Preference heterogeneity for HIV pre-exposure prophylaxis care among gay, bisexual, and other men who have sex with men in the United States: a large discrete choice experiment

BackgroundPrEP uptake among Black and Latino gay, bisexual, and other men who have sex with men (GBM) remains low in the United States. The design and implementation of PrEP delivery programs that incorporate the preferences of Black and Latino GBM may overcome barriers to uptake. We aimed to identify preferences for PrEP care among high-priority GBM in the U.S. with a large discrete choice experiment. MethodsWe conducted two discreet choice experiments (DCE) to elicit care preferences for (1) Starting PrEP and (2) Continuing PrEP care among GBM clinically indicated for PrEP. The DCE web-based survey was nested in a longitudinal cohort study of GBM in the U.S., implemented with video and audio directions among 16-49 year-old participants, not using PrEP, and verified to be HIV-negative. All participants were presented with 16 choice sets, with choices determined by BLGBM and PrEP implementation stakeholders. We calculated overall utility scores and relative importance and used latent class analyses (LCA) to identify classes within the Starting and Continuing PrEP DCE. Multivariable analysis was performed to identify factors associated with class membership. FindingsAmong 1514 participants, mean age was 32 years; 46{middle dot}5% identified as Latino, 21{middle dot}4% Black, and 25{middle dot}2 White; 37{middle dot}5% had an income less than USD $20,000. Two latent classes were identified for Starting PrEP: Class 1 (n=431 [28{middle dot}5%]) was driven by preference for more traditional in-person care, and Class 2 (n=1083 [71{middle dot}5%]) preferred flexible care options and on-demand PrEP. In a multivariable model, having a sexual health doctor (adjusted OR 0{middle dot}7, CI 0{middle dot}5, 0{middle dot}9), having a primary care provider (OR 0{middle dot}7,CI 0{middle dot}5, 0{middle dot}9, p= 0{middle dot}023), and concerns over PrEP side effects (OR 1{middle dot}1, CI 1{middle dot}0,1{middle dot}2, p= 0{middle dot}003) were all associated with class membership. InterpretationThe different preferences identified for PrEP care indicate the need for diverse care and formulation choices to improve PrEP uptake and persistence. Addressing these preferences and understanding the factors that shape them can inform the implementation of programs that increase PrEP uptake.