University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Home > Global Health Literature Digest > Male Circumcision
Male circumcision for HIV prevention in sub-Saharan Africa: who, what, when?
Global Health Sciences Literature Digest
Published April 17, 2009
Journal Article

White RG, Glynn JR, Orroth KK, et al. Male circumcision for HIV prevention in sub-Saharan Africa: who, what, when? AIDS 2008;22:1841-50.

In Context

Circumcision has been demonstrated to reduce HIV infection among men(1,2,3) and is recommended as a prevention tool in high prevalence settings.(4) Its population-level impact and cost-effectiveness, however, still are not known. Factors that may influence the effectiveness of circumcision on HIV incidence include the target group, the co-occurrence of sexually transmitted infections (STIs), risk compensation, and resumption of sexual activity before complete wound healing. This study examines the effect of various factors on the impact of circumcision on HIV incidence.


To estimate the variation in the population-level impact and cost-effectiveness of male circumcision by target age group, coverage, time to scale-up, level of risk compensation, and circumcision of HIV-infected men


A high HIV prevalence setting in sub-Saharan Africa

Study Design

STDSIM is an individual-level model that simulates the natural history and transmission of sexually transmitted diseases (STD), i.e. STIs. The model allows for changes in the characteristics of the populations examined over time. The baseline scenario used characteristics from a high HIV prevalence setting between 1997 and 2006 and with 25% of the male population circumcised. It was assumed that condoms were not used in long-term sexual partnerships. The simulated STIs were assumed to increase susceptibility to and infectivity of HIV. The presence of multiple co-factors in an individual was assumed to be an additive and overall cofactor effect for each partner in a HIV serodiscordant partnership. Lack of circumcision was assumed to increase directly the risk of HIV transmission among men, and the magnitude of the per-contact risk was fitted using clinical trial data from Kenya.2 Estimates of the per-contact risk of transmission from men who resumed sex before complete healing, which was assumed to take four weeks, was computed using a Bernoulli model and determined to be 9.8%. It was assumed that 15% of men engaged in intercourse prior to healing.

It was also assumed that lack of circumcision doubled the per-contact risk of acquiring syphilis and chancroid and that HSV-2 ulcers were decreased by 50% among circumcised men. The point estimate of HIV incidence among uncircumcised men in the Kenya trial was 59%, with lower and upper limits of 30% and 76%, respectively. These values were used in each simulated scenario.


Estimated HIV incidence from January 1, 2004 through December 31, 2006


A) In individual-level impact data, circumcision of 100% of men aged 15-24 years in the estimated population (based upon Kisumu, Kenya) was assumed to have occurred on January 1, 2004. The proportion of the observed trial data that were due to direct and indirect effects of circumcision was estimated by removing direct and indirect effects in the model.

B) Population-level impact was estimated using the individual-level impact of the various scenarios, all of which were assumed to start on January 1, 2007. The default intervention resulted in a linear increase in the proportion of circumcised men increasing from 25% to 75% within five years. It was assumed that risk compensation was negligible. Additional simulations were done for various age groups. Results are based on means over 500 runs. The cost-effectiveness of the intervention was calculated over 2, 5, 10, 20, 30, 40, and 50 years.

Impact on HIV incidence was calculated as one minus the mean annual incidence rate ratio in people aged 15-49 years. The cost and number of HIV infections averted was calculated. Circumcision costs were based on data from the clinical trials and adjusted for inflation.(5,6)

The sensitivity analysis varied the coverage, scale-up period, risk compensation, circumcision of HIV-infected men, and condom use and included a scenario in which a strong effect of circumcision on male-to-female transmission was assumed. As well, the cost-effectiveness analysis among people aged 15-49 years was recalculated under different scenarios: removing the effect of chancroid and syphilis, increasing the proportion of men who resumed sex before wound healing, the effect of the intervention in populations with different HIV prevalence rates and trends, and cost of surgery.


A) In the baseline scenario, during the two-year simulated trial period, more than 95% of the estimated impact was due to the direct effects of circumcision and less than 5% due to its effect on HSV.

B) In population impacts, the default intervention had the greatest impact on men. The impact on incidence targeting men 15-49 years of age was 8.1% at two years, 31.8% at 20 years, and 50.6% at 50 years. Among women the impact was 1.3% at two years, 21.5% at 20 years, and 43.5% at 50 years. Over time, targeting people aged 15-24 years had a larger impact on incidence in both sexes. Circumcision among neonates would begin to reduce HIV incidence in the general population after 20-30 years.

C) The default intervention would avert 28.3, 307.2, and 745.1 infections per 1000 circumcised men over two, 20, and 50 years. Until 10 years, more infections are averted per 1000 circumcised men by targeting the ages of 25-29 years or 30-34 years, but over 20 or more years, targeting those aged 15-49 years or 15-24 years would avert a similar number of infections. The cost of averting one HIV infection in the default scenario targeted at people aged 15-49 years at two years was $1,806, at 20 years was $195, and at 50 years was $89. Over the first 10 years, cost-effectiveness was highest if men aged 25-34 years were circumcised and was predicted to be cost-saving compared with lifetime HIV treatment costs.

D) Coverage and scale-up were linearly related to the intervention impact, and increases in risk behavior reduced impact. More infections were averted with increases in condom use.

E) Circumcising HIV-infected men did not substantively affect impact on incidence. CONCLUSIONS: Circumcision is a cost-saving intervention under a variety of possible scenarios. Recommendations to expand the targeted age group would increase the impact on HIV incidence. Efforts to minimize risk compensation will maximize its effectiveness.

Quality Rating

This was a high-quality, cost-effectiveness modeling study that provided the impact of the intervention on HIV incidence, which is a benefit even in the absence of cost-effectiveness, and on cost. Appropriate estimates were used and plausible variations in scenarios were assessed. In addition, the objectives were clear and measurable, a sensitivity analysis was performed, the variables used were plausible and obtained from appropriate references, there was sufficient time to consider outcomes, cost measurements were appropriate, and the conclusions was justified by the results.

Programmatic Implications

Circumcision greatly reduces the risk of HIV infection among men. Current recommendations(4) for sub-Saharan Africa are to target 12-30 year olds. The findings from this study suggest that at least initially, targeting older men may be more cost-effective. Targeting boys before sexual debut may be efficient but not as cost-effective because of the long time between circumcision and when men are at highest risk of HIV infection. Similarly, circumcising neonates, which is less expensive than surgery in older individuals and ensures fully a fully healed wound prior to sex, is not cost-effective until many years later. For circumcision to have the greatest impact on HIV incidence, the target group should be widened and scale-up should be accelerated.


  1. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, Puren A. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 Trial. PLoS Med 2005;2:e298.
  2. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, Krieger JN, et al. Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet 2007;369:643-56.
  3. Gray RH, Kigozi G, Serwadda D, et al. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet 2007;369:657-66.
  4. UNAIDS/WHO. WHO/UNAIDS technical consultation male circumcision and HIV prevention: research implications for policy and programming. Montreux: UNAIDS/WHO; 6-8 March 2007.
  5. Kahn JG, Marseille E, Auvert B. Cost-effectiveness of male circumcision for HIV prevention in a South African setting. PLoS Med 2006;3:e517.
  6. Krieger JN, Bailey RC, Opeya J, et al. Adult male circumcision: results of a standardized procedure in Kisumu District, Kenya. BJU Int 2005;96:1109-13.