World Health Organization (WHO) and Joint United Nations Program on AIDS (UNAIDS) recommend male circumcision (MC) as an additional HIV prevention measure. This study aimed to assess three models of promoting MC and their effects on preventing HIV infection among drug users in western China. We carried out a cohort study in three provinces of western China. HIV seronegative male drug users were recruited from methadone maintenance therapy clinics and cluster randomized into three intervention models. At baseline, 6, and 9 months of follow-up, changes in MC knowledge, the acceptability of MC, MC surgery uptake, and the costs of model implementation were analyzed. Of 1,304 male drug users who were screened, 1,218 were enrolled in the study. The participants' knowledge about MC was significantly increased after interventions by all three models. The one-stage model led to the highest increase in MC acceptability and the greatest increase in MC uptake. Multivariable Cox regression analysis showed that the one-stage model was also the most effective method to promote MC uptake, compared with the two-stage model [rate ratio (RR) = 0.602; 95% confidence interval (CI), 0.420-0.862] and three-stage model (RR = 0.555; 95% CI, 0.382-0.807). The HIV incidence rate in the MC group was lower than that in the non-MC group (RR = 0.234; 95% CI, 0.056-0.974). Moreover, the one-stage model required the lowest cost per circumcision. The one-stage model is the most effective and the most cost-effective intervention to increase MC uptake among male drug users in western China and could decrease the HIV incidence rate, based on a short follow-up investigation.
Keywords: AIDS; China; Guangxi; HIV; circumcision; drug user.