Can hepatitis C virus prevalence be used as a measure of injection-related human immunodeficiency virus risk in populations of injecting drug users? An ecological analysis


Peter Vickerman, London School of Hygiene and Tropical Medicine, Keppel Street, London, WC1E 7HT, UK.


Background  Human immunodeficiency virus (HIV) outbreaks occur among injecting drug users (IDUs), but where HIV is low insight is required into the future risk of increased transmission. The relationship between hepatitis C virus (HCV) and HIV prevalence among IDUs is explored to determine whether HCV prevalence could indicate HIV risk.

Methods  Systematic review of IDU HIV/HCV prevalence data and regression analysis using weighted prevalence estimates and time–series data.

Results  HIV/HCV prevalence estimates were obtained for 343 regions. In regions other than South America/sub-Saharan Africa (SAm/SSA), mean IDU HIV prevalence is likely to be negligible if HCV prevalence is <30% (95% confidence interval 22–38%) but increases progressively with HCV prevalence thereafter [linearly (β = 0.39 and R2 = 0.67) or in proportion to cubed HCV prevalence (β = 0.40 and R2 = 0.67)]. In SAm/SSA, limited data suggest that mean HIV prevalence is proportional to HCV prevalence (β = 0.84, R2 = 0.99), but will be much greater than in non-SAm/SSA settings with no threshold HCV prevalence that corresponds to low HIV risk. At low HCV prevalences (<50%), time–series data suggest that any change in HIV prevalence over time is likely to be much smaller (<25%) than the change in HCV prevalence over the same time-period, but that this difference diminishes at higher HCV prevalences.

Conclusions  HCV prevalence could be an indicator of HIV risk among IDUs. In most settings, reducing HCV prevalence below a threshold (30%) would reduce substantially any HIV risk, and could provide a target for HIV prevention.