Methodology and predictive accuracy of the prospective preference assessment for randomized trial enrollment: A literature review.

Objective: The success of a randomized controlled trial (RCT) depends, in part, on the willingness to participate (WTP) of eligible subjects. Prospective preference assessments (PPA) offer insights into future RCT enrollment by evaluating eligible individuals' WTP in the RCT. We sought to summarize the methods and findings of published PPAs and to compare the WTP estimates of published PPAs to the actual enrollment rates from these trials. Design: We performed a systematic literature search using PubMed for studies that assessed eligible participants' WTP in a hypothetical trial. We abstracted sample size, proposed interventions, WTP response options, WTP percentage, presence of qualitative analyses, and assessment of trial feasibility from each publication. We defined WTP as a response of "probably" or "definitely" willing in the PPA; in sensitivity analyses, we included only "definitely" willing responses. We searched for registered and published RCTs connected to each PPA and extracted enrollment data from the corresponding trial publications where available. Results: We identified 40 PPAs. The median WTP across all PPAs was 54.9 ​% (range: 13 ​%-92.4 ​%), and 42.1 ​% (range: 7 ​%-90.2 ​%) when including only "definitely willing" responses. We found ten registered RCTs; five are completed and one is ongoing. Four PPAs with a RCT provided both "definitely willing" and "total willing" estimates. In three of these four RCTs, the actual enrollment fell between the "definitely willing" and "total willing" PPA estimates. Conclusion: "Definitely willing" and "total willing" findings in a PPA may provide useful upper and lower boundaries on participation for future trial planning.