Outcomes

Anticipated absolute effects per year

Incidence Rate Ratio

(95% CI)

No of Participants

(studies)

Quality of the evidence

(GRADE)

Risk without circumcision (Control)

Risk difference with circumcision  (95% CI)

HIV incidence (0-12 months)

16 per 1000

8 fewer per 1000 (from 4 to 10 fewer)

IRR* 0.50 (0.34-0.72)

10,908 (3 studies)

High

HIV incidence (0-24 months)

10 per 1000

5 fewer per 1000 (from 4 to 7 fewer)

IRR 0.46 (0.34-0.62)

10,908 (3 studies)

High

Adverse Events: The rate of adverse events for medical circumcision ranged from 1.72% to 7.65%. Reported adverse events included: pain, excessive bleeding, infection, damage to the penis, delayed wound healing, insufficient skin removed, problems with urinating, problems with appearance, and swelling. All three trials found that men in the circumcision group were more likely to practice risky sexual behaviors than controls.

About quality of evidence (GRADE)
 High: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low: We are very uncertain about the estimate.

* IRR = incidence risk ratio

Findings

Interpretation

Equity - Which of the PROGRESS groups examined

Participants in the three RCTs were from South Africa, Uganda, and Kenya. All participants were sexually active men between the ages of 15 and 49. All the men were HIV uninfected at enrollment.

There is no evidence on the effectiveness of circumcision as a preventive procedure prior to the start of sexual activity but this is something that could be explored in future studies. The men most likely to benefit from a medical circumcision intervention should be selected.

Equity Applicability

Circumcision in the included studies was performed using commonly used surgical techniques and local anaesthesia. In many countries, circumcision is performed by a traditional healer as part of the rites of initiation into manhood

The results of the study are specific to circumcision performed as a medical procedure and do not support traditional circumcisions which have higher rates of adverse effects. Policymakers who want to implement a circumcision intervention will need to find a way to incorporate the medical procedure of circumcision into these traditional customs.  Patients undergoing medical circumcision need to receive information about wound care and identifying possible complications.

All three trials were conducted in areas with high rates of HIV transmission.

The risk of HIV infection in the population should be considered before planning a medical circumcision intervention. There is no information yet to describe the effectiveness of male circumcision on female partners.

Cost-equity

The review does not comment on the cost-effectiveness of medical male circumcision.

The cost-effectiveness of the intervention needs to be explored. Policy makers need to consider the costs included in the medical procedure, including local anaesthetic and follow up care when needed. The costs of HIV infection should be weighed against the cost of medical circumcision. This review reports that the number needed to treat is 62 to prevent 16 HIV infections assuming a control risk of 3% (30 per 1000)

Monitoring & Evaluation for PROGRESS groups

Longer term evaluation of the effectiveness of medical circumcision is needed. Other studies should examine the feasibility of implementing this intervention in different contexts as well as the related social and cultural issues.

All three trials found that men in the circumcision group were more likely to practice risky sexual behaviours, such as unprotected sex, than controls. During the two year study follow up period circumcised men were protected but more follow up studies are needed to determine if circumcision benefits persist long term.