Psychiatric diagnoses are associated with means selection in united states suicide deaths.

PURPOSE: The purpose of this study was to examine whether the choice of means by persons who die by suicide is associated with a prior psychiatric diagnosis. METHODS: In this cross-sectional study, we analyzed suicide surveillance data from 18 states reporting to the National Violent Death Reporting System (NVDRS) between 2003 and 2014. NVDRS compiled data from multiple sources (e.g., coroner's reports, police reports, death certificates) on every violent death within reporting jurisdictions, including information on indicated psychiatric disorders and suicide means. We assessed whether the selected suicide means were associated with diagnoses using multinomial logistic regression. RESULTS: Adjusted models suggested that, compared to decedents using firearms, those using poisoning were more likely to have each psychiatric disorder examined, including bipolar disorder (aOR: 2.17 [95% CI 2.03-2.32]), schizophrenia (aOR: 1.81 [1.61-2.04]), depression (aOR: 1.64 [1.58-1.70]), anxiety disorder (aOR: 1.46 [1.35-1.57]), and PTSD (aOR: 1.41 [1.22-1.64]). A far greater proportion of individuals who died from less common means (other than firearms, suffocation, or poisoning) had schizophrenia (aOR: 4.52 [4.00-5.11]). CONCLUSIONS: Many existing and proposed means restriction interventions have focused on firearms. Additional focus on access to potential agents of poisoning (e.g., the type and quantity of medication administered to patients) among individuals with psychiatric diagnoses may be warranted.