ASSESSING NON-CONSENT BIAS WITH PARALLEL RANDOMIZED AND NONRANDOMIZEDCLINICAL-TRIALS

In some randomized clinical trials, a large proportion of patients eli gible for randomization may withhold consent: to be randomized. When t he subjects in the randomized trial differ from the eligible populatio n with respect to characteristics that are associated with the magnitu de or the treatment effect, there may be non consent bias, i.e., the t reatment effect for those in the randomized trial may not reflect the treatment effect for the eligible population. In response to this prob lem, some investigators have conducted, in addition to the randomized trial, a separate nonrandomized but otherwise identical trial consisti ng of those patients who are eligible for randomization, but instead c hoose their own treatment. Observed baseline covariate data can be use d to adjust for differences between the randomized population and the eligible population when estimating the treatment effect for the eligi ble population. After adjusting, different outcomes for the randomized versus nonrandomized treated groups and/or the randomized versus nonr andomized control groups reflect the presence of hidden non-consent bi as resulting from differences between the trial population and the eli gible population with respect to unobserved covariates. A sensitivity analysis can display how hidden non-consent bias can account for an im balance in the treatment groups with respect to an unobserved covariat e. PI parallel randomized and nonrandomized trial which compares adeno idectomy versus medical treatment for children with recurrent otitis m edia [Paradise et al. Efficacy of adenoidectomy for recurrent otitis m edia in children previously treated with tympanostomy-tube placement. J Am Med Assoc 1990; 263: 2046-2073] is used as an illustration. (C) 1 991 Elsevier Science Inc.