Comparing geographic accessibility of hiv testing and hiv burden among adolescents and young adults in tampa bay, florida

Florida adolescents and young adults (AYA, ages 13-24) experience rates of new HIV diagnosis at nearly double the national average, with Tampa Bay among the regions with highest transmission. Understanding the geo-social factors that contribute to this disparity is critical to HIV prevention efforts. “Testing deserts” have been used to describe how limited geographic access to testing impacts disease burden. Applying this concept to HIV, we sought to determine the association between zip code-level testing site accessibility and AYA HIV burden to identify potential local testing deserts. Because adults have traditionally been prioritized in HIV outreach, we also sought to determine this association using adult (ages 25-44) HIV burden. We hypothesized that test site distribution would more closely follow trends in adult, rather than AYA, HIV burden. Study data were drawn from AIDSVu, an open-access online mapping tool that includes HIV surveillance and testing site data by zip code. HIV burden for AYA and separately for adults was defined using three measures: new HIV diagnoses (per 100K), prevalence (per 100K), and percent virally suppressed. We characterized the geographic accessibility of testing sites by zip code by using geocoding software (ArcGIS Online) to create a 15-minute walking distance radius around each test site and calculate the percent surface area per zip code within walking distance to at least one site. We used walking distance to limit transportation bias. We then mapped testing site accessibility and HIV burden to visually identify testing deserts. Pearson’s correlation coefficients were calculated to quantify and compare the association of test site accessibility to HIV burden for both AYA and adults. Of the 97 Tampa Bay zip codes analyzed, 36.1% had 0% surface area within walking distance to at least one test site. Mapping revealed that most test sites were concentrated around regions with the highest adult, rather than AYA, new diagnosis and prevalence rates, a trend that was confirmed quantitatively with r=0.51 for the relationship between test site availability and adult prevalence rate vs. r=0.09 for AYA. With this disparity in mind, we identified 4 new potential AYA testing deserts that have high AYA new HIV diagnosis and prevalence rates but few local test sites. In contrast to trends in the association between testing availability and new diagnosis and prevalence rates, we found that zip codes with greater access to testing tended to have greater AYA percent viral suppression. Quantification of this trend showed a modest, positive, statistically significant correlation (r=0.47). Our analysis indicates an overall paucity of HIV testing sites in regions with high AYA new diagnosis and prevalence rates. The current geographic distribution of test sites follows HIV diagnosis and prevalence rates among adults—a demographic that has remained a long-standing focus of HIV prevention—leaving AYA vulnerable across several testing deserts. These newly identified testing deserts represent opportunities to enhance access to HIV testing sites for Tampa Bay AYA, which in turn may be associated with improvement in other measures of AYA HIV burden, such as percent viral suppression.