Monday, September 27, 2010

Some Statistics Are More Equal Than Others

I recently wondered out loud why those who were baying for mass male circumcision in high HIV prevalence African countries didn't seem to be as interested in the possibility that washing with soap and water might be just as effective. Well, one Thomas J Coates has received 800,000 dollars to evaluate the feasibility of a "post-coital genital hygiene study among men unwilling to be circumcised" (compared to an estimated $50 million to circumcise just over a million members of the Luo tribe).

If this 'genital hygiene' involved something as basic as soap and water, that might be worth celebrating. But I doubt if it is. It's more likely to involve 'penile wipes' or some other technology. (Nothing wrong with technology, but let's just keep it appropriate.) And the study is not to establish if such measures would protect as well as circumcision, better or less well. It is to establish if men would find the practice acceptable, convenient, practicable and if adherence is high.

The project information says "If we find that men are able to practice consistent washing practices after sex, we will plan to test whether this might protect men from becoming HIV infected in a later study." It's surprising that they are going ahead with the plan to circumcise a million people, even though penile hygiene may be all that's required!

But I am not sure what they mean by 'able'. I assume most men are able to wash their penis, whether circumcised or not. I'm also wondering why pre-coital hygiene is not also being observed, with a view to reducing male to female transmission of HIV (and other sexually transmitted infections), not just female to male transmission, which is often less of a risk.

If "adherence is high" (and if men are 'able'), genital hygiene might then undergo randomized controlled trials to be considered as a HIV prevention intervention. If adherence is not high, some serious questions would need to be raised about the way the research was carried out. Personal cleanliness is considered extremely important everywhere I have been to in Kenya, Tanzania and Uganda, whether running water is available or not. But by the time the results are available, many more men will have been circumcised, possibly unnecessarily, possibly at considerable risk to their health.

Male circumcision is, all things considered, an odd HIV prevention intervention in a country with relatively low health care standards and low levels of access. Outside of the Luo tribe, circumcision rates are between 90 and 100%, yet HIV prevalence ranges from 0.8% among the ethnic Somalis to 7.9% among the Luhya and Maasai. That's nearly 10 times higher. And among the Embu, Somali and Meru tribes, men are just as likely to be infected as women (see table below).

In contrast, men are far less likely to be infected with HIV than women in most other tribes. Only three Luo men are infected for every four Luo women. But only 1.6 Luyha men are infected for every 10 Luyha women. So circumcision may protect men, but to widely varying degrees.

In some tribes, circumcision may not be giving much protection at all. It's hard to compare, because there are very few uncircumcised sexually active members of most tribes, but that information would hardly increase confidence in the effectiveness of circumcision, would it? In fact, it has never been clear what sort of protection male circumcision gives, if any. There is even evidence that Luos are more susceptible to HIV infection for reasons unrelated to their circumcision status.

But the passion with which mass male circumcision is advocated as a HIV prevention intervention smacks of an almost religious fervor. After all, lower HIV rates are found among those who practice female genital mutilation (FGM), too, mutilation of the worst sorts. Thankfully, none of the circumcision enthusiasts are advocating for FGM. But the highest rates of HIV are found among the people who don't practice FGM in any form (Luo and Luhya). Perhaps we shouldn't be so easily swayed by the promise of 'up to 60% protection' against HIV, or any other opaque statistics.

Tribe
HIV+
Female HIV Prev
Male HIV Prev
Ratio
HIV+ Male
Circ'd
Male Circ'd
Embu
3.0
2.7
3.2
118.5
3.3
97.8
Kalenjin
1.8
2.1
1.4
66.6
1.5
93.8
Kamba
4.1
5.5
2.4
43.6
2.5
99.2
Kikuyu
4.1
5.9
1.7
28.8
1.6
98.0
Kisii*
4.7
5.1
4.3
84.3
4.4
97.0
Luhya
7.9
12.0
1.9
15.8
2.0
95.9
Luo
20.2
22.8
17.1
75
16.4
21.5
Maasai
7.9
8.2
7.8
95.1
8.8
90.1
Meru
5.3
5.3
5.4
101.9
5.1
91.6
Miji/Swah
3.2
3.5
2.7
77.1
2.7
98.9
Somali*
0.8
0.8
0.8
100
0.8
99.2
Taita/Taveta
2.7
3.7
1.4
37.8
1.4
100
Other
3.1
5.0
1.1
22
1.1
72.8

*FGM rates almost 100%


allvoices

7 comments:

Mark Lyndon said...

Circumcision is a dangerous distraction in the fight against AIDS. There are six African countries where men are *more* likely to be HIV+ if they've been circumcised: Cameroon, Ghana, Lesotho, Malawi, Rwanda, and Swaziland. Eg in Malawi, the HIV rate is 13.2% among circumcised men, but only 9.5% among intact men. In Rwanda, the HIV rate is 3.5% among circumcised men, but only 2.1% among intact men. If circumcision really worked against AIDS, this just wouldn't happen. We now have people calling circumcision a "vaccine" or "invisible condom", and viewing circumcision as an alternative to condoms. The South African National Communication Survey on HIV/AIDS, 2009 found that 15% of adults across age groups "believe that circumcised men do not need to use condoms".

