In their recent Personal View, Claire Tully and colleagues1 explore the ethical implications of evaluating experimental vaccines through randomised controlled trials and comment on the stepped-wedge study design as an adaptation to minimise risk.1 The authors identify the lack of need for a placebo group as the main advantage of the stepped-wedge design; however, many other benefits of this method are worth mentioning. In a stepped-wedge design trial, such as the Sierra Leone Trial to Introduce a Vaccine against Ebola (STRIVE) trial of health-care workers in Sierra Leone,2 the Ebola vaccine is sequentially rolled out to participants in clusters, such as clinics or hospitals, throughout several time periods. By the end of the study, all participants will have received the intervention.3 When the intervention is expected to confer a large benefit, a stepped-wedge design mitigates the ethical dilemma of non-treatment, such as in the case of a parallel control group, or withdrawal of treatment as would occur in a standard crossover study. We agree with Tully and colleagues that a consensus on vaccine efficacy would be reached more quickly with a randomised controlled trial because stepped wedge trials take longer to do than randomised controlled trials.1 However, this fact is less problematic when a shorter time between intervention and outcomes is expected, such as in the case of an Ebola outbreak. Another benefit of this study design is increased statistical power achieved by doing several comparisons, both within and between groups.4 Although a stepped-wedge trial has no placebo group, direct comparisons are still made between groups receiving the treatment and those who are not, and groups can act as their own controls before and after the intervention, so fewer units are needed. Logistical and financial considerations in resource-limited areas of west Africa pose barriers to the initiation of Ebola vaccine trials. A stepped-wedge design would eliminate the need to simultaneously introduce costly interventions and provide medical staff and equipment to study sites. Finally, a stepped-wedge design trial would assess the intervention effectiveness in the context of the community and take into consideration temporal changes such as seasonality. Stepped-wedge trials can be randomised, whereby the order of administering the intervention is established randomly, or observational, thereby increasing flexibility in design. A downside is that several collection points would be needed; however, in group-sequential designs the data would be collected routinely and analysed regardless.3 We declare no competing interests.
Piszczek, J., & Partlow, E. (2015, July 1). Stepped-wedge trial design to evaluate Ebola treatments. The Lancet Infectious Diseases. Lancet Publishing Group. https://doi.org/10.1016/S1473-3099(15)00078-X