Clinical Analysis

Unsolved mysteries of the African circumcision trials

VMMC advocates conducted surgical trials on Africans to establish the efficacy of male circumcision as an HIV-prophylaxis.* The World Health Organisation accepted their findings as evidence of a 50–60% protective effect of male circumcision against female-to-male HIV transmission. A Cochrane review accepted them as evidence of a 38–66% protective effect. Others, including our organisation, have questioned the findings themselves on the basis of the following weaknesses:

  • To qualify informed consent, circumcised subjects were told that at best, the procedure could only be partially protective, and that additional protection would be necessary to prevent HIV infection. Did the trials measure the protective effect of male circumcision or sex education?
  • Why was the 50–60% protective effect attributed to circumcision alone when additional sex education was provided to circumcised subjects?
  • The uncircumcised control group was given a head-start to contract HIV (a “lead-time” bias) while the circumcised group was instructed to abstain from sex for at least six weeks to allow wounds to heal. Why wasn’t the clock started at the same time for both groups?
  • The trials were terminated early for ethical reasons. Could premature termination have amplified the effect of the lead-time bias? Might the delta between the groups have diminished over time?
  • One-third of subjects in the South African trial reported having no unprotected sex throughout the trial period (Green et al., 2008). Were all infections (hetero)sexually transmitted as assumed? Why was there no control for common bloodborne vectors?
  • Why were 700 subjects lost to follow-up?
  • The trial authors included previously established circumcision advocates. Could researcher bias have influenced the results?
  • Sexual behaviour differs by culture and region. Are behavioural trials on men in three high-risk districts applicable to all boys and men in fourteen countries? (NOTE: USAID statistics from the start of the VMMC campaign show that HIV rates were highest among circumcised men in the majority of countries surveyed, half of which were subjected to VMMC nonetheless.)
  • African HIV rates have increased from the start of VMMC roll-out. Why are the trial findings not manifesting in the real world?

Male-to-female HIV transmission is more common than female-to-male transmission, and thus has a greater impact on the epidemic. So what of the “buried” male-to-female trial?

  • The male-to-female trial found that male circumcision increased women’s HIV risk by the same margin that it is claimed to reduce men’s risk. Why is male circumcision’s alleged “indirect benefit” to women prioritised over this established direct risk to women?
  • The male-to-female trial was terminated early because too many women were becoming infected through their partners’ circumcision wounds. If this effect occurred in a clinical trial setting with optimal counseling, what is happening in real world settings?
  • Why aren’t women informed of this critical finding?
  • What does this significant increase in women’s HIV risk from male circumcision mean for the epidemic at large?

*The female-to-male trials were conducted in Orange Farm, South Africa; Rakai, Uganda; and Kisumu, Kenya. The male-to-female trial was conducted in Rakai.

What’s a mucous membrane anyway? Click here to see our analysis of VMMC science.