RESUMEN
INTRODUCTION: HIV prevalence among state prison inmates in the United States is more than five times higher than among nonincarcerated persons, but HIV transmission within U.S. prisons is sparsely documented. We investigated 88 HIV seroconversions reported from 1988-2005 among male Georgia prison inmates. METHODS: We analyzed medical and administrative data to describe seroconverters' HIV testing histories and performed a case-crossover analysis of their risks before and after HIV diagnosis. We sequenced the gag, env, and pol genes of seroconverters' HIV strains to identify genetically-related HIV transmission clusters and antiretroviral resistance. We combined risk, genetic, and administrative data to describe prison HIV transmission networks. RESULTS: Forty-one (47%) seroconverters were diagnosed with HIV from July 2003-June 2005 when voluntary annual testing was offered. Seroconverters were less likely to report sex (OR [odds ratio] = 0.02, 95% CI [confidence interval]: 0-0.10) and tattooing (OR = 0.03, 95% CI: <0.01-0.20) in prison after their HIV diagnosis than before. Of 67 seroconverters' specimens tested, 33 (49%) fell into one of 10 genetically-related clusters; of these, 25 (76%) reported sex in prison before their HIV diagnosis. The HIV strains of 8 (61%) of 13 antiretroviral-naïve and 21 (40%) of 52 antiretroviral-treated seroconverters were antiretroviral-resistant. DISCUSSION: Half of all HIV seroconversions were identified when routine voluntary testing was offered, and seroconverters reduced their risks following their diagnosis. Most genetically-related seroconverters reported sex in prison, suggesting HIV transmission through sexual networks. Resistance testing before initiating antiretroviral therapy is important for newly-diagnosed inmates.