Aboriginals are over-represented in Canada's HIV epidemic and are commonly infected with HIV via injection drug use (IDU); however, little is known about the impact of Aboriginal ethnicity on mortality after starting highly active antiretroviral therapy (HAART). Therefore, we compared mortality rates between Aboriginal and non-Aboriginal HIV patients and between IDU and non-IDU HIV patients after they initiated HAART.
We conducted a retrospective cohort study of antiretroviral-naïve patients starting HAART January 1999-June 2005 (baseline), followed until December 2005. We constructed two Cox proportional hazards models, one to estimate all-cause and one to estimate HIV-related mortality hazard ratios (HRs), considering sex, and baseline age, CD4 cell count, HIV RNA level, calendar year, and HAART regimen as potential confounders.
The 548 study patients were followed for 1,889.8 person-years; 194 (35%) were Aboriginal, 255 (46%) were IDUs. We observed 55 deaths; 47% were HIV-related. In multivariable models, Aboriginals experienced higher all-cause (HR = 1.85, 95% CI = 1.05-3.26, p = 0.034) and HIV-related (HR = 3.47, 95% CI = 1.36-8.83, p = 0.009) mortality rates compared to non-Aboriginals; and, compared to patients with other exposures, IDUs experienced higher all-cause (HR = 2.45, 95% CI = 1.31-4.57, p = 0.005) but similar HIV-related (p = 0.27) mortality rates.
Compared to non-Aboriginals, Aboriginal HIV patients suffer higher all-cause and HIV-related mortality rates after starting HAART. The strongest and most significant predictor of higher all-cause mortality was IDU. Future research should examine reasons for the observed poorer survival of Aboriginal and IDU HIV patients after initiating HAART to develop interventions to improve the prognosis for these vulnerable populations.