Published online by Cambridge University Press: 14 February 2017
It has been suggested that screening interventions may be effective for suicide prevention. Few studies, however, have reported their effects on outcome measures, including death by suicide among middle-aged adults.
We used a quasi-experimental parallel cluster design with matched community-based intervention and control municipalities (total eligible population: 90 000) in Japan. At-risk residents within the intervention area were invited for universal depression screening and subsequent care/support. We compared changes in suicide incidence of adults aged 40–64 years for the 4-year pre- and post-implementation periods in the intervention group with the control group and the whole country. Incidence rate ratios (IRRs) of the outcomes were adjusted for age group, gender and interaction terms, using mixed-effects negative binomial regression models. Suicide rates among intervention and control subgroups were compared.
The screening procedure was offered to 52% of the intervention group, and 61% of those contacted responded over the implementation period. Suicide rates decreased more in the intervention group [IRR 0.57, 95% (CI) 0.41–0.78; F 1,36 = 12.52, p = 0.001] than the control group (IRR proportion 1.63, 95% CI 1.06–2.48; F 1,82 = 5.20, p = 0.025) or the whole country (IRR proportion 1.64, 95% CI 1.16–2.34; F 1,42 = 8.21, p = 0.006). Sensitivity analyses confirmed the results from the primary analysis. There were lower suicide rates among both respondents and non-respondents to the screening than in the control group during the implementation period.
Prevention efforts involved in the depression screening intervention were probably successful in reducing suicide rates.