DMPA and HIV: why we need a trial.

Recent articles by Ralph et al. [1,2] and Gollub and Stein [3], as well as an accompanying commentary by Jones [4] in the current issue of Contraception, have challenged the concept of conducting a randomized trial to determine whether injectable depot medroxyprogesterone acetate (DMPA) increases the risk of HIV acquisition. They all raise concerns about the evidence, the methodology, the ethics, the feasibility and the return on investment of such an undertaking. While those authors acknowledge the uncertainty of the evidence regarding the effect of hormonal contraception, particularly DMPA, on increased HIV risk, they argue that the observational data are of sufficiently high quality to inform women of the potential risks and recommend against investing in a randomized trial to answer the question. The widespread use of DMPA in areas of high HIV incidence means that the issue of whether DMPA increases HIV susceptibility relative to other methods is of major global health importance and requires the strongest evidence possible to resolve any uncertainty. As members of the multiorganizational collaborative team called ECHO (Evidence for Contraceptive Options and HIV Outcomes), we are currently developing a protocol for a randomized trial of contraception and HIV acquisition. In this Commentary, we consider the evidence, ethics and feasibility of a possible experimental design. First is the quality of the current observational evidence. In February 2012, a World Health Organization (WHO) expert panel reviewed all the available studies on hormonal