See also http://www.iasociety.org/Default.aspx?pageId=11&abstractId=2197431
"Conclusions: We find a protective effect of circumcision in only one of the eight countries for which there are nationally-representative HIV seroprevalence data. The results are important in considering the development of circumcision-focused interventions within AIDS prevention programs."

http://apha.confex.com/apha/134am/techprogram/paper_136814.htm
"Results: … No consistent relationship between male circumcision and HIV risk was observed in most countries."

This 1993 study found that "partner circumcision" was "strongly associated with HIV-1 infection [in women] even when simultaneously controlling for other covariates."

The Wawer study showed a 54% higher rate of male-to-female transmission in the group where the men had been circumcised. The figures were too small to show statistical significance, but there will be no larger scale study to find out if circumcising men increases the risk to women. Somehow that's considered unethical, yet it's considered ethical to promote male circumcision whilst not knowing if the risk to women is increased (by 54%?, 25%?, 80%? - who knows?)

Mark Lyndon said...

Others are questioning the promotion of male circumcision btw:

French AIDS Council:
"Even though the WHO insists on the idea that, beyond male circumcision, the use of other forms of prevention remains essential, it is very likely that people who mistakenly believe themselves to be adequately protected will no longer use condoms."
...
"Implementation of male circumcision as part of a draft of preventative measures could destabilize health care delivery and at the same time confuse existing prevention messages. The addition of a new tool could actually cause a result opposite to that which was originally intended."

Rozenbaum W, Bourdillon F, Dozon J-P, et al. Report on Male Circumcision: An Arguable Method of Reducing the Risks of HIV Transmission. Conseil National du SIDA, 2007: 1-10.

and as this South African paper puts it:
"Those promoting circumcision argue that circumcision is an additional tool that will ultimately reduce infections more than just relying on condoms, monogamy and abstinence. However, African males are already lining up to be circumcised, thinking they will no longer need to use condoms. Rather than complementing ABC programs, promoting circumcision will undermine the ABC approach by diverting funds and encouraging risk compensation behavior, ultimately leading to an increase in HIV infections."

ABC (Abstinence, Being faithful, and especially Condoms) is the way forward. Promoting genital surgery will cost African lives, not save them.

Simon said...

Hi Mark
Thank you for your comments and links. As always, I agree with much of it. And condoms are about all most people have for protecting themselves against sexually transmitted HIV.

However, I have never seen any credible evidence that either abstinence or being faithful have any impact on HIV transmission, only copious amounts of not very credible evidence.
S

Mark Lyndon said...

It's disturbing that no-one really knows just how much HIV transmission in Africa is non-sexual. I saw a presentation that suggested that over a third may be caused by medical procedures. Circumcision itself seems to be infecting some people, since virgin males are more likely than to be HIV+ if they've been circumcised.

I think it's hard to argue that abstinence and fidelity wouldn't help, but I think the main focus of AIDS prevention should be condoms. Sadly, a lot of funding comes from organizations who are against condoms (or targeting sex workers). It's not that long ago that a third of the entire PEPFAR budget had to be spent on abstinence programmes :-(

Simon said...

I don't believe that abstinence and fidelity programs wouldn't help, I believe they haven't helped. In other words, behavior programs have had little demonstrable impact on behavior and behavior has little correlation with HIV prevalence.

It's easy to find examples in Demographic and Health Surveys, where there is very little correlation that suggests that HIV is strongly related to sexual behavior.

Of course, all the big and expensive behavior change programs claim to have an impact, that's the advantage of being well funded. But I've seen no credible evidence. Any drop in prevalence or even incidence has been hailed as a success, but this does not demonstrate a correlation.

Uganda is a good example of a country where the epidemic struck hard and early, declined early, stagnated and then started to increase again. Kenya is probably another example.

But notice how the behavioral enthusiasts are not now churning out articles to say that their programs have failed or that sexual behavior and HIV are not very closely correlated?

Also, much though I dislike emphasis on abstinence and other prurient and useless 'strategies', there was never a time that PEPFAR mandated that one third of funding be spent on them. Rather, one third of prevention funding was, at one time, supposed to be spent on abstinence programs. That's 10% or less, not one third.
S

Mark Lyndon said...

You're right, it was a third of prevention funding.

Even now though, there are still restrictions on the promotion and distribution of condoms:
In countries with generalised HIV epidemics, at least half of all money directed towards preventing sexual HIV transmission should be for ‘activities promoting abstinence, delay of sexual debut, monogamy, fidelity, and partner reduction’.
(text of legislation)

PEPFARWatch has some interesting information.

Even if there were no risk compensation, and not a single dollar of prevention funding was diverted, I genuinely believe that promoting male circumcision in Africa will result in more new HIV infections, particularly amongst women.

Simon said...

Thanks Mark. Yes, I don't believe the figures show that circumcision is effective. And the conditions in African health facilities, certainly in Kenya, would not be conducive to safe operations or follow up. Also, the epidemic in parts of Kenya and parts of many other African countries overwhelmingly affects women, so a strategy based on massaged figures, like mass male circumcision, increases the risk for women more than men.