An ecological analysis to examine stigma and undiagnosed HIV infections among men who have sex with men (MSM) in the United States, National HIV Behavioral Surveillance System, 2017

Abstract Background: Studies have been conducted to assess the prevalence of undiagnosed HIV infections however, at present, there is no study that compares and evaluates the association of perceived HIV stigma, gay related enacted stigma, and undiagnosed HIV infections among MSM by geographic region in the United States (U.S). This ecological study identifies clusters of cities with similar stigma measures in the U.S. Methods: We conducted the analysis using data from National HIV Behavioral Surveillance (NHBS) collected during the 2017 MSM cycle. We used unsupervised machine learning procedure known as k-means clustering to identify clusters of cities with similar stigma measures. We also examined correlations between percentages of undiagnosed HIV infections both with perceived HIV stigma and gay related enacted stigma. In addition, we used American Community Survey (ACS), 2013-2017, data to characterize each city in terms of key population level characteristics. Results: Cities were grouped in two distinct clusters after the implementation of the clustering technique; each cluster was different in terms of their racial makeup. There was a moderate correlation (0.50) between percentage of perceived HIV stigma and undiagnosed HIV infections and a low correlation (0.23) between percentage of gay related enacted stigma and undiagnosed HIV infections among MSM. Conclusion: Stigma reduction interventions should be tailored by city to address the underlying HIV burden among MSM communities. This will not only improve the health of the MSM communities but the larger community as a whole. Future studies should include more cities from the Southern U.S and examine how stigma affects MSM populations, especially of color.