Reader Comments

Post a new comment on this article

Are the conclusions justified in “Size matters: The number of prostitutes and the global HIV/AIDS pandemic”.

Posted by adamsonmuula on 01 Oct 2007 at 18:12 GMT

RE: Are the conclusions justified in “Size matters: The number of prostitutes and the global HIV/AIDS pandemic”.
Adamson S. Muula
Department of Community Health, University of Malawi, College of Medicine, Private Bag 360, Chichiri, Blantyre 3, Malawi
The paper by Talbott 1 is timely as it comes at a time when the world community and national governments, especially in southern Africa, consider whether to make national health policy male circumcision in an effort to curb the incidence of HIV. This paper certainly has moved beyond the individual level variables but “group or community level” variables to explain the disparity of HIV prevalence among nations. I have several areas of concerns that I wish to discuss:
1. The author has written; “There have been numerous other research papers that have purported to show the correlation between circumcision and reduced HIV/AIDS transmission rates during heterosexual sex.” I found the term “purport” rather unfortunate at this time when three randomized controlled trials (RCTs) carried out in eastern and southern Africa have consistently shown the protective effect of circumcision. Although I recognize that the term purport could have at least two meanings i.e. to imply and the second meaning; to make false appearance of being, the term (purport) is less fitting now after evidence has so far been accumulated. The evidence now is not by implication neither are the claims false. This is not to suggest that the RCTs have answered all our questions about circumcision.
2. Throughout the manuscript, the author reported that they conducted a linear and multiple regression analysis. Without an explicit description of the link function, the reader is persuaded to think that the author used a general linear method since most of data could be analyzed as continuous variables (although after transformation, other link functions can also be used). The linear model : E(Y|x)=f(α + β1x) can just as well be used for general linear regression just as it can be used for logistic regression analysis as long as the link function is appropriately defined. I therefore find the author’s reference to linear regression (instead of general linear methods) and multiple regressions unfortunate. The general linear regression is just a subset of generalized linear regressions. Because of the error reporting linear and multiple regressions, the author did not report whether general linear regression was actually used.
3. Furthermore, I find the conclusion by the author that “This paper provides strong evidence that when conducted properly, cross country regression data does not support the theory that male circumcision is the key to slowing the AIDS epidemic.” Talbott’s analyses looked at, among other issues, whether the proportion of Muslims in a country was associated with HIV prevalence. What needs to be appreciated is that while the proportion of Muslims in a country could be a surrogate indicator of male circumcision, the study did not assess the association between the prevalence circumcision and HIV. We need to exercise a lot of caution in ascribing an association between an outcome and a variable that was not measured. This is not to say that circumcision, just like any other interventions, will likely have different “effects” depending on other variables operational in a setting.
4. Finally, although the effect of circumcision and HIV acquisition has been shown (and is not purported) and there have been explanations as to how circumcision could biologically prevent HIV acquisition, our knowledge on the interaction of circumcision with other social, especially behavioral variables, is still minimal. In the real world outside the RCTs, being a Moslem may mean more than what has been measured in studies to data as has been previously discussed by Madani. 2
The author declares that he has no competing interests.
1. Talbott JR. Size matters: the number of prostitutes and the global HIV/AIDS pandemic.
PLoS ONE. 2007 Jun 20;2(6):e543.
2. Madani TA. Sexually transmitted infections in Saudi Arabia.
BMC Infect Dis. 2006 Jan 10;6:3.