Predictive Properties of the C-SSRS Suicide Screen Among Individuals in Veterans Health Administration Care.

OBJECTIVE: In 2018, Veterans Health Administration (VHA) mandated adoption of the Columbia-Suicide Severity Rating Scale (C-SSRS) for health system suicide screens. This study evaluates the predictive validity of C-SSRS screens in VHA specialty mental health (SMH) and emergency department (ED) settings. METHODS: For 867,023 patients screened in SMH clinics from 10/1/2018-11/30/2020 and 1,035,317 screened in EDs from 11/1/2019-11/30/2020, proportional hazards analyses evaluated whether C-SSRS screens (rated as minimal, low, moderate, or high risk) were associated with non-fatal suicide attempts and with suicide deaths within the subsequent 7, 30, and 365 days. Analyses adjusted for patient characteristics, prior year mental health treatment, prior year suicide attempts, and most recent VHA REACH VET suicide prediction algorithm risk score. RESULTS: For the SMH cohort, 61.27% of those who died by suicide and 78.29% of those who attempted suicide within a year initially screened as low, moderate, or high risk; for the ED cohort, 15.39% of those who died by suicide and 46.29% of those who attempted suicide within a year initially screened low, moderate, or high. In both settings, C-SSRS risk level was positively associated with risk of suicide attempts and deaths (hazard ratios ranged from 1.76 to 6.62 depending on outcome, setting, and risk level), adjusting for covariates. CONCLUSION: In VHA SMH and ED populations, the C-SSRS identifies patients who are at increased risk of suicide deaths and attempts. Follow-up should be completed for patients who screen at low risk or higher on the C-SSRS, particularly in EDs.