Background & aims: Previous studies reported increased morbidity and mortality related to liver disease among human immunodeficiency virus (HIV)-infected patients with hepatitis C co-infection. However, the long-term effect of hepatitis C virus (HCV) co-infection on the mortality of HIV-infected patients remains unclear.
Methods: By using national Veterans Affairs (VA) databases, we performed a retrospective cohort study of HIV patients hospitalized between October 1991 and September 2000. Mortality rates and hazard rate ratios (HRRs) for mortality were calculated for the entire cohort as well as after excluding patients with pre-existing liver disease, with follow-up through September 2001 after discharge. Multivariable adjustment for differences in demographics, comorbidities, and HIV disease severity was performed. Separate analyses were performed for patients identified during the highly active antiretroviral therapy (HAART) era.
Results: We identified 18,081 patients, of whom 5320 patients had dual HCV/HIV infection and 12,761 patients had HIV monoinfection. The number of deaths per 100 patient-years was 7.33 in the dual infection group and 14.13 in the HIV monoinfection group during 22,054 and 40,655 person-years of follow-up, respectively. The mortality rate ratio between HCV/HIV dual infection and HIV monoinfection was .53. In Cox multiple regression, the dual HCV/HIV infection group had an adjusted HRR for mortality of .55 compared with the HIV monoinfection group (95% CI, .51-.58, P < .0001), after controlling for age, race, sex, year of diagnosis, and HIV disease severity. These findings persisted in several sensitivity analyses. However, in the HAART era, if patients with liver disease at baseline were excluded, the HRR for mortality was .83 (95% CI, .73-.94, P = .003).
Conclusions: Co-infection with hepatitis C is associated with a significant decrease in the mortality of HIV-infected patients. However, this effect was less pronounced during the HAART era.