Predictors of liver-related death among people who inject drugs in Vancouver, Canada: a 15-year prospective cohort study

INTRODUCTION While HIV/AIDS remains an important cause of death among people who inject drugs (PWID), the potential mortality burden attributable to hepatitis C virus (HCV) infection among this population is of increasing concern. Therefore, we sought to identify trends in and predictors of liver-related mortality among PWID. METHODS Data were derived from prospective cohorts of PWID in Vancouver, Canada, between 1996 and 2011. Cohort data were linked to the provincial vital statistics database to ascertain mortality rates and causes of death. Multivariate Cox proportional hazards regression was used to examine the relationship between HCV infection and time to liver-related death. A sub-analysis examined the effect of HIV/HCV co-infection. RESULTS AND DISCUSSION In total, 2,279 PWID participated in this study, with 1,921 (84.3%) having seroconverted to anti-HCV prior to baseline assessments and 124 (5.4%) during follow-up. The liver-related mortality rate was 2.1 (95% confidence interval [CI]: 1.5-3.0) deaths per 1,000 person-years and was stable over time. In multivariate analyses, HCV seropositivity was not significantly associated with liver-related mortality (adjusted relative hazard [ARH]: 0.45; 95% CI: 0.15-1.37), but HIV seropositivity was (ARH: 2.67; 95% CI: 1.27-5.63). In sub-analysis, HIV/HCV co-infection had a 2.53 (95% CI: 1.18-5.46) times hazard of liver-related death compared with HCV mono-infection. CONCLUSIONS In this study, HCV seropositivity did not predict liver-related mortality while HIV seropositivity did. The findings highlight the critical role of HIV mono- and co-infection rather than HCV infection in contributing to liver-related mortality among PWID in this setting.