The lifetime risk of suicide in schizophrenia: a reexamination

The psychiatry literature routinely quotes a lifetime schizophrenia suicide prevalence of 10% based on 1 meta-analysis and 2 studies of chronic schizophrenics.

To build a methodology for extrapolating lifetime suicide prevalence estimates from published cohorts and to apply this approach to studies that meet inclusion criteria.

We began with a MEDLINE search (1966-present) for articles that observed cohorts of schizophrenic patients. Exhaustive bibliography searching of each identified article brought the total number of articles reviewed to 632.

Studies included in the meta-analysis observed a cohort of schizophrenic patients for at least 2 years, with at least 90% follow-up, and reported suicides. Articles are excluded for systematic age bias (ie, adolescents).

Extracted data included sample size, number of deaths, number of suicides, percentage of follow-up, and diagnostic system used. Data were extracted independently by 2 of us, and differences were resolved by consensus after re-review.

Studies were divided into 2 groups: 32 studies of schizophrenics enrolled at various illness points (25 578 subjects) and 29 studies of schizophrenics identified at either illness onset or first admission (22 598 subjects). Regression models of the intersection of proportionate mortality (the percentage of the dead who died by suicide) and case fatality (the percentage of the total sample who died by suicide) were used to calculate suicide risk in each group. The estimate of lifetime suicide prevalence in those observed from first admission or illness onset was 5.6% (95% confidence interval, 3.7%-8.5%). Mixed samples showed a rate of 1.8% (95% confidence interval, 1.4%-2.3%). Case fatality rates showed no significant differences when studies of patients diagnosed with the use of newer systems were compared with studies of patients diagnosed under older criteria.

This study estimates that 4.9% of schizophrenics will commit suicide during their lifetimes, usually near illness onset.