The ethics of cluster randomized trials: response to a proposal for revision of the Ottawa Statement

Cluster randomized trials are commonly used to evaluate public health, knowledge translation, and health service interventions. Cluster trials raise novel ethical issues, however, and the Ottawa Statement on the Ethical Design and Conduct of Cluster Randomized Trials (2012) provides researchers and research ethics committees with needed guidance. In this journal, van der Graaf et al. reflect on the Ottawa Statement and propose three revisions. In this paper, we respond to each of these proposed revisions.

First, van der Graaf et al. argue that patients who are merely indirectly affected by study interventions ought nonetheless to be considered research participants. We disagree. So long as the practice change is evidence based and the physician continues to make individualized judgments regarding patient care, patient liberty and welfare interests are not substantially affected. Second, although they agree that health providers who are targeted are research participants, they argue that such providers ought to be treated differently and should not be allowed to withdraw from a study too easily. In our view, this position fails to weigh adequately the potential for coercion and harms faced by employees in research. Third, they argue that the potential for bias may require blinding participants to allocation and study interventions in the consent process of a cluster trial. We agree on this point and support this approach in a limited set of cases.

While we reject two of van der Graaf et al.'s proposed revisions, we agree that further guidance on informed consent and study bias is needed.