摘要：Epidemiological theory and many empirical studies support the hypothesis that there is a protective effect of medical male circumcision (MMC) against HIV infection. We aim to test this claim with evidence from a high endemic community in South Africa. Methodology: We analysed a cross-sectional dataset, including HIV test outcome and MMC status of males aged 15–49, who participated in the HIV Incidence Provincial Surveillance System study between June 2014 and June 2015. We applied two propensity score methods, inverse probability of treatment weighting (IPTW) and full matching, to approximate a randomised trial investigating the hypothesis that MMC protects against HIV infection. We benchmarked our results using the conventional regression adjustment. Models were adjusted for identified confounders. Results: The adjusted odds ratios (OR) provided evidence of a protective effect of MMC against HIV acquisition: IPTW (OR = 0.623, 95% CI: 0.467 – 0.830); Full matching (OR = 0.607, 95% CI: 0.440 – 0.837); Regression (OR = 0.579, 95% CI: 0.426 – 0.789). Conclusion: There is a need to create more programmes that emphasise MMC as an important HIV preventive intervention.
关键词：HIV Medical male circumcision Propensity score Kwazulu-Natal South Africa
