Male circumcision and HIV infection: the ethics of a landmark study

In an earlier post, I mentioned some of the preliminary and partial results of a randomized controlled trial of the use of male circumcision to reduce risk of HIV infection in South Africa (the ANRS 1265 trial). Those results were presented at the International AIDS Society in Rio last July, and an article further detailing the study methods and findings have just appeared in the November issue of PLoS Medicine. In an unusual move, the editors of PLoS Medicine have seen fit to write an editorial to explain why it choose to publish the trial results – after having them combed over by six reviewers -- and the issue also includes a commentary by the head of one of the IRBs that originally approved the protocol. It is safe to say that the study is regarded as ethically hot, and its publication is unlikely to cool it down.

ANRS 1265 was a randomized controlled trial with one intervention arm (1640 men who received circumcisions after recruitment) and one control arm (1654 uncircumcised men). During the course of the study, initially planned to take 21 months, 20 men in the intervention arm became HIV positive as opposed to 49 in the control arm. This translates into a 60% protection rate – a rate comparable to an efficacious vaccine and achieved by means of a relatively simple surgical intervention. For decades, there have been observational studies and meta-analyzes suggesting the protective qualities of circumcision against female-to-male HIV transmission, but this study seems to finally offer experimental proof. The importance of the findings is unquestionable.

But where are the ethical flashpoints? There seem to be at least two related ones. First, while it would seem to be logical to limit the participants to HIV negative men, the researchers choose not to use a positive HIV test as an exclusion criterion, reasoning that such a criterion would risk subjecting prospective participants to stigmatization. Thus the presence of some HIV positive men in the trial was virtually guaranteed. Second, while the researchers drew blood to test for syphilis and HIV, interviewed participants in detail about their sexual activities, offered HIV counseling, and encouraged the men to be HIV tested at local clinics, they did not inform participants of their HIV status if they did not wish to be informed. The study design, in fact, called for researchers to be blinded as to the HIV status of the participants. So on the one hand, the trial presupposed that some of these at-risk men (both circumcised and uncircumcised) would be or become HIV positive during the trial, but on the other hand the trial left no room for a ‘duty to warn’ those who were or became HIV positive, even if the participant revealed in interviews that he was engaging in unsafe sex. As the article puts it:

They [the investigators] considered it unethical to inform participants of their HIV status without their permission, even if they thought that participants should be aware of their HIV status.

Some might say that in the name of avoiding stigmatization and respecting autonomy, the researchers were studying (at least in some cases) HIV positive persons engaging in harmful practices without protecting third parties. And why did the investigators need to be blinded in regard to the HIV status of the men in the first place? Couldn’t the same results be generated by an ethically less controversial trial design?

Readers are invited to read the PLoS article and draw their own ethical conclusions.