The American Journal of Preventive Medicine has recently published a discussion on the “external validity” of the three African randomized control trials that examined circumcision as a preventive measure for HIV infection. The nutshell version of this discussion is that clinical conditions that show promise do not necessarily translate into a real world net benefit.
Recent editorials have asked the global health community to scale up male circumcision for HIV prevention in regions with HIV epidemics following the publication of three randomized controlled clinical trials (RCCTs) in Africa (in South Africa, Uganda, and Kenya).1–5 One editorial concluded: “The proven effıcacy of MC [male circumcision] and its high cost-effectiveness in the face of a persistent heterosexual HIV epidemic argues overwhelmingly for its immediate and rapid adoption.”6 This “Current Issue” review questions not the internal validity of the studies, but their external validity, an issue that has been discussed more generally in two commentaries,7,8 an editorial,9 and a systematic review of research on prevention trials10 in this journal. External validity is the issue that questions the generalization from the RCCT results to a policy of “immediate and rapid adoption” of circumcision of men across Africa.
Five dimensions of external validity should be weighed before the global health community can determine that male circumcision is a widely effective, cost-effective, or cost benefıcial use of resources, as well as an effective and safe method for controlling the HIV epidemic in Africa. These trials provide a case illustration of how a policy might be adopted without due consideration of external validity in experimental trials that appear to have established internal validity for a short-term reduced risk of infection.
Recommending mass circumcision by generalizing from the particular RCCTs to the diverse populations of Africa highlights problems of external validity identifıed in several areas of preventive medicine and public health research. Studies published since the RCCTs show that (1) male circumcision is not correlated with lower HIV prevalence in some sub-Saharan populations 14,15; (2) circumcision is correlated with increased transmission of HIV to women 20; and (3) male circumcision is not a cost effective strategy.17,28 This new evidence warrants caution and further study before recommending circumcision campaigns. In addition, ethical considerations, informed consent issues, and possible increase in unsafe sexual practices from a sense of immunity without condoms must be weighed.
The global health community understands that the most important modifıable factor in sexually transmissible HIV is human behavior. 40 The policy questions to be considered are not whether a link exists between male circumcision and reduced rates of HIV infection, but, rather, whether mass circumcision is (1) an ethical and safe public health choice, and (2) the most cost-effective use of limited resources.