Association for Public Policy Analysis & Management

Background A number of clinical randomized controlled trials have shown that medical male circumcision reduces HIV-1 infections. Studies have shown limited evidence of risk compensation following male circumcision but are generally based on self-reported data in short term follow-ups. The aim of the study is to assess the long-term impacts of medical male circumcision on risky sexual behaviors as well as HIV-1 and HSV-2 infection.

Methods We implemented a clustered randomized controlled trial of a male circumcision intervention with 4 years of follow-up. 2,663 male students enrolled in 33 secondary schools near Lilongwe, Malawi were randomly assigned to an intensive transportation intervention group or a control group that received a less intensive and delayed transportation intervention. For both transportation support interventions, students could choose either a direct pick-up service or a transportation voucher that could be redeemed at the assigned clinic performing free circumcisions. For the intensive intervention, the direct pick-up service was offered six-times and the transportation voucher was offered twice over a period of 18 months. For the less intensive intervention, the direct pick-up service and the transportation vouchers were offered once over a period of 6 months. The primary outcomes are HIV-1 and HSV-2 infections four years after the offer and a series of self-reported measures of sexual behavior four years after the offer. Analysis was intention to treat.

Findings Between October 2011 and May 2012, eligible male participants were assigned to the intensive intervention group (n=1,342) and to the less intensive and delayed control group (n=851). The male circumcision take-up rate of the intervention group is higher than that of the control group after two years (adjusted OR 5.72, 95% CI 3.14-10.44), and the difference was attenuated but sustained after four years (adjusted OR 2.48, 95% CI 1.52-4.05). We find that the intervention group is more likely to be infected with HSV-2 than the control group four years after the offer (adjusted OR 1.49, 95% CI 1.08-2.07). However, we do not find significant differences in the prevalence of HIV-1. In addition, we find a significant changes in sexual behavior among those in the intervention group compared to the control group, especially in attitude toward a use of condom.

Interpretation Using a community-based trial in male circumcision scale-up project for male secondary school students with the longest follow-up to date and a comprehensive set of outcomes, we provide evidence that male circumcision led to increased HIV-1 and HSV-2 risk behavior. Our results suggest that as male circumcision programs continue to be scaled-up, the monitoring of risk compensation remains important as are efforts to supplement these programs with educational and public health campaigns aimed at reducing risky sexual behavior.