Treatment (as Prevention) Availability and Individuals' Behavior: A Cost-Effectiveness Analysis of Cabotegravir Long-Acting Injectable PrEP.

Background: The US Food and Drug Administration (FDA) has approved long-acting injectable agents for preexposure prophylaxis (PrEP) of human immunodeficiency virus (HIV) infection. These agents are urgently required to enhance both the available options and the efficacy in HIV prevention. Long-acting PrEP treatments offer a less frequent and more discreet alternative, potentially preferred by many individuals. These treatments address adherence issues commonly associated with traditional PrEP regimens and can provide increased effectiveness. Consequently, these benefits might alter individuals' risk perceptions, potentially influencing their behaviors, such as changes in condom use and the number of sexual partners. Such behavioral shifts can affect the treatment's overall value, encompassing its costs and benefits.

Objective: This paper aims to model this endogenous behavior concerning the availability of long-acting PrEP. Using this model, we then conduct a cost-effectiveness analysis comparing long-acting PrEP (cabotegravir) with the standard daily oral pill treatment (tenofovir disoproxil fumarate/emtricitabine [TDF/FTC], branded as Truvada).

Methods: We developed a conceptual framework using evolutionary game theory to model endogenous behavioral responses to long-acting PrEP among US-based individuals at an elevated risk for HIV, specifically cisgender men who have sex with men and transgender women. We defined two theoretical strategies-protective (P) and nonprotective (NP)-reflecting distinct behavioral responses in terms of condom use and number of sexual partners. These two behavioral groups were integrated as separate populations in a Markov health-state transition model. While the Markov model remains consistent across both populations, certain parameters vary to reflect the behavioral impacts on economic and health outcomes.

Results: The endogenous model offers a stylized representation of how behavioral responses to long-acting PrEP may shape population distribution across two behavioral strategy types. Notably, if the entire population is protective, the long-acting PrEP is less likely to be cost-effective at US$109,769 per quality-adjusted life year (QALY) gained. In contrast, if everyone is nonprotective, the treatment is highly cost-effective at US$54,536 per QALY gained. Using a conservative estimate, where the population is evenly split between protective and nonprotective, the treatment is deemed cost-effective at US$85,152 per QALY gained. It is observed that the incremental costs of the long-acting PrEP, when compared with traditional PrEP, are marginally higher for the protective group than the nonprotective group. This is attributed to the lower mortality in the protective subgroup for both PrEP types, which results in higher overall costs due to increased survival rates. Despite a smaller population size at the end of the simulation, the nonprotective group exhibits greater gains in QALYs, stemming from the combined effects of their riskier behaviors and the high efficacy of the long-acting PrEP.

Conclusions: Our findings suggest that, even when considering potential shifts toward riskier behaviors, introducing long-acting PrEP remains cost-effective when compared with daily oral PrEP treatments.