Over 20 years, interventions have targeted HIV among female sex workers (FSWs) in Kenya given their central role in new infections. To determine the effects of these interventions, we assessed the prevalence and correlates of HIV among a random sample of FSWs and modelled prevalence estimates from studies since 1993. FSWs aged 16-34 years were enrolled through multi-stage sampling. Regression models identified correlates of HIV infection. Generalised linear mixed modelling estimated temporal changes in prevalence between 1993 and 2016. 882 FSWs were enrolled. Prevalence rose from 3.6% among 16-20-year-olds to 31.6% among 31-34-year-old FSWs. Those aged 31 to 34 years had greater odds of HIV compared to those 16 to 20 years (AOR 14.2, 95% CI, 5.5-36.8). Infection was less prevalent among FSWs with tertiary education compared to those with primary or no education (AOR 0.23, 95% CI, 0.07-0.78). There was an overall 30% reduction in prevalence from 1993 to 2016 with an average annual decline of 3%. About one in ten FSWs in Mombasa are currently infected with HIV. Considering FSWs' central role in sustaining population-level infections, these initiatives require continued support, focusing on reducing transmission from older FSWs and those with less education.
Keywords: HIV; Kenya; complex sampling; female sex workers; key populations; prevalence.