefore of critical importance.

Guidelines and regulatory actions, no matter how sophisticated, are all based on generalities and cannot be expected to determine clinical decisions in all circumstances. Ultimately, the uniqueness of each clinical situation is something that clinicians and their patients need to address together within the societal and legal context they share. A “delicate consideration of clinical judgment, duty of care, and legal obligations”^{4, p 141} seems like a good place to start.