If you take the time to look carefully at Auvert(1)/Bailey(2)/Gray(3), the three "Randomized Controlled Trials" looking at HIV/Circumcision, and the conclusions stated therein, it's easy to see it's a total load of steaming nonsense.
Let me emphasize that you, working in an office or a construction site or a restaurant or a taxi, you can easily see how totally unsupported the grand claims are. You just need to put on your thinking cap, and take a close look.
Why am I taking the time to walk through a process which demonstrates that the conclusions drawn from the African trials are unfounded? I'm doing it mainly because some community members have asked that I address the issue. | |
You might think it's crazy to suggest that you can figure out for yourself that most of the reporting you've hear about these studies is mindless drivel regurgitating, amplifying, and twisting already unjustified conclusions, but really, it's not that hard. You might think it's a contradiction: If it's not that hard, why do the unsupported claims seem to prevail in the consciousness of the public, of the media, and of some policy-makers? I suppose that's a diary about herd mentality, group-think or some other sociological phenomenon. In any case, it's off-topic here.
You can see these claims are not supported even if you aren't a "trained scientist." That's because you are, in fact, a trained scientist, although you can never have too much training. The scientific method is for everyone. Once understood, it's merely synonymous with rational thought.
So, at last, let's get started.
Employing the cardinal rule of science, we make the following assumption:
The African RCTs are total bullshit until proven otherwise.
But, you might be told, it's already established, it's entrenched, it's on top of the mountain and we're at the base. Bullshit. No number of news articles or political or philanthropic endorsements mean anything whatsoever to scientific thought. What it does do is inform your web of trust. Most of the time we rely on sources we put some trust in and don't investigate deeply for ourselves. Nobody has time to fact and logic test everything. We develop trusted sources and give them the benefit of the doubt much of the time, often in specialized areas of expertise. But once we decide to verify something ourselves, as we're bothering to do here, what others have concluded is of no consequence. These studies have no validity in the slightest until we see it with our own critical eyes.
And, it won't be difficult to see how these African RCTs don't even come close to showing what their authors claim, let alone the "news" headlines resulting from the pathetic game of "telephone" that so often is called "journalism."
Let's ease our way into this with some facts, which will be especially useful to those do not have previous familiarity with the studies.
All three studies bear considerable similarity in how they were conducted, and we'll be looking mainly at things they have in common. For clarity, we'll sometimes cite specific examples from one of the studies.
A few facts common to each trial:
- In each study, all participants started at the same time.
- The study ended 18 months after it began (4).
- The circumcised group was instructed to refrain completely from sexual activity for 6 weeks after surgery (5) or "until healing is complete" (6), while the intact group was not instructed to refrain from sexual activity for any period of time.
- The number of participants going from testing HIV negative to HIV positive in each group are absolute numbers, for example 22 of 1393 from the intact group and 47 of 1391 from the circumcised group (7).
- Intervention group had no prepuces (foreskins) for most of the study period. The control group had foreskins.
- Intervention group had a surgery taking weeks to heal. Control group had no surgery.
The flaws are already shining like the sun. Where, even, to begin.
Validity rests on addressing reasonably identifiable confounds
Good science means eliminating damn near every confound. The further you get from that, the less valid your results. The confounding factors identifiable by any ordinary person stack from here to the moon.
Now, while the conclusions of others mean nothing here, we're free and appropriately resourceful to look for confounds wherever they may be most easily found. Why not start with some published correspondence?
Hugh Young via PLoS correspondence:
A prime requirement in any controlled study is that as far as possible, all conditions apart from the one being tested should be the same.
I already mentioned this, but since it's absolutely fundamental, it bears repeating.
Confound: Maximally similar placebo surgery was not performed on the control group. One group had a penis surgery, the other did not.
Honestly, we could just stop here. It's over. Checkmate in 4 moves. Knockout in the first minute. Seriously, WHO do they think they're fooling? WHO could be so gullible or misguided? Not the Brazilians. Not the Chinese. But it wouldn't matter if every man, woman, and child on Earth swallowed this load. The studies fall on their own lack of merit.
If you had even the slightest hope of making the two groups maximally similar in every way except that one gets circumcised, you need to do placebo circumcision surgery on the control group. The prepuce cannot be removed since its presence is part of the experiment, so a preputioplasty or dorsal slit might be done instead.
This point bears repeating.
A similar surgery but without prepuce removal must be performed on the control group to address the elephant-sized confounds of failing to do so.
