処方ベンゾジアゼピン治療開始と自殺行動のリスク

Benzodiazepines (BZD) are widely prescribed for mental health conditions, but research has been unable to determine the effects of initiating BZD on suicidal behavior. This study examined the temporal dynamics of suicidal behavior surrounding BZD initiation and explored patterns of risk within patient subgroups.

We examined incident BZD users (N = 691,517, ages 13-64) from Optum's de-identified Clinformatics® Data Mart Database from 2016-2019. We compared risk of suicidal behavior immediately before initiation (60 days prior) and during medicated periods (1-30 days, and 31-365 days post initiation) to a baseline period 3-12 months prior to initiation within individuals, which accounts for all time-stable confounding. Suicidal behavior was defined by ICD-10-CM codes for suicide attempts and intentional self-harm during emergency department encounters and hospitalizations. We statistically adjusted for measured time-varying covariates, including other medications and psychosocial treatment. Additionally, we ran analyses stratifying by psychiatric diagnosis to assess risk in different patient populations.

Compared to baseline, risk of suicidal behavior was elevated in the period before initiation, OR=3.54; 95% CI, 3.23-3.89. Risk further increased in the first 30 days after initiation [OR=5.05; 95% CI=4.41-5.77] and remained elevated during the 31-365 days after initiation [OR=3.62; 95% CI=3.08-4.26]. Stratifying by diagnosis revealed similar temporal patterns.

We found elevated risk for suicidal behavior occurred prior to BZD initiation. Risk was subsequently highest during short-term treatment but remained elevated during longer-term treatment. More research is needed to understand the underlying mechanisms, such as the pharmacologic effects of BZD initiation or factors associated with reasons for initiating.