Wednesday, December 5, 2012

Problems with RCT's & Circumcision

What you'll find those with a pro-circumcision agenda doing is that they cite that only data derived from RCT research is valid in determining the validity of any circumcision research outcomes. ie. if its not RCT derived it is not valid.

While RCT's tend to be the gold standard of research studies, they are not without fault, and can lead to very misleading future predictions, particlularly when the assumptions they are based on are incorrect.

1. The first incorrect assumption of the African RCT circumcision trials was blaming the male foreskin as the cause of HIV infections.  Because some African population groups of non-circumcised men, had higher HIV infection rates than circumcised populations, the male foreskin was seen as the predominant cause of the differences in HIV infection rates.  What the researchers ignored was the fact that many African circumcised populations had higher hiv infection rates than non-circumcised populations.  Therefore the correct assumption about the cause of sexually transmitted  HIV infections has to be behavioural, eg. unsafe sex with multiple sex partners.

2. The second incorrect assumption was that circumcision confers an absolute lifetime protection against HIV infection, and that an 18 to 24 month trial was enough to prove this.  WRONG, the studies needed to have been longitudinal, over a complete lifetime, to determine the validity of circumcisions protective effect over a lifetime. If as seems to be the case that circumcision in a highly controlled environment with extra resources, than are normally available in aAfrica, can confer a small delay in time before infection occurs, then the assumption of life-time protection is also WRONG.  The desperation of researchers, and particularly the WHO in trying to do something about the HIV epidemic in Africa, has led to premature endoresment of something the research did not find or prove, ie. lifetime protection against HIV.  The RCT's did not prove a lifetime protective effect from circumcision, there is no evidence from the RCT';s of this claim. 

3. The third incorrect assumption is that men wont change their behaviour after circumcision, believing they are protected from HIV, and therefore disinhibiting their behaviour and practicising less safe sex behaviour.  Recent evidence from Africa confirms behavioural disinhibition is already occurring.  Therefore the assumption that men wont change their behaviour which will increase risk is also wrong.

The WHO has endorsed medical circumcision in Africa becuase it is claiming circumcision will reduce millions of infections, but this is made on an assumption that is not proven and most probably incorrect.  If you just change the assumptions slightly, ie. that circumcision may increase behavioural dis-inhibition and increase in unsafe sex behaviours, then the reduced infections become meaningless.  If as most data shows, that circumcision does not provide a lifetime protective effect against infection, then the protective effect of circumcision is incorrect, and again the predictions of reduced infections are meaningless.

Lets look at some research evidence which disputes the lifetime protective effect of circumcision. For instance,there is much data from around the world that shows some circumcised populations have higher HIV infection rates, than non-circumcision populations.  The EU where circumcision is uncommon has much lower rates of HIV than the USA where circumcision is very common. What this tells us, is that over a lifetime, whatever protective effect circumcision may have, is insignificant over a whole life, and meaningless in populations where circumcised groups have higher HIV infections than non-circumcised pouplations.

Therefore if you set up RCT's to last 18 to 24 months, end them early, and resource them in ways that will never be available in real life, and the reality is that the protective effect of circumcision is only short term, lets say 12 to 24 months, & not effective over a lifetime of 30 to 50 years of sexual activity as claimed, And you advertise and convince men that the protective effect of circumcision is for a lifetime, then you have made a grave public health error. What you have done is created a false sense of securty, and potentially created an environment where behavioural disinhibition occurs, where circumcised populations begin to practice less safe sex, & you end up with a public health disaster with higher HIV infections.

RCT's as with Computers, are only as good as the theoretical assumptions they are based on, or the data that is put into them.  Its Basically garbage in & garbage out!!!!!!!!!!

4 comments:

  1. Thanks for this posting this. I would argue that the African RCTs aren't true RCTs at all; they really don't fit the criteria. It amazes me that the "results" of these "studies" are accepted as gold despite all the ethical and scientific flaws.

    ReplyDelete
  2. In my opinion, those studies are inherently unethical. I'm not presently in the mood to write... but when I am, sometimes I, also, write blog articles. All who are reading this are invited to read Karl Hegbloom Is in. There are several articles there regarding the "circumcision" issue.

    The word "circumcision" is a deprecated euphemism for the atrocity that is more accurately referred to as "genital mutilation." It is not a "rite." It is a crime.

    ReplyDelete
  3. 1. We can never be sure if the sexual behaviours of the circumcised treatments and intact controls are broadly similar. In one respect, they cannot be, because the circumcised men are told to abstain from sex for 6 weeks after getting cut. This would not matter very much, if the RCTs lasted 5-10 years. In fact, they were cut short after 18-24 months.

    More generally, it cannot be said that we truly understand the etiology of AIDS in eastern and southern Africa. When questioned about their sexual practices, Africans tend to give answers that they believe the investigators want to hear.

    2. It is quite possible that all, if anything, circ achieves in Africa is to delay the inevitable. We cannot know whether this is the case because the RCTs were cut short.

    3. On the relation between circumcision and relaxed attitudes towards unsafe sex, read:
    http://en.wikipedia.org/wiki/Risk_compensation

    There is also NO research on whether circumcised men are more less willing to use condoms. If less, circumcision could prove a disaster over the long haul.

    ReplyDelete
    Replies
    1. There may be no official research, however, several reports out of Africa have stated that recently circumcised men, and indeed have unsafe sex with multiple sex partners in the belief, they are now protected from HIV, I have blogged about this also, with references to original media reports

      Delete