The Lancet, February 2007, vol.369, iss.9562, pp.643-656. Available From:Link to Source
This study rigorously evaluates the effectiveness of male circumcision as a strategy to prevent human immunodeficiency virus (HIV) infection in Kenya. Focusing on young men aged 18 to 24 years residing in the large western city of Kisumu, the researchers randomly assign half of the 2,784 participants into the intervention group (circumcision) and the remaining into the control group (delayed circumcision). Both groups are then followed for 24 months, with multiple rounds of HIV testing, medical examinations and behavioural interviews.
Theoretically, male circumcision can lead to a reduction in the risk for HIV infection, because the inner surface of the human foreskin has a nine times greater density of HIV target cells compared with cervical tissue. The present study, together with similar studies in Uganda and South Africa, helps quantify the magnitude of these prevention effects. The experimental subjects are recruited from an area where about 10 per cent of adult men are circumcised at baseline. It is an area of high HIV risk, with prevalence rates of 25 per cent amongst women and 18 per cent amongst men in 2003.
The statistical analysis focuses on comparing rates of HIV incidence across treatment groups. In addition to reporting results for the whole study period, the authors disaggregate the findings by examining four 6-month periods, corresponding to months 0 to 6, 6 to 12, 12 to 18 and 18 to 24 of the study respectively. For each period, the authors report average rates of HIV infection for each group, as well as their 95 per cent confidence intervals. To investigate possible channels through which the intervention may have affected the risk for acquiring HIV, the authors also examine reported sexual practices such as condom use and multiple partners. Finally, they report the total number and severity of injuries caused by the surgery.
The key outcome of the study is that male circumcision is very effective at reducing HIV infection risk in young men. For the study period as a whole, the treatment group had an estimated HIV incidence rate of 2.1 per cent (95 per cent confidence interval = 1.2 per cent to 3.0 per cent), compared with 4.2 per cent for the control group (95 per cent confidence interval = 3.0 per cent to 5.4 per cent). In other words, circumcision cut the rate of new infections by half. This difference is statistically significant at the 1 per cent level (p = .0065). When this overall effect is disaggregated by looking at separate time periods, the circumcision group had lower rates of infection throughout (0.8 per cent vs. 1.0 per cent during months 0 to 6, 0.2 per cent vs. 1.4 per cent during months 6 to 12, 0.0 per cent vs. 0.7 per cent during months 12 to 18, and 1.0 per cent vs. 1.2 per cent during months 18 to 24).
In terms of reported sexual behaviour, there is evidence of some small effects. For instance, 22 per cent of treatment and 21 per cent of control individuals reported consistent condom use in the previous 6 months at baseline. At 24-month follow-up, however, the corresponding numbers are 36 per cent and 41 per cent. This effect has an associated p of .0326 and is therefore statistically significant at the 5 per cent level. In terms of the incidence of multiple sexual partners, both treatment and control individuals show significant declines after study enrolment, with those in the circumcision group less likely to report its presence than those in the control group (29 per cent vs. 33 per cent at 12-month follow-up). Finally, there was a very small number of mild surgery-related complications, suggesting that male circumcision is indeed a powerful tool in the fight against HIV.