Circumcision is not the only fruit

Lopping off foreskins reduces HIV transmission. But we haven’t always known that. Indeed when I was in grad school (not ALL that long ago) I remember being given a paper overlaying a map of tribal circumcision practices in Africa with HIV rates. The paper concluded that circumcision might slow the spread of HIV; it was held up as an example of bad epidemiology.

HIV is low where circumcision is high, yes, just as TB is low where Mercedes ownership is high. But that doesn’t mean circumcision protects you from HIV any more than driving a Merc protects you from TB. And then the lecture about confounding factors such as religion, which may affect both the “explanation” — circumcision — and the thing we’re trying to explain — HIV. Leading to next week’s lecture, about how all will be neatly explained by the randomised controlled trial, the Rolls Royce of study designs.

It may have been a weak paper, but it turns out it was true. We know that because colleagues did go on to do the Rolls Royce studies, randomly assigning some guys to get snipped and others to stay in one piece. The men who lost their foreskins were 60% less likely to get HIV than the men who didn’t go under the knife. According to Jeffrey Klausner and colleagues, writing in Future Medicine, “male circumcision is the only modality for preventing sexual HIV transmission that has been proven to work by the highest standards of scientific evidence; specifically, findings from multiple randomized controlled trials”.

Hang on a minute. That implies that other methods have been proven not to work. But that is absolutely not true. Condom promotion has worked exceptionally well in many settings, most famously in the brothels of Thailand, slashing new sexual infections from nearly 120,000 a year in 1989 to 12,000 a year in 2005 (see AIDS in Asia: warning – gargantuan pdf file). It hasn’t been shown to work by the “highest standards of scientific evidence” precisely because it IS so efficient. We KNOW condoms work to prevent the spread of HIV. It would therefore be unethical to randomly assign people who might be at risk for infection to a “No condom use” group, just for the sake of the highest standards if scientific evidence. Sometimes, common sense trumps randomised controlled trials. If you’re not convinced, please read the British Medical Journal’s seminal paper: Parachute use to prevent death and major trauma related to gravitational
challenge: systematic review of randomised controlled trials

This post was published on 11/01/08 in Science, Uncategorized.

Send this post to a friend Send this post to a friend


You can follow the comments on this post via this RSS feed.

Tags: .

  1. Comment by Hugh G Rection, 12/01/08, 02:22:

    Plummer was lecturing about the association between HIV prevalence and circumcision in 1989. Sexual concurrency and HIV prevalence was being openly discussed in progressive HIV circles in the early 90s. All that is new is old …


  2. Comment by Candy, 12/01/08, 06:32:

    It would therefore be unethical to randomly assign people who might be at risk for infection to a “No condom use” group, just for the sake of the highest standards if scientific evidence.

    Yes. And yikes.

  3. Comment by Lee Rudolph, 17/01/08, 03:44:

    What, two comments already and neither has taken the “seminal paper” bait? Well, I won’t either, then…

  4. Comment by Hugh7, 11/06/08, 01:46:

    “the highest standards of scientific evidence; specifically, findings from multiple randomized controlled trials”

    Not quite the highest. They weren’t double-blinded or placebo-controlled (and perhaps couldn’t be), and the drop-out (“lost to study”) rates were about five times higher than the number of known seroconversions. (If you’d undergone painful, marking surgery to prevent HIV and then been tested elsewhere – as they were encouraged to be; it was considered “unethical” to tell them they were HIV+ – and found you had it, would you go back to the people who’d done that to you?)

    The experimental group had to abstain from sex for six weeks after their operations, and given more condom advice than the control group. (Maybe being told to use condoms when you’ve just been circumcised focusses the mind more?)

    The sleeve technique used in the Ugandan trial removes more tissue than the forceps-guided technique used in the other two countries, yet all three methods produced the same amount of “protection”. (The forceps-guided method is very similar to the tribal technique used in Lesotho and blamed for the high rate of HIV there.)

    And since all the men had volunteered to be circumcised sooner or later, they were not a random selection of the population.

  5. Comment by Hugh7, 14/06/08, 11:39:

    You say “according to Jeffrey Klausner and colleagues…”. As you know, sometimes the “colleagues” are more influential than the lead writer. In this case they include Daniel T Halperin, who has been at the forefront of promoting circumcision for over a decade. He is on record as thinking his descent from a ritual circumcisor meant “maybe in some small way I’m ‘destined’ to help pass along [circumcision] to people in [other] parts of the world … .” (Cover Story: The Case for Circumcision. By Gordy Slack. The East Bay Express Online. May 19-24, 2000.) Whatever else that it, it’s not science.

    Check those “colleagues” and the same few names keep coming up again and again: Daniel T Halperin, Robert Bailey, Malcolm Potts, Helen Weiss, Bertran Auvert, Frank Klausner, Stephen Moses, Roger Short, Xavier Castellsagué, Stefan Bailis. Are they more interested in preventing AIDS or promoting circumcision?

    And “60% reduction” sounds impressive, but what it means is that they circumcised a total of 5,400 men and found that after less than two years, 64 (1.2%) had HIV, compared to 137 (2.5%) in similar-sized groups left intact – 39 circumcisions for each HIV infection deferred (not prevented). How does the cost of that compare to education or condoms?

    Circumcision is now being marketted in conjunction with partner reduction – if the HIV rate declines, which will have done the job, and which will get the credit?

  6. Comment by Hugh7, 08/07/08, 04:08:

    Following my comment #4, here is a graph showing how very easily the number of circumcised men who were “protected” from HIV could have been swallowed up in those “lost from study”:

  7. Comment by cosmopolite, 19/07/10, 12:37:

    The randomised clinical trial is a Rolls Royce in theory. But the practice here left a great deal to be desired, for reasons that include those Hugh7 lays out above.

    Around the North Atlantic, nonMoslem adult Europeans are not circumcised, Americans mostly are, and Canadians are somewhere in between. The North Atlantic nation with the highest rate of seropositivity is, by far, the USA. For me, that is conclusive evidence that routine infant circumcision is irrelevent to the battle against AIDS in the First World.

  8. Comment by Dave, 09/01/11, 08:07:

    Circumcision does not prevent HIV and condoms and having safer sex to a better job of preventing HIV transmission than any sort of pointless genital mutilation does.

Comments are closed at this time.