Considerations for open-label randomized clinical trials: Design, conduct, and analysis

Abstract

Randomization and blinding are regarded as the most important tools to help reduce bias in clinical trial designs. Randomization is used to help guarantee that treatment arms differ systematically only by treatment assignment at baseline, and blinding is used to ensure that differences in endpoint evaluation and clinical decision-making during the trial arise only from the treatment received and not, for example, the expectation or desires of the people involved. However, given that there are times when it is not feasible or ethical to conduct fully blinded trials, we discuss what can be done to improve a trial, including conducting the trial as if it were a fully blinded trial and maintaining confidentiality of ongoing study results. In this article, we review how best to design, conduct, and analyze open-label trials to ensure the highest level of study integrity and the reliability of the study conclusions.