High mortality by natural and unnatural causes: a 10 year follow up study of patients admitted to a poisoning treatment centre after suicide attempts.

OBJECTIVE To describe mortality by suicide and other causes of death in a group of patients who attempted suicide, and to identify predictive factors. DESIGN 10 year follow up study based on records of suicide attempters in 1980. SETTING Poisoning treatment centre at a general hospital. SUBJECTS 974 patients aged 15 and over referred to the poisoning treatment centre after deliberate self poisoning. MAIN OUTCOME MEASURES Death by different causes registered in the Danish death cause register. RESULTS In 10 years of follow up 306 patients died: 103 by suicide, 131 from natural causes, and 31 by accident; five were murdered, and in 36 cases the cause of death was uncertain. The standard mortality ratio was 550. Cause specific standardised mortality rates were 2960 for suicide, 236 for natural causes, 1256 for accidents, and 5459 for uncertain causes. In a Cox regression analysis, high risk factors for subsequent suicide were: more than one previous suicide attempt (relative risk 2.25), living alone (2.28), and age (1.03 per year). Predictors of death by natural causes were receiving a pension (1.69), drug misuse (2.72), more than one previous suicide attempt (2.25), age (1.06 per year), and male sex (2.49). The group of patients fulfilling at least one high risk criterion for later suicide differed significantly from the rest of the patient group in incidence of suicide, but both sensitivity and specificity were low. CONCLUSIONS Most patients who attempted suicide were at high risk of succeeding because the risk factors, though significant, are not very specific. A strategy to prevent suicide must be directed toward the majority of those who attempt suicide.