(121) Patterns of Psychotropic Medicine Use and Initiation in The Year Prior to Death by Suicide: A Population-Based Study

Publication Number: 795

Background: Prescribed medicines are commonly used to treat mental illnesses that increase the risk of suicide but are also often implicated in suicide death by poisoning. Understanding how medicine use changes before death by suicide can help to identify potentially harmful medicine use practices.

Objectives: To quantify changes in dispensing and initiation of antidepressants, benzodiazepines, and antipsychotics in the year prior to death by suicide.

Methods: We used national coronial data linked with medicine dispensing claims for all people who died by suicide in Australia (2013-2019). We used a data-driven approach to identify structural breaks (breakpoints) in linear regression models of weekly dispensing and initiation rates for all medicines over time.

Results: Our study included 14,344 people (24% female, median age 44 years). In the year prior to death, women were more likely than men to be dispensed antidepressants (62% [n=2164] vs 42% [n=4535]), benzodiazepines (46% [n=1600] vs 29% [n=3126]), and antipsychotics (26% [n=892] vs 17% [n=1814]). For antidepressants, the dispensing rate increased by 0.3% per week (95%CI 0.2%-0.4%) until a breakpoint 15 weeks prior to death, when it started increasing by 1.8% per week (95%CI 1.5%-2.1%). For benzodiazepines, dispensing increased by 0.4% per week (95%CI 0.4%-0.5%) until 9 weeks prior to death, increasing to 4.8% per week (95%CI 4.1%-5.6%) after. For antipsychotics, dispensing increased by 0.2% per week (95%CI 0.1%-0.4%) until 23 weeks prior to death, increasing to 1.0% per week (95%CI 0.8%-1.3%) after. In the 6 months prior to death, treatment-naïve women were more likely than men to initiate antidepressants (26.2% [n=523] vs 18.4% [n=1553]), benzodiazepines (21.4% [n=568] vs 15.2% [n=1479]), and antipsychotics (9.9% [n=293] vs 6.0% [n=596]). Breakpoints in the weekly rate of initiation of antidepressants and benzodiazepines were observed at approximately 3 months prior to death, and 4 weeks prior to death. In the 4 weeks prior to death, the initiation rate increased rapidly by 16.2% per week (95%CI 5.0%-28.5%) for antidepressants and 18.2% (95%CI 9.9%-27.4%) per week for benzodiazepines. The initiation rate was highest the week prior to death for both medicine classes. No breakpoint was observed in antipsychotic initiation over time. Similar patterns of change were observed for men and women.

Conclusions: Women who died by suicide were more likely to be dispensed psychotropic medicines than men. In both sexes, dispensing and initiation increased more rapidly closer to the date of death, with benzodiazepine initiation seeing the most dramatic changes in the month before death. This may reflect worsening of symptoms or stockpiling, and indicate an opportunity for intervention.