COST-EFFECTIVENESS OF HIV PREVENTION PORTFOLIOS FOR PEOPLE WHO INJECT DRUGS IN THE UNITED STATES

Purpose: HIV pre-exposure prophylaxis (PrEP) for people who inject drugs (PWID) in the US remains a controversial intervention. A recent analysis demonstrated that PrEP can reduce national HIV incidence but would pose substantial budget challenges to do so. To identify the highest-value portfolio of HIV prevention programs, we evaluate the cost-effectiveness of PrEP in the context of three other PWID-targeted interventions: opioid agonist therapy (OAT), needle and syringe exchange programs (NSP), and enhanced testing and antiretroviral treatment (ART).

Methods: We adapted an empirically calibrated dynamic model of the US HIV epidemic and used it to simulate transmission between PWID, men who have sex with men, and lower-risk heterosexuals. Intervention portfolios were defined by coverage levels in the PWID population for each of the four prevention programs: at their status quo level or else at low (40% of the eligible population), medium (60%), or high (80%) coverage expansion. For each portfolio, we averaged discounted lifetime costs and QALYs for a 20-year analytic time horizon over 182 calibrated input parameter sets, and we estimated incremental cost-effectiveness ratios.

Results: We first considered expansion of each prevention program with all others at status quo levels. Considered singly, OAT, ART, and NSP expanded up to high coverage levels all achieved costs per QALY gained below $50,000. PrEP cost more than $300,000 per QALY gained even at low coverage expansion. Over 20 years, low coverage of PrEP averts 5,000 more infections than low NSP, which averts 7,000 more infections than low OAT, a program that, unlike PrEP, is not targeted solely to uninfected PWID. Thus, investing in OAT alone, while the most cost-effective choice, may not directly prevent as many new infections. This provides a motivation for combining programs. Figure 1 shows the efficient frontier when we consider all combinations of the four programs and reveals how to build the highest-value prevention portfolio: maximize OAT coverage, then maximize ART coverage, then maximize NSP coverage.

Conclusions: A cost-effective prevention program portfolio for PWID in the US should maximize OAT coverage and include combinations with ART and NSP. Although PrEP for PWID has the potential to avert many HIV infections, at current prices, it has the least favorable cost-effectiveness of the evaluated interventions.