Abstract
To improve current and future use of existing (oral, injectable) and potential future (implants, douches) pre-exposure prophylaxis (PrEP) products, we must understand product preferences relative to one another, among gay and bisexual men (GBM), and physicians who prescribe PrEP. We completed an online discrete choice experiment (DCE) with separate groups of GBM and/or physicians from the United States, South Africa, Spain, and Thailand. Participants were presented information on PrEP products, including daily pills, event-driven pills (2-1-1 regimen), injections, subdermal implants (dissolvable, removable), and rectal douches. Next, they completed a choice exercise in which they were shown 10 screens, each presenting 3 of the aforementioned products at a time with 11 attributes for physicians and 10 attributes for GBM. For the attributes that were not constant, one level was shown per screen for each product. Participants selected the product they preferred most and rated their likelihood to select (GBM) or recommend (physicians) that product. Data were modeled using hierarchical Bayes estimation; resulting model coefficients were used to develop attribute importance measures and product preferences. For GBM across all countries, if all aforementioned PrEP products were on the market at the same time, over 90% of GBM would use some form of PrEP; 100% of physicians would recommend at least one of the PrEP products. There were variations in product preference by country. GBM in the United States and Thailand preferred the injection (21.7%, 22.9%, respectively), while the dissolvable implant was preferred in South Africa and Spain (19.9%, 19.8%, respectively). In the United States, South Africa, and Spain (where physician data were available), physicians were most likely to recommend the dissolvable implant (37.2%, 40.6%, 38.3%, respectively).