Date Published: August 7, 2018
Publisher: Public Library of Science
Author(s): Rebecca Kahn, Annette Rid, Peter G. Smith, Nir Eyal, Marc Lipsitch
Abstract: In a Policy Forum, Marc Lipsitch and colleagues discuss trial design issues in infectious disease outbreaks.
Partial Text: In outbreaks of emerging infectious diseases for which no proven efficacious vaccines exist but investigational vaccines have been developed, it is important both to rapidly test the investigational vaccines and, if effective, to deploy them. Following the 2014–2016 Ebola epidemic, the World Health Organization (WHO), the Coalition for Epidemic Preparedness Innovations, and other bodies committed to developing investigational vaccines for emerging infectious diseases [1,2]. They aim to evaluate them for immunogenicity and safety, so that promising candidates will be available for efficacy testing and possible deployment when an epidemic occurs.
When designing and implementing randomized efficacy trials for investigational vaccines after safety and immunogenicity data have been collected (in Phase 1 and 2 trials), some key choices must be made. In the current regulatory system, randomized trials are considered the gold standard and, except in rare circumstances, have been required for vaccine licensure [6,7]. We restrict our scope to randomized trials of a single vaccine against an emerging infectious disease for which no effective vaccine exists. We assume that all participants, whether in the intervention or control group (if any), will have access to the best currently available other preventative measures (e.g., information on how to prevent infection).
Table 1 summarizes the major designs that have been used or proposed for vaccine trials. Some have not been employed in epidemic settings.
We argue that individually randomized trials with a placebo control should be the default strategy for evaluating investigational vaccines during epidemics. Placebo-controlled trials typically maximize the social and scientific value of the trial, and objections to using placebo, such as a duty to “rescue” individual participants—with an unlicensed investigational vaccine candidate of unproven efficacy—are rarely persuasive. Depending on the pathogen as well as statistical, fairness, and feasibility considerations, trial participants may be selected either from the general population or from a group at high risk of exposure to infection. Starting the trial at approximately the same time for all participants, in a parallel rollout, will minimize the time required to obtain a result and maximize social and scientific value. If resources are limited and/or incidence is spatiotemporally variable, a stepped rollout may be necessary, in which recruitment to the trial is staggered over time.