The bar of scientific validity doesn't give the slightest care for what researchers find practical, for what their budgets will allow, for what ethics boards will permit, or for what participants will sign up for. |
The PLoS correspondence continues:
In the Auvert study, the men from the intervention group were instructed, in effect, as follows: "When you are circumcised you will be asked to have no sexual contact in the six weeks after surgery. To have sexual contact before the skin of your penis is completely healed could lead to infection if your partner is infected with a sexually transmitted disease. It could also be painful and lead to bleeding. If you desire to have sexual contact in the six weeks after surgery, despite our recommendation, it is absolutely essential that you use a condom"
Umm... Okay... the act of going to a clinic for penis surgery, being counseled about penis surgery, spending the extra time at the clinic, learning about the special requirements for condom use due to a healing wound, being exposed to those white folks in white coats more. Remember now, the grand conclusions of each of these studies rests on the difference of a few handfuls of individuals (e.g. 25 in Bailey) out of a few thousand participants (2784 in Bailey).
So the men in the intervention group were given very different instructions about sexual behavior than those in the control group—in precisely the field where their risk of HIV infection was most affected. This could have differentially affected their sexual behavior, and perhaps how they reported it. The time they spent waiting for and recovering from their surgery could also have exposed them to more safe-sex information and influence than the control group.
More instruction about condoms, more exposure to the clinic, more time to learn from the staff:
Confound: More counseling about and exposure to condoms in the clinic may give one group a much better chance of correctly using the condom when they report having used one.
These folks may have never even seen a condom before. Additional discussion and instruction could be very influential.
Now, let me be clear: The motives of the study leaders are no part of this. Their work stands or falls on its own. But since their work falls so quickly and so badly, we can spare a little time to examine what the heck these people could be thinking.
Let's start out being charitable: These folks probably really want to help stop HIV. If their pro-circumcision chocolate should happen to fall into the stopping HIV peanut butter, what could be more delicious? We've each got our own passions and motives, that's human nature.
3/29/2007 San Francisco Chronicle, Circumcision pushed in AIDS fight:
Robert Bailey, a medical anthropologist at the University of Illinois in Chicago who carried out one of the three African studies, called the WHO recommendations a strong endorsement. "It's a great day,'' he said. "It is something I have been working toward for 12 years.''
Well, he's persistent all right, and delighted that people actually believe these unsubstantiated conclusions from methodologically moronic studies. When I spend a couple hours writing an article, I can't proofread it right away, because having just written it I'm prone to skipping the same typos or grammatical errors that slipped through in the first place. Could the failure to see what terribly inconclusive and flawed science this is be affected by having spent more than a decade trying to prove intactness facilitates HIV infection? Ultimately irrelevant, but amusing to ponder briefly.
Following Passion Leads Epidemiologist to HIV Prevention:
Bailey read an article that stated that in the areas of Africa where HIV/AIDS was most prevalent the men were not circumcised. In fact, men who were not circumcised were 2-8 times more likely to contract HIV than men who were, the article stated.
You read an article? You heard of one pair of populations where there was a correlation, but you also checked to make sure there aren't just as many anecdotal tidbits that suggest the opposite, right? Right??
Circumcision Rate: USA 85%, UK 7%
HIV Prevalence/100,000: USA 16.0, UK 2.4
Noticing one non-circumcision society with higher HIV, or even several, sure doesn't suggest much if one can just as easily find the reverse too. And indeed there are many such pairs.
Bailey decided his next challenge would be to prove that circumcision could indeed help prevent HIV. But it was 1996 and no one was talking about circumcision as a way to prevent HIV.
"I went all over the world giving talks in the 1990’s telling people to look at the evidence and that we need to start delivering circumcision," he said. "A lot of people laughed, but they finally see that this has great potential."
Ok. I get wanting to help people. I get wanting to stop HIV. But to want to prove circumcision can help do that? Why someone would dedicate themselves to doing that, I don't care, and I don't need to. That's because one of the most fundamental by-products of the scientific method is that it totally strains out the motivations of its practitioners.
Providing the circumcision procedure gives healthcare workers an opportunity to also provide other important treatment to young men who may not otherwise come to a health facility, he said.
Huh, so he noticed that too, eh? Yet he, the leader of the study, didn't bother to develop a methodology which avoided this difference between the two groups???? Wow. Just wow. It's almost like he wanted a particular result. Hmmm. Hard to say. But again, it just doesn't matter.
Bailey is now working on opening a center of excellence in Kisumu, where healthcare workers are trained in circumcision and counseling.
Is that his sign in the second picture in the article which says "CIRCUMCISION DONE HERE" with a price quote in local currency?
OK, enough. Really, it makes no difference. People will follow their passions. A study gets done because people are sufficiently motivated to get the resources and do the work. The work stands or falls on its own merits. |
Back to the science. Where were we?
Oh, right, one group had surgery and post-surgical counseling and special instructions on condom usage and spent lots more time in the clinic. The other went home and resumed their normal routine.
