OBJECTIVES

Among adolescents and young adults with nonfatal self-harm, our objective is to identify risk factors for repeated nonfatal self-harm and suicide death over the following year.

METHODS

A national cohort of patients in the Medicaid program, aged 12 to 24 years ( = 32 395), was followed for up to 1 year after self-harm. Cause of death information was obtained from the National Death Index. Repeat self-harm per 1000 person-years and suicide deaths per 100 000 person-years were determined. Hazard ratios (HRs) of repeat self-harm and suicide were estimated by Cox proportional hazard models. Suicide standardized mortality rate ratios were derived by comparison with demographically matched general population controls.

RESULTS

The 12-month suicide standardized mortality rate ratio after self-harm was significantly higher for adolescents (46.0, 95% confidence interval [CI]: 29.9-67.9) than young adults (19.2, 95% CI: 12.7-28.0). Hazards of suicide after self-harm were significantly higher for American Indians and Alaskan natives than non-Hispanic white patients (HR: 4.69, 95% CI: 2.41-9.13) and for self-harm patients who initially used violent methods (HR: 18.04, 95% CI: 9.92-32.80), especially firearms (HR: 35.73, 95% CI: 15.42-82.79), compared with nonviolent self-harm methods (1.00, reference). The hazards of repeat self-harm were higher for female subjects than male subjects (HR: 1.25, 95% CI: 1.18-1.33); patients with personality disorders (HR: 1.55, 95% CI: 1.42-1.69); and patients whose initial self-harm was treated in an inpatient setting (HR: 1.65, 95% CI: 1.49-1.83) compared with an emergency department (HR: 0.62, 95% CI: 0.55-0.69) or outpatient (1.00, reference) setting.

CONCLUSIONS

After nonfatal self-harm, adolescents and young adults were at markedly elevated risk of suicide. Among these high-risk patients, those who used violent self-harm methods, particularly firearms, were at especially high risk underscoring the importance of follow-up care to help ensure their safety.