The epidemiology of recent hepatitis C infection among persons who inject drugs

Background Globally, 71 million people ate infected with hepatitis C virus (HCV) and there are approximately 1.75 million new infections per year. The World Health Organization (WHO) has announced the goal of eliminating viral hepatitis as a public health problem by 2030, including a reduction in HCV incidence by 90%. Accurate HCV incidence estimates will be critical for setting baselines for these goals as well as for monitoring and evaluation of interventions and progress. Traditional methods to estimate HCV incidence rely on longitudinal follow-up of high-risk populations. These studies are logistically difficult, resource intensive, and the study participants enrolled in such studies are subject to biases and may not reflect the underlying population at risk. Objectives Our study aims to evaluate methods to estimate incident HCV infections in persons who inject drugs (PWID) using data derived from cross-sectional studies and to identify factors associated with recent infection to better understand the epidemiology of HCV transmission among this high-risk population. Methods Using 1,840 samples collected from 875 individuals enrolled in five prospective cohort studies of persons who inject drugs (PWID) in the US and India we evaluated the utility of a modified Genedia avidity assay coupled with viremic confirmation to estimate HCV incidence in cross-sectional serosurveys. We evaluated and modified the utility of the Osmond method (MOA), previously used to estimate cross-sectional HIV incidence in young men who have sex with men, to estimate HIV incidence among PWID using cross-sectional data collected from over 14,000 PWID using respondent driven sampling (RDS) from 15 sites across India enrolled in the National Collaboration on AIDS (NCA) study. MOA estimates of HIV incidence were compared with HIV incidence estimates from a previously validated multiassay approach (MAA) and applied what we learned from this HIV application to estimate HCV incidence in this population using the MOA. We used multilevel logisitic regression to identify factors associated with recent infection among 1,640 PWID enrolled in the NCA study who reported injection for ≤ 1 year. Results Using an avidity index cutoff of <40%, this Genedia avidity based approach had an overall mean duration of recent infection (MDRI) of 113 days (95%CI: 84-146), and a false recent rate (FRR) of 0.4% (95%CI: 0.0-1.2) and 4.6% (95%CI: 2.2-8.3) among HIV negative and positive individuals, respectively, and did not differ by HCV genotype. Further, the calculated sample sizes needed to estimate incidence in high burden settings were