Circumcision in men and the prevention of HIV infection: a 'meta-analysis' revisited

Abstract

There is debate on the role of male circumcision in HIV transmission. Most case-control and cohort studies from Africa have shown an association between a lack of circumcision and an increased risk of HIV infection in men. The evidence is conflicting, however, with cross-sectional surveys from Tanzania and Rwanda either showing no relationship or an association in the opposite direction. A recent review and meta-analysis of the literature concluded that the risk of HIV infection was lower in uncircumcised men (combined odds ratio 0.94, 95% confidence interval 0.89 to 0.99). However, the analysis was performed by simply pooling the data from 33 diverse studies, which is an inappropriate method for combining studies. We re-analysed the data, stratifying by study, and found that an intact foreskin was associated with an increased risk of HIV infection: combined odds ratio 1.43 (1.32 to 1.54) with a fixed effect model and 1.67 (1.25 to 2.24) with a random effect model. There was significant between-study heterogeneity (P < 0.0001) which was partly explained by stronger associations in studies in high-risk groups. The results from this re-analysis thus support the contention that male circumcision may offer protection against HIV infection, particularly in high-risk groups where genital ulcers and other STDs 'drive' the HIV epidemic. A systematic review is required to clarify this issue. Such a review should be based on an extensive search for relevant studies, published and unpublished, and should include a careful assessment of the design and methodological quality of studies. Much emphasis should be given to the exploration of possible sources of heterogeneity. In view of the continued high prevalence and incidence of HIV in many countries in sub-Saharan Africa, the question of whether circumcision could contribute to prevent infections is of great importance, and a sound systematic review of the available evidence should be performed without delay.