Isotretinoin and risk factors for suicide attempt: a population-based comprehensive case series and nested case-control study using 2010-2014 French Health Insurance Data

Background Although the causal role of isotretinoin in suicidal behaviour is controversial, suicide attempts (SA) do occur among patients taking isotretinoin. Objectives To describe patient profiles and the management of isotretinoin among patients who committed or attempted suicide under treatment. To assess the risk factors for SA under isotretinoin. Methods We performed a comprehensive case series of suicides and SAs under isotretinoin, and a case-control study, using Nationwide French Health Insurance database. The main analysis compared cases (subjects with a SA during a course of isotretinoin) to controls, individually matched for age, gender and rank of the current course; controls were to be exposed to isotretinoin at the index date (date of SA for the corresponding cases). The patients' psychiatric history at isotretinoin initiation was studied. In a secondary analysis, patients who continued their isotretinoin treatment after their SA were compared to patients who discontinued it. Results In all, 328 018 subjects started a course of isotretinoin between 1 January 2010 and 31 December 2014 and 184 patients were hospitalized for a SA; half of them had a psychiatric history at initiation. In the multivariate analysis, psychiatric history and history of anxiety alone were risk factors for SA [Odds ratio (OR), 18.21; 95% confidence interval (CI), 9.96-33.30 and 4.78; 95% CI, 2.44-9.33, respectively]. Among 176 cases of SA with sufficient follow-up, 103 (58.5%) carried on with their treatment after their SA. Treatment initiation by a dermatologist was inversely associated with the continuation of the treatment after a SA (OR, 0.38; 95% CI, 0.18-0.80). Conclusions Suicide attempts under isotretinoin are rare events, and our results suggest that most of the patients concerned have a risk-prone profile detectable at the time of treatment initiation. The risk-benefit ratio of continuing isotretinoin after a SA warrants further careful evaluation.