These results are already as discredited as could possibly be necessary, but let's recap and elaborate:
Confound: Circumcised group had added exposure and interaction in the clinical setting. Control group lacks this experience.
For the placebo surgery to be maximally similar, it must involve a penis wound, but obviously not removal of the foreskin.
Confound: Circumcised group experienced a penis surgery. Control group did not.
Who the hell knows all the ways two groups are different when one just had a penis surgery and one did not? One might well behave quite differently after a penis surgery! It's easy to envision it being a behavior-modifying event. It's a gigantic can of worms (i.e. confound) that any serious effort must try to eliminate with an incisive penis surgery.
Confound: Circumcision group got much more counseling, like directions for care of their circumcision wound. Control group lacked this extra counseling.
Surgery means lots more time getting instructions, interacting with staff, being exposed to the clinical setting, being reminded of the need for condoms, etc.
Confound: Circumcision group heard about and learned a lot more about condoms, since they were at the clinic more and there are special precautions to take after the surgery.
Suppose you were a rural African, possibly having never used or even seen a condom before, and then some white folks show you these little squishy packages, odd things, to start wearing when make sexy time (it's nice...) | | No! Intervention group only must abstain for 6 weeks! |
Just ponder for a moment how enormous a difference between two groups it is for one of them to abstain from sex for six weeks and nurse a penis wound, and the other not. How could anyone think the control group doesn't need penis surgery too? Sometimes I am just at a loss for how such brazenly asserted conclusions from these studies can go largely unchallenged despite being based on such profound stupidity.
How condom instructions are presented, how many times, with what urgency, with what consequences for failure, how well acquainted you are with the staff explaining it, in what context, over what time period, can make a big difference. They must all be the same in both groups.
Would being around the clinic a lot more and having penis surgery and getting special counseling about how important it is that you wear that strange rubbery thing affect you differently than, well, getting a lesser amount of instruction over a shorter time period, then just going back to your previous routine? These systematic differences between the two groups may affect how successfully they deploy this western invention during their most intimate interpersonal moments, while reporting proper usage.
The coffin is nailed shut and the corpse is emitting a noxious odor, but let's see what the Washington Blade has to add:
4/20/2007 article entitled N.Y. circumcision plan derided as ‘insane’:
(All mental bell chimes and italics are mine)
‘Bad science’
...
Ryan McAllister, 30, a biophysics research fellow at Georgetown University in Washington, said findings from the three studies are inapplicable to gay men [Ding!]. He also noted the studies used questionable methods [Ding Ding!].
"As far as I can tell, it’s pretty bad science," said McAllister, who is gay.
He said the studies, which followed groups of circumcised and uncircumcised men for two years to track emergence of HIV infections, were poorly done [Ding Ding Ding!].
McAllister said researchers in Africa started tracking infections for both groups immediately after circumcision, even though most men cannot resume sexual activity for about six months following the operation [Ding Ding Ding Ding!].
He also said circumcised men were advised to use condoms while uncircumcised men were not [to the same extent] [Ding Ding Ding Ding Ding!], and men in neither group were asked whether they had engaged in anal intercourse.
The study was only 18 months long. Um... hello? That's six more weeks of sexual activity in the "control" group, which could lead to infection!
But actually it's not even a full 18 months. The antibody test doesn't turn positive for anywhere from 2 weeks to 6 months (typically about 4 weeks) after infection (8,9,10). That reduces the effective duration to 17 months or less, exaggerating the circumcised group's slow start even more.
Let's take a side-by-side look at the experiences of the two groups (this is a dramatic illustration. Control group may not have gone to market):
Intervention: | Control: |
Prepare for surgery | Go home |
Pre-surgery information | Be at home like usual |
Undergo surgery | Go to market as usual |
Post-surgery instructions | Nothing out of the ordinary |
Wound healing for weeks | Normal routine for weeks |
Ponder dire warning to use condom if violating waiting period | No dire warning to ponder |
Warned to abstain from sex for 6 weeks | Enjoy the wives/etc. (8) |
Determine when wound is "fully healed" | Traditional day out fishing |
Post-recovery behavior adaptations | No penis changes to adapt to |
Take the newly healed meat-sculpture out for a first spin | Yet another totally ordinary day |
Enough. This parrot is stone cold dead. In addition to presence or absence of a foreskin, these groups have monumentally divergent experiences. You know, like penis surgery versus no surgery! My tour through this wasteland of Bad Science (BS) is complete.
Should your appetite for investigation be not yet satisfied, have a look at this evaluation from Doctors Opposing Circumcision. Remember, just as we don't care about the motives behind the study leaders, what DOC has to say stands or falls on its own merits. Decide for yourself. You can do it.
Commentary:
1 2 3 4 5 6 7 8 9 10 11
Reasoned Opposing Viewpoint:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18