Related Experiment Videos
Changes in beliefs identify unblinding in randomized controlled trials: a method to meet CONSORT guidelines
Judy R Rees1, Timothy J Wade, Deborah A Levy
1Department of Community and Family Medicine, Center for Environmental Health Sciences, and Norris Cotton Cancer Center, Dartmouth Medical School, Lebanon, NH 03756, USA. judith.rees@dartmouth.edu
Contemporary Clinical Trials
|April 20, 2005
Summary
Assessing participant beliefs, not their accuracy, is key to detecting bias in blinded trials. Comparing changes in belief over time, rather than a single snapshot, offers the best method for evaluating unblinding.
Area of Science:
- Clinical Trials
- Methodology
- Research Integrity
Background:
- Double-blinded trials are crucial for minimizing bias in research.
- Current methods for assessing blinding effectiveness, such as participant accuracy in identifying interventions, may be insufficient.
- The CONSORT guidelines emphasize reporting blinding success, but clear evidence standards are lacking.
Purpose of the Study:
- To investigate the effectiveness of sequential participant belief assessment in detecting unblinding in randomized controlled trials.
- To determine whether participant beliefs or their correctness are more indicative of potential bias.
- To identify the optimal method for assessing blinding effectiveness in clinical trials.
Main Methods:
- Conducted the Water Evaluation Trial, a randomized controlled trial in northern California in 1999.
- Sequentially interrogated participants about their perceived intervention assignment (active or placebo) using multiple questionnaires.
- Analyzed intergroup differences in the correctness and change of participants' beliefs over time.
Main Results:
- Participants across both groups showed a tendency to believe they received the active intervention.
- A statistically significant intergroup difference in belief correctness was observed early in the trial.
- Sequential belief assessments revealed significant group differences in belief patterns not detected by single measures, with 31% of participants changing their beliefs during the trial.
Conclusions:
- Participant beliefs, not their accuracy, are more directly linked to potential bias, especially for self-reported outcomes.
- Assessing the change in beliefs from the start to the end of a trial provides a more sensitive measure of unblinding than cross-sectional assessments.
- An intergroup comparison of the change in beliefs is the optimal method for assessing unblinding in randomized controlled trials.