点击显示 收起
Department of Obstetrics, Gynecology and Reproductive Sciences, University of California, San Francisco, San Francisco
As has been discussed elsewhere [1, 2], a 3-arm microbicide trial, comprising a microbicide arm, a placebo arm, and a "condom-only" arm (and in which all 3 groups are equally counseled to use condoms), allows for all comparisons: the placebo arm controls for potential behavioral differences that might result from the use of an unblinded control, and the condom-only arm allows a comparison with the current "real world" standard of prevention. However, the interpretation of the results of a trial are more challenging when it contains 3, rather than 2, treatment arms. The use of multiple arms is also more expensive and more difficult to standardize across sites.
In light of these considerations, the choice of a control group should be directly linked to the primary objective of the study. Comparisons with a placebo arm primarily address the research objective of whether, when a study controls for behavior, an active agent is more effective than a placebo. As has been noted elsewhere [1, 2], it is not unreasonable to assume that a placebo might be protective. In this case, if there were no unblinded control, a null or insignificant effect would likely resultas could have occurred in the COL-1492 study [3]thus eliminating from further consideration both the active and inactive products. In contrast, comparisons with a condom-only arm address the research objective of finding something that is better than (or at least as good as) the only method known to be effective in the prevention of heterosexual transmission of HIVthat is, consistent use of male condoms.
As is suggested by Drs. Stein and Susser [4], excellent rates of recruitment and retention are the hallmarks of a successful trialand that, if it were the case that the use of an unblinded-control arm resulted in differential retention rates by the study arms, this difference could severely bias the interpretation of study results. Thus far, however, the possibility that retention will be worse in a condom-only arm has not been borne out in similarly designed trials (e.g., [5]). In addition, strategies such as offering a variety of condom types and making it clear that the control arm is a "delayed" interventionmeaning that participants in the control arm will receive the study product once (and if) its effectiveness is determinedgo a long way toward maintaining good retention rates.
Although Drs. Stein and Susser also argue for a "proof of concept" to detect a difference between preparations (i.e., active agent vs. placebo), in the face of a relentless epidemic for which few successful strategies for prevention have been identified it is far more important to gauge the success of an intervention compared with the success of known, effective real-world prevention interventions. If scarcity of current resources dictates that we can have only 2 study arms, it is essential to maximize our chances of detecting a difference. Given a choice between (1) differentiating the effect of an active ingredient versus that of an inert vehicle of delivery or (2) finding an intervention (be it an active ingredient, placebo, or behavior change) that works better than current prevention interventions, I would choose the latter.
References
1. Stein ZA, Myer L. Susser M. The design of prophylactic trials for HIV: the case of microbicides. Epidemiology 2003; 14:803. First citation in article
2. Padian N. The design of prophylactic trials for HIV. Epidemiology 2003; 14:834. First citation in article
3. Van Damme I, Ramjee G, Alary M, et al. Effectiveness of COL-1492, a nonoxynol-9 vaginal gel on HIV-1 transmission in female sex workers: a randomized controlled trial. Lancet 2002; 360: 9717. First citation in article
4. Stein ZA, Susser MW. Control groups in microbicide trials: in defense of orthodoxy . J Infect Dis 2005; 191:13767 (in this issue). First citation in article
5. Roddy RE, Zekeng L, Ryan KA, Tamoufe U, Weir SS, Wong EL. A controlled trial of nonoxynol 9 film to reduce male-to-female transmission of sexually transmitted diseases. N Engl J Med 1998; 339:50410. First citation in article