Five years on from circumcision trial, nine in ten participants are circumcised and HIV incidence is two-thirds lower

This article is more than 11 years old. Click here for more recent articles on this topic

Five years after the ending of one of the three big randomised controlled trials of male circumcision as an HIV prevention measure, four out of five men who were in the control arm of the trial and thus not circumcised have opted to get circumcised, a follow-up study presented to the 18th Conference on Retroviruses and Opportunistic Infections has found.

The study also found that, if anything, the protective effect ascribed to circumcision appears to have strengthened over time.

The post-trial analysis was conducted on the randomised controlled trial of male circumcision as an HIV prevention measure conducted in Rakai, Uganda, in 2005-6 (Gray 2007). In this study 4996 HIV-negative men aged 15 to 49 were randomised either to be immediately circumcised or to be offered circumcision at the end of the trial.


voluntary male medical circumcision (VMMC)

The surgical removal of the foreskin of the penis (the retractable fold of tissue that covers the head of the penis) to reduce the risk of HIV infection in men.


The surgical removal of the foreskin of the penis (the retractable fold of tissue that covers the head of the penis) to reduce the risk of HIV infection in men.

control group

A group of participants in a trial who receive standard treatment, or no treatment at all, rather than the experimental treatment which is being tested. Also known as a control arm.

safer sex

Sex in which the risk of HIV and STI transmission is reduced or is minimal. Describing this as ‘safer’ rather than ‘safe’ sex reflects the fact that some safer sex practices do not completely eliminate transmission risks. In the past, ‘safer sex’ primarily referred to the use of condoms during penetrative sex, as well as being sexual in non-penetrative ways. Modern definitions should also include the use of PrEP and the HIV-positive partner having an undetectable viral load. However, some people do continue to use the term as a synonym for condom use.

risky behaviour

In HIV, refers to any behaviour or action that increases an individual’s probability of acquiring or transmitting HIV, such as having unprotected sex, having multiple partners or sharing drug injection equipment.

The study was designed to last two years but was terminated early in December 2006 when it was found that HIV infections were just under half as common (efficacy, 51%) in men who had been randomised to be circumcised compared with men in the control group.

Later analyses showed that this efficacy underestimated the true effectiveness of circumcision. The HIV infection rate in men who actually got circumcised was 58-60% lower than in men who remained uncircumcised, and 70% lower in men with high numbers of partners.

Dr Xiangrong Kong of Johns Hopkins University told the conference in Boston that by the end of the fifth year after the study ended just over 80% of the control group, who had not been circumcised during the trial, had opted for circumcision and, out of 2916 men who were uncircumcised at the last scheduled visit during the trial, only 372 men now remained uncircumcised. Including the intervention group and excluding those lost to follow-up, 90% of those who entered the study had been circumcised.

Looking at men who were not circumcised during the trial, HIV incidence in men who got circumcised, in the post-trial period was one infection per 181 men per year after circumcision (0.55%), and in men who remained uncircumcised one infection per 60 men a year (1.67%). Circumcision was thus 68% effective. If the trial period was included this made very little difference and efficacy still stood at 67%.

If the men who got circumcised during the trial were included, then the overall efficacy of circumcision over the whole period from the start of the study was 73%.

There are data on sexual behaviour for the first 2.8 years since the end of the trial. Before the trial, there had been concerns that circumcision might produce behavioural disinhibition in men and an increase in unsafe sex, especially once men knew circumcision worked.

In the original trial, 18% of participants reported consistent condom use during the trial and 52% did not use them at all. During the follow-up period, condom use declined by 4.3% in consistent condom users to 13.5% and the proportion who never used them increased by 6% to 58.2%.

But there was no difference in decreases in condom use between circumcised and uncircumcised men, and in fact condom use levels now are almost exactly what they were at baseline before the start of the study. The declines in condom use therefore probably reflect reduced availability of condoms and safer sex advice post-trial, rather than any disinhibiting effect of circumcision.

There was no change in the number of non-marital sexual partners, and a 9.4% decrease in the number of men who reported alcohol use during sex, again with no difference between circumcised and uncircumcised men.   

These findings are remarkably similar to a post-trial analysis 3.5 years after the end of one of the other two circumcision efficacy trials, in Kisumu, Kenya, presented at the 2008 International AIDS Conference, which found a long-term efficacy of 65% for circumcision and no increase in risk behaviour.      

Abstract and webcast

You can view the abstract from this research on the official conference website:

Abstract 36:

You can also watch webcasts of presentations made at the conference.

The webcast from the conference session HIV Prevention: HSV2, Topical and Oral PrEP, and Circumcision, includes the speaker Xiangrong Kong.

We rely on donations to continue our work to help people with HIV, support us today at:


Kong XR et al. Longer-term effects of male circumcision on HIV incidence and risk behaviors during post-trial surveillance in Rakai, Uganda. Eighteenth Conference on Retroviruses, Boston, abstract 36, 2011.

Gray RH et al. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. The Lancet 369(9562):657-666. 2007. Original study abstract available here.