Comparison of risk factors in fatal and non-fatal repetition of self-harm: A population-based cohort study of 84,430 Taiwanese adults who presented to hospital with self-harm

BACKGROUND: Individuals who self-harm show an increased risk of self-harm repetition. We aimed to compare factors associated with fatal and non-fatal self-harm repetition among adults who presented to hospitals with self-harm. METHODS: We conducted a cohort study using data from Taiwan's National Suicide Surveillance System (NSSS) from 2012 to 2017, linking data to national mortality records to identify fatal repetitions (i.e., suicide). We calculated incidence rates for repeated self-harm and suicide, examined risk factors using Cox proportional hazards models, and calculated population attributable fraction (PAF) for each factor. RESULTS: The overall incidence rate of non-fatal and fatal self-harm repetition was 67.6 (95% confidence interval [CI] 66.5-68.6) and 10.4 (95% CI 10.0-10.8) per 1000 person-years, respectively. Risk factors for non-fatal self-harm repetition included female sex, younger age, being divorced/separated, lower education, overdose as the index self-harm method, a history of psychiatric disorders, and psychiatric and non-psychiatric hospitalizations in the past year. Risk factors for fatal self-harm repetition included male sex, older age, being single/divorced/separated, using higher lethality methods at the index episode, and a history of psychiatric disorders. Depressive disorders showed the highest PAF (19.28%) for non-fatal self-harm repetition, while male sex showed the highest PAF (20.56%) for fatal repetition among all risk factors. CONCLUSIONS: Non-fatal and fatal self-harm repetition shared some risk factors but also had distinct ones. A better understanding of the shared and distinct mechanisms underlying the risk of non-fatal and fatal self-harm repetition can improve intervention strategies.