Implementation of an integrated community-based suicide prevention programme, Gujarat, India: cluster randomised controlled trial.

Abstract

In low- and middle-income countries (LMICs), suicide is a major problem. Research on the effectiveness of large-scale suicide prevention interventions is limited. To test the effectiveness of an integrated intervention (school-based prevention; reducing access to means of suicide; increased identification and management of suicide risk) in reducing deaths by suicide and suicide attempts; and to evaluate the implementation and effectiveness of sub-interventions. In this pragmatic cluster randomised controlled trial, 124 villages from Mehsana, India, were randomly assigned to either intervention or control arm. The intervention comprised school-based awareness intervention, community pesticide storage and training of community health workers (CHWs) to recognise, support, refer and follow up people at risk. Intention-to-treat analysis using mixed-effects Poisson regression tested the primary outcome (suicide attempts plus deaths by suicide), and multilevel linear models assessed sub-interventions. The primary outcome was captured through a novel suicide surveillance system. There was no statistically significant difference in the primary outcome between the intervention (54 of 62 consenting villages) and control (62 villages) arms. Separately, the intervention arm showed a 43% reduction in risk of death by suicide at 12 months (suicide rate 30.7 versus 43.6 per 100 000 person-years in intervention versus control arm; incidence rate ratio 0.57, 95% CI: 0.32-1.02, adjusting for baseline and clustering). Most students (≥90%, <i>n</i> = 2330/2560) from 47 schools received the intervention and had lower depression and suicidal ideation than controls at month 3. Nearly all villages (52/54, 96.2%) provided pesticide lockers (<i>n</i> = 8370 households, 88.83% uptake). Compared with controls, CHWs in the intervention arm had significantly higher knowledge, confidence and skills, and identified 108 at-risk individuals. The intervention increased identification without significantly reducing suicide attempts, but reduced suicide deaths. This trial, involving 116 villages and a multicomponent intervention implemented at scale, advances suicide prevention and complex intervention research, especially in LMICs.