Model-based cost-effectiveness estimates of testing strategies for diagnosing hepatitis C virus infection in people who use injecting drugs in Senegal.

ABSTRACT

BACKGROUND: Scaling-up the access to hepatitis C virus (HCV) diagnostics for people who use injecting drugs (PWID) is essential to reduce the HCV incidence in low and middle-income countries. METHODS: A decision tree model was developed to compare the cost-effectiveness of 12 strategies for diagnosing HCV in Senegal with a health sector perspective. Strategies included HCV-Ab screening and confirmation of viraemia (based on HCV-RNA or HCV core antigen detection) or only the latter step. Laboratory assays and decentralized tools (point-of-care (POC) tests and dried blood spot (DBS) samples) were included. The base-case assumed a 38.9% seroprevalence, as reported in the PWID population of Dakar. RESULTS: Compared to the cheapest strategy (POC HCV-Ab followed by POC HCV-RNA (S5)), one strategy remained un-dominated in the base-case POC HCV-Ab followed by venepuncture-based laboratory HCV-RNA (S3). Above a lost to follow-up testing rate of 2.3%, combining POC HCV-Ab with HCV-RNA on DBS (S4) became more cost-effective than S3. Above this threshold, a single-step POC HCV-RNA (S12) was also found un-dominated (ICER to S5=€3,297.50). S5, S12 and S4 cost €14.21, €17.03 and €36.55/screened individual. Incremental cost-effectiveness ratios (€/additional true positive case) were 2,164.82 (S12 versus S5) and 3,297.50 (S4 versus S12). Whenever HCV seroprevalence reached 55.5%, S12 became more cost-effective than S5. Moreover, S4 required a budget 2 to 2.5 times higher than S5 or S12 for diagnosing 90% of HCV-infected PWID in Dakar. CONCLUSION: A two-step POC-based strategy (S5) would be the most cost-effective option among those proposed in this study for diagnosing HCV in PWID in Senegal. This study illustrates how the lack of secure financing and of data on PWID in LMICs, render difficult to identify the most sustainable strategy in those countries, as well as its implementation.