Influenza vaccination in the elderly: Is a trial on mortality ethically acceptable?

The effectiveness of influenza vaccination in the elderly has long been a topic of debate, fueled by the absence of direct evidence from randomized trials on its effect on mortality and the methodological limitations of observational studies pointing this direction. It is argued that new placebo-controlled trials should be undertaken to resolve this uncertainty. However, such trials may be ethically questionable. To bring this discussion forward, we provide a comprehensive overview of the ethical challenges of an influenza vaccine efficacy trial designed to evaluate mortality in the elderly. An important condition in the justification of a trial is the existence of genuine uncertainty in regard to the answer to a research question. Therefore an extensive analysis of the existing levels of knowledge is needed to support the conclusion that an effect of vaccination on mortality is uncertain. Even if a so called "clinical equipoise" status applies, denying a control group vaccination would be problematic because vaccination is considered "competent care" and withholding vaccination could substantially increase patients' risk for influenza and its complications. Given the high burden of disease and already proven benefits of vaccination, such a trial is unlikely to meet the Declaration of Helsinki stating that the importance of a trial must outweigh the risk patients are exposed to. While a placebo-controlled trial in vaccine refusers may be considered, such a trial is unlikely to meet substantial methodological barriers regarding trial size and generalizability. We conclude that a new trial is unlikely to provide for a direct answer, let alone change current policy. At the same time, given the lack of consensus on the ethical acceptability of a placebo-controlled trial on the effect of influenza vaccination on mortality in the elderly, we invite researchers considering such trials to address the ethical challenges discussed in this manuscript.