In a case-control study, the basic population were patients who between 1968 and 1985, when aged 15-24 years, were admitted to the regional poisoning treatment centre because of deliberate self poisoning or self injury. Cases (n=62) consisted of those who by the end of 1985 had died locally from either suicide (n=41) or possible suicide (n=21). Controls (n=124) were patients who were known not to have died locally during the study period. Two matched controls were selected for each case.
Univariate analysis showed that risk of death due to suicide and possible suicide was associated with six factors: social class V (OR 2.7, 95% CI 1.1 to 6.7), unemployment (2.8, 1.4 to 5.8), previous inpatient psychiatric treatment (4.9, 2.2 to 10.9), substance misuse (3.3, 1.6 to 6.8), personality disorder (2.1, 1.03 to 4.4), and previous attempted suicide (2.3, 1.2 to 4.4). Multivariate analysis identified two factors as significantly contributing to the model that best discriminated between the cases and controls: substance misuse (alcohol or drugs, or both) (adjusted OR 3.9) and previous inpatient psychiatric treatment (3.7). These factors seemed to be associated with suicide after attempted suicide in both the short term (less than 12 months) and the long term (one year or more) and were also identified when the analysis was restricted to subjects who definitely died by suicide and their controls.
Primary/Secondary Keywords