Patient acceptability of the modified Zelen approach to randomized trials: a survey of the CAPS THA cohort

Randomized controlled trials (RCTs) are the gold standard for evaluating interventions, but when comparing established care pathways, informed consent processes may alter usual preference-related behavior, thereby introducing bias. Zelen designs randomize participants before consent to reduce these effects. The Patient Satisfaction and Costs Associated with Total Hip Arthroplasty (CAPS-THA) trial used a modified Zelen approach to compare inpatient vs outpatient discharge after total hip arthroplasty (THA). Participants were prerandomized and initially consented to an observational study of their assigned discharge pathway; randomization and full study details were disclosed after final follow-up. This survey aims to understand how participants feel about the modified Zelen design and its justification and whether the deception affected their trust in medical research.

We conducted an anonymous cross-sectional survey of all available CAPS-THA participants (n = 290), distributed by email (n = 193) or mail (n = 97). Items assessed understanding of the design, perceived ethical acceptability, responses to disclosure, trust in medical research, and willingness to participate in future studies using similar designs. Descriptive statistics summarized responses.

Of 290 eligible participants, 235 completed the survey (81% response rate). Mean age was 70.1 ± 10.2 years, 48% female, and 64% (150/235) reported postsecondary education. Ninety percent (210/235) reported a preference toward one comparison group. If informed of both pathways upfront, 36% (84/233) indicated that they would have wanted to learn more about the alternative pathway, suggesting potential for preference-related behavior and satisfaction effects under a conventional RCT. After disclosure, 76% (177/233) agreed that the delayed consent design was justified, and 91% (208/229) agreed that the study was conducted ethically with patients' best interests in mind. Trust in medical research was reported as unchanged (57%) or increased (37%) by 94% (219/233) of respondents. Seventy-one percent (163/229) expressed strong willingness to participate in a similar study in the future.

Nearly all participants viewed the modified Zelen approach as ethically acceptable and reported no decrease in trust following disclosure. These findings support the responsible use of Zelen designs, paired with a robust debriefing and meaningful postdisclosure choice, when preference-related bias may threaten trial validity.

Randomized controlled trials are the best ways to compare medical treatments because they reduce biases. However, if patients in a research setting are told about another treatment option that they would not usually hear about in routine care, they may start to prefer that option or feel disappointed with their assigned treatment. This could change their health-related behavior in ways that would not normally occur, potentially biasing the results and limiting their usefulness for guiding routine care. One way to reduce this problem is to use a modified Zelen design, where patients are randomly assigned to a treatment group but not told about the randomization or alternative treatment being studied until the end of their study participation. This approach may improve the quality of the research, but it also raises concerns about honesty, choice, and trust. In the CAPS-THA study, patients having hip replacement surgery were randomly assigned to either stay in the hospital for one to two nights or go home the same day as their surgery. At first, patients were not told that they were part of a randomized trial, only that the researchers wanted to learn more about their recovery. Once participants completed the study, they were informed they had been randomized, given details about the purpose of the randomized trial, and why the research team decided to withhold this information until study end. During this discussion participants were asked for their consent to include their data in the results of the trial. To understand how patients felt about the Zelen approach, we surveyed former CAPS-THA participants. Of 290 eligible participants, 235 completed the survey. Most (91%) felt that the Zelen approach was justified and ethical. Almost all said that their trust in medical research stayed the same or increased after the study was explained to them. Many also said that they would be willing to take part in a similar study in the future. These findings suggest that, when there are good reasons for using a modified Zelen design, such as situations where patient preferences could affect the results of the study, patients felt it an acceptable design.