The aim of this review was to evaluate the scientific evidence supporting the hypothesis that male circumcision reduces the risk of HIV infection and consequently the incidence of acquired immunodeficiency syndrome (AIDS).
Patients and Methods
We performed a literature search of the major databases (Medline, Embase, Cochrane Library, Biosis and Science Citation Index) for papers published in the period 1999 to 2008, using the terms “male circumcision”, “HIV infection” and “sexually transmitted infection,” plus the combination of the search terms “foreskin” and “HIV receptor” to identify 1,048 articles. We reviewed the abstracts to identify 278 articles meriting detailed review. This detailed review considered how well individual studies were designed and carried out, using a standard checklist to provide a systematic quality rating for individual studies. This process identified a total of 80 papers, which were rated following the level of evidence and grade of recommendation scales modified from the Oxford Center for Evidence-Based Medicine.
Detailed analysis of the selected articles on male circumcision and HIV infection risk revealed the following. Systematic reviews, meta-analyses and modeling studies: there were 11 papers, 10 positive (favoring circumcision) and 1 negative; of the 10 positive studies, 4 were level 3 evidence, 5 were level 2 and 1 was level 1 evidence. Randomized controlled trials: there were 3 studies, all positive with level 1 evidence. Non-randomized cohort studies: there were 6 papers, 5 were positive (2 level 3 and 3 level 2 evidence) and 1 was negative (level 3 evidence). Casecontrol studies: there were 12 studies, 11 positive (all level 3) and 1 negative (level 3 evidence). Case series: there were 2 studies, both positive (level 3 evidence). Expert opinion: there were 34 studies, 30 positive (15 level 4, 15 level 3 evidence), 2 negative (both level 4) and 2 neutral (both level 4 evidence). Cost-effectiveness studies: there were 3 studies, all positive, all level 2 evidence. Pertinent biological studies: there were 3 studies, all positive, all level 4 evidence. The three large, exceptionally well-done randomized, controlled trials of adult male circumcision among consenting, healthy men in three African countries enrolled a total of 10,908 uncircumcised, HIV-negative adult men. The cumulative HIV infection risk estimated using intention-to-treat Kaplan-Meier analysis showed an overall rate ratio (RR) of 0.42 (95% confidence interval (CI) 0.31–0.57), corresponding to a protective effect of 58% (95% CI 43-69%). Meta-analysis of the “as-treated“ results of the three trials showed even stronger protection against HIV infection in the circumcision group (summary RR 0.35, 95% CI 0.24–0.54).
Rigorous analysis of the available scientific evidence clearly supports a positive recommendation that male circumcision should be actively promoted in populations at high risk of HIV infection. There is a need to provide safe male circumcision services for high-risk populations, because this is one of very few proven HIV prevention strategies. Male circumcision provides a much-needed addition to the limited HIV prevention armamentarium. The challenges to implementation must now be faced.