LAUSR

PURPOSE Residing in areas with little spatial accessibility to HIV pre-exposure prophylaxis (PrEP) providers, or PrEP deserts, contributes to low PrEP uptake. This study examines and characterizes the spatial distribution of PrEP accessibility in the United States over time. METHODS We conducted spatial network analyses and geographic mapping to explore the spatiotemporal distribution of persistent PrEP deserts (census tracts with suboptimal accessibility in 2016 and 2020), new PrEP deserts (tracts with suboptimal accessibility in 2020 but not 2016), new PrEP oases (tracts with suboptimal accessibility in 2016 but not 2020), and persistent PrEP oases (tracts with optimal accessibility in 2016 and 2020). We used polytomous logistic regression to determine area-level factors associated with these four spatiotemporal PrEP accessibility types. RESULTS There was a reduction of 52.8% in the prevalence of 30-minute PrEP deserts from 2016 (28,055 tracts) to 2020 (13,240 tracts) and an increase of 33.5% in 30-minute PrEP oases from 2016 (44,259 tracts) to 2020 (59,074 tracts). Of all tracts, 12,487 (17.3%) were persistent PrEP deserts, 753 (1.0%) were new PrEP deserts, 15,568 (21.5%) were new PrEP oases, and 43,506 (60.1%) were persistent PrEP oases. Overall, persistent PrEP oases were more likely to be of higher socioeconomic status, racially/ethnically diverse, located in urban areas, and located in the Northeast compared with other spatiotemporal PrEP accessibility types, with variation by urbanicity and U.S. Census region. CONCLUSIONS Efforts to improve PrEP accessibility should be especially focused in disadvantaged communities in nonurban areas and the South, Midwest, and West. Monitoring changes in the spatial accessibility of PrEP over time and determining the factors associated with such changes can help to evaluate progress made towards improving PrEP accessibility.