Identifying preferred program delivery attributes for long-acting injectable cabotegravir-rilpvirine in three U.S. HIV clinics

Abstract

BACKGROUND: Long-acting injectable cabotegravir and rilpivirine (CAB/RPV-LA) may overcome adherence barriers and provide treatment choice among people with HIV (PWH). However, little is known about the factors most likely to facilitate CAB/RPV-LA use.

METHODS: We conducted a discrete choice experiment (DCE) at three urban U.S. HIV clinics, stratified by care engagement status (sub-optimal care engagement defined as most recent HIV RNA ≥200 copies/mL or <2 primary care visits with no HIV RNA measurement in past year). The DCE had 6 attributes: visit location, conversation with one's HIV provider at injection, extra support services, visit length, extended hours, and cost. We used mixed logit regression to estimate preference weights and latent class analysis to identify preference groupings. We examined associations between covariates, attribute preferences and latent class groups, and estimated shares of preference towards hypothetical scenarios.

RESULTS: Between December 2021 - May 2022, 370 PWH participated. Median age was 46, 34% were cisgender female/gender minority, 59% African American/Black, 13% Latino/a/x, 34% experienced unstable housing, 19% endorsed illicit substance use, and 27% sub-optimal engagement. Cost and visit location were the most important attributes. Latent class analysis identified three groups: 1) cost-conscious (44%), 2) location-driven (34%), and 3) provider/support services-oriented (22%). Sub-optimally engaged PWH preferred extra support services. Shares of preference suggested a predicted use of CAB/RPV-LA exceeding 80% across delivery scenarios with more preferred features.

CONCLUSION: Participants strongly preferred CAB/RPV-LA delivery at no cost and accessible at their current HIV clinic. Preferences of PWH with sub-optimal engagement were heterogeneous, suggesting the need for tailored, differentiated delivery programs offering choice.