Refusal rates and waivers of informed consent in pragmatic and comparative effectiveness RCTs: A systematic review

Background: Pragmatic and comparative effectiveness randomized controlled trials (RCTs) aim to be highly generalizable studies, with broad applicability and flexibility in methods. These trials also address recruitment issues by minimizing exclusions. The trials may also appeal to potential subjects because of lower risk and lower burdens of participation. We sought to examine rates of refusal and uses of waivers of informed consent in pragmatic and comparative effectiveness RCTs. Methods: A systematic review of pragmatic and comparative effectiveness RCTs performed wholely or in part in the United States and first published in 2014 and 2017. Results: 103 studies involving 105 discrete populations were included for review. Refusal data was collected for 71 RCTs. Overall, studies reported an average rate of 31.9% of potential subjects refused participation; on an individual basis, 38.4% of people asked to take part refused at some point during recruitment. 23 trials (22%) were performed, at least in part, with a waiver of informed consent, 7 (30%) of which provided any form of notice to subjects. Conclusions: Overall refusal rates for pragmatic and comparative effectiveness RCTs appear roughly the same as other types of research, with studies reporting about a third of people solicited for participation refuse. Moreover, informed consent was waived in 22% (95% Binomial exact Confidence Interval 13.9-30.5%) of the trials, and further study is needed to understand when waivers are justified and when notice should be provided.

Automated summary

The study found that overall refusal rates for pragmatic and comparative effectiveness RCTs are similar to other types of research, with approximately one-third of potential subjects refusing to participate. About 22% of the trials waived informed consent, indicating a need for further research to determine when waivers are justified and when notice should be provided.