HIV Insight: Suicide Mortality among People Accessing Highly Active Antiretroviral Therapy for HIV/AIDS in British Columbia: A Retrospective Analysis

Suicide rates have been reported at elevated levels among people living with HIV/AIDS. We sought to characterize longitudinal suicide rates among people living with HIV/AIDS who are accessing free highly active antiretroviral treatment (HAART) in British Columbia and evaluate the sociodemographic, clinical and behavioural factors associated with suicide in this population.

Retrospective analysis of all patients in the HAART Observational Medical Evaluation and Research (HOMER) cohort who were 19 years of age and older who started treatment between August 1996 and June 2012. The primary outcome variable was death due to suicide. Data on deaths were obtained monthly through a linkage with the British Columbia Ministry of Health Vital Statistics Agency. Logistic regression and Cox proportional hazards models were used to identify factors independently associated with suicide mortality.

A total of 993 deaths among 5229 patients accessing treatment were recorded, of which 82 (8.2%) were caused by suicide. Death from suicide peaked at 961 deaths per 100 000 person-years in 1998 and declined to 2.81 deaths per 100 000 person-years in 2010. Cox regression analysis showed that a history of injection drug use (adjusted hazard ratio [AHR] = 3.95, 95% confidence interval [CI] 1.99–7.86) or having no experience with an AIDS-defining illness (AHR = 4.45, 95% CI 1.62–12.25) were factors independently associated with suicide. This model showed a 51% reduction (AHR = 0.49, 95% CI 0.45–0.54) in the suicide rate per calendar year.

Deaths from suicide declined substantially over time, and factors other than progression of HIV disease, such as injection drug use, may be important targets for intervention to reduce suicide risk.

Below: Comparison of suicide rates in the HOMER cohort with suicide rates in the general population of British Columbia and as a proportion of all-cause mortality from 1997 to 2011. To ensure full-year comparisons, data starting from 1997 to 2011 were used in the graph. HOMER = HAART Observational Medical Evaluation and Research.