Slow progress in expansion of voluntary circumcision coverage

HIV incidence already falling, but cost savings will only be seen over long term

Gus Cairns
Published: 07 March 2013

Progress towards greater coverage of voluntary male circumcision (MC) in Africa remains slow, and is due as much to men’s fear of the operation as it is to resource constraints, despite low and decreasing rates of complications, the 20th Conference on Retroviruses and Opportunistic Infections (CROI 2013) heard on Monday.

Increased circumcision rates are already leading to lower HIV incidence in both men and women in areas where coverage is high, and also to lower rates of other sexually transmitted infections (STIs), human papillomavirus (HPV) and genital ulcer disease in female partners of circumcised men.

But because of the cost of scale-up, cost savings are unlikely to be seen for 10-15 years, despite lower treatment costs for HIV and STIs. Higher cost savings would be achieved over the long run by circumcising infants as well as adults and adolescents, but this will depend on parents seeing a benefit in it.

Dr Connie Celum of the University of Washington in Seattle introduced a poster discussion session on circumcision. She noted that in a high-coverage area such as Rakai in Uganda, where one of the three randomised controlled trials (RCTs) that showed male circumcision significantly reduced HIV transmission was conducted, HIV incidence had already halved in the male population (from 1.2% a year to 0.6%).

However only three countries in Africa – Kenya, South Africa and Zambia – had reached even 20 to 30% of their target for uptake of MC, and only four more (Tanzania, Swaziland, Botswana and Zimbabwe) had achieved even 10,000 by April 2011. This was partly due to resource limitations and the training up of staff, but also to the reluctance of some men to be circumcised.

Uptake of circumcision - Rakai

A survey from Rakai showed that while 97% of men there knew that circumcision prevented HIV and 96% knew where to get circumcised, only 28% of non-Muslim men in the province were as yet circumcised, rising to 35% in 25-29 year olds, the age most at risk from HIV. Generally, those who were at highest risk were more likely to get circumcised, but not very much more: for instance, 35% of men who had more than one partner were circumcised.

This does represent a doubling of circumcision rates from before the RCT there – 12% of non-Muslim men were circumcised in 2004 – but clearly a large number of men are still reluctant to consider it. When asked, 45% of men who had not been circumcised said they were undecided, 27% said they were interested and 26% said they would definitely not get circumcised.

When asked why they had not been circumcised, two-thirds of men said they were mainly concerned about pain and 45% about things going wrong and being injured. Only 12% said they did not feel at risk of HIV and so did not need circumcision. Presenter Xiangrong Kong commented that uptake rates were far too low for Rakai to reach its target of 80% of men circumcised by 2016.

Dr Celum commented that demand stimulation was clearly going to be as important for circumcision scale-up as resource allocation. In particular, more work was needed in how to incorporate medical MC into traditional rites of passage and how to move towards infant MC.

Adverse events decreasing - Kenya

A study from Nyanza in Kenya, across Lake Victoria from Rakai, found that adverse events for male circumcision had fallen considerably since circumcision programmes were started there in 2008. The rate of moderate or severe adverse events (AEs) happening during the operation, such as severe bleeding, was only 0.23% and the rate of post-operative AEs (in the week after the operation), such as swelling or infection, was 2.14%.

These rates fell 20-fold and 4-fold, respectively, between 2008 and 2011, though because only 28% of men actually returned for the seven-day follow-up, the true rates of post-operative AEs cannot be determined. Trained nurses had the lowest rates of adverse events but were also in shorter supply than other medical officers.

Cost-effectiveness of circumcision – new model

How much could be saved by scaling-up voluntary medical MC? A mathematical modelling study by Johns Hopkins University in the US shows that programmes would only actually start to be cost saving 10 to 15 years after implementation.

The model used data from Rakai and posited four different scenarios: a fast or slow scale-up of adolescent and adult MC (ages 15-49); or adding in a fast or slow scale-up of infant male circumcision as well. It then forecast the savings, for the whole of Uganda, in terms of HIV and other STIs (penile and cervical cancer, genital ulcer disease, and trichomoniasis and bacterial vaginosis in women) averted over five years, and over 25 years.

It was quite a conservative model, using quite high estimates of the cost of the MC operation (US $42 in adults and $17 in infants, where Connie Celum cited costs as low as $30 for adults and $5 for infants) and a low estimate of 51% for MC efficacy against HIV, whereas the efficacy seen in the other two RCTs in Kisumu and Orange Farm was about 60%.

It found that there would be immediate cost savings from infections averted of around $200,000 after five years for a gradual scale-up or $600,000 for a rapid scale-up. However these would be offset by the cost of scale-up, which would be about $1 million for a gradual scale-up or $1.8 million for a rapid one.

By 25 years from now, however, the savings from adults and adolescent MC would be $15 million and the cost $9 million, resulting in a net saving of $6 million. If infants were also circumcised, the net saving would be $13 million.

Mothers and infant circumcision - Botswana

Botswana is an example of a country which, despite having one of the highest HIV prevalences in the world, has so far put far fewer resources into voluntary medical male circumcision, with only 12,800 operations by April 2011, than it has into HIV treatment (although, because of Botswana's small population, this is actually a higher proportion than Kenya’s 250,000).

Researchers from Harvard Medical School and the University of Botswana gave mothers of baby boys a questionnaire asking their opinions on circumcision, and offered it for their babies if they were interested in the idea. A total of 768 questionnaires were given out and 547 (71%) completed. A large majority – 93% – of the women said they were interested in having their boys circumcised, but only 302 (55%, or 40% of all women approached) had them circumcised in the end.

The women were on average 26 years old and while only 12% were married, 90% were supported by a primary partner. In this very high-prevalence and treatment-literate country, 98% of the women had had an HIV test and 64% knew that male circumcision helped prevent HIV. One-third of the women was HIV positive and one-fifth had a partner they knew had HIV, with one in six had a partner of unknown status.

Most of the women said they made their own decisions about their baby’s health, but 15% deferred to their male partner and 7% to their mother. Women who made their own decisions were 48% more likely to bring their baby forward for circumcision and 20% more likely if their partner had HIV; conversely, if their partner made the decision, they were 34% less likely to bring the baby forward for circumcision.

All women were asked what they thought was the most important reason a mother might choose to have her son circumcised. Fifty-one per cent of those who had their babies circumcised mentioned HIV as the most important reason, but 25% said for cultural or religious reasons and 29% for general hygiene.

Health benefits to women - Rakai

What about the benefit of circumcision to women themselves? Another study from Rakai looked not at HIV but at human papillomavirus (HPV) infections in women partners of circumcised and uncircumcised men. The baseline of the survey was the date of the male partner’s circumcision, in the case of women with circumcised partners.

Nearly all the women had at least one type of HPV at the beginning of the survey but only 14% and 13% of women with circumcised and uncircumcised partners respectively had three or more types. Two years later, the proportion of women with three or more types had not increased in partners of circumcised men but had increased to 17% in partners of uncircumcised men.

The proportion of women who acquired a new HPV infection in the first year was 40% and 44% after the first year but during the second year was 31% in partners of circumcised men and 47% in partners of circumcised men – a significant 35% difference.

There was also a significant difference in genital ulcer disease (GUD). While 15% and 16% of the women with circumcised and uncircumcised partners respectively had GUD at baseline, 12% and 21% respectively had it by year two. Benefits to women such as lower rates of STIs and HIV are expected to accumulate over time.


Celum C Introduction to session 14, Themed Discussion: Implementation and Cost-effectiveness of Male Circumcision. 20th Conference on Retroviruses and Opportunistic Infections (CROI), Atlanta, 2013.

Kong XR et al. Male Circumcision Coverage by Risk Profiles: Rakai, Uganda. 20th Conference on Retroviruses and Opportunistic Infections (CROI), Atlanta, abstract 1009, 2013. View abstract 1009 on the conference website.

Chesang K et al. Factors Associated with Voluntary Medical Male Circumcision Adverse Events: Nyanza Province, Kenya. 20th Conference on Retroviruses and Opportunistic Infections (CROI), Atlanta, abstract 1008, 2013. View abstract 1008 on the conference website.

Kacker S Costs and Effectiveness of Male Circumcision Scale-up for the Prevention of HIV and Other Sexually Transmitted Infections: Sub-Saharan Africa. 20th Conference on Retroviruses and Opportunistic Infections (CROI), Atlanta, abstract 1010, 2013. View abstract 1010 on the conference website.

Plank R Uptake of Neonatal Male Circumcision as Part of HIV Prevention Efforts in Botswana: Maternal Motivators and Barriers. 20th Conference on Retroviruses and Opportunistic Infections (CROI), Atlanta, abstract 1011, 2013. View abstract 1011 on the conference website.

Tobian A et al. Male Circumcision Decreases High-risk of Human Papillomavirus Viral Shedding in Female Partners: Analyses from a Randomized Trial in Rakai, Uganda. 20th Conference on Retroviruses and Opportunistic Infections (CROI), Atlanta, abstract 1012, 2013. View abstract 1012 on the conference website.

A webcast of the session in which this research was presented, Implementation and cost-effectiveness of male circumcision, is available on the conference website.



Find details of HIV services in Uganda, the latest news from the country, and a selection of resources from local organisations.

Find out more about Uganda >

NAM is partnering with gTt (Barcelona), GAT (Lisbon) and LILA (Como) to deliver the CROI 2013 bulletins, which have also been made possible thanks to support from Bristol-Myers Squibb. NAM’s wider conference news reporting services have been supported by Abbott, Boehringer Ingelheim, Janssen, Roche and ViiV Healthcare. The funders have no editorial control over the content of the materials.

Community Consensus Statement on Access to HIV Treatment and its Use for Prevention

Together, we can make it happen

We can end HIV soon if people have equal access to HIV drugs as treatment and as PrEP, and have free choice over whether to take them.

Launched today, the Community Consensus Statement is a basic set of principles aimed at making sure that happens.

The Community Consensus Statement is a joint initiative of AVAC, EATG, MSMGF, GNP+, HIV i-Base, the International HIV/AIDS Alliance, ITPC and NAM/aidsmap