ABSTRACT
Unequal allocation is often employed in clinical trials with two or more treatment arms based on ethical or power considerations [35]. The detailed reasons for unequal allocation uncovered in the review by Dumville [7] include improving the acceptability of the trial by increased allocation to the experimental group, increasing the patients’ exposure to the treatment, mitigating the higher dropout in one of the groups, increasing power for secondary analyses, reducing the cost of the treatment by placing fewer patients on the more expensive treatment, and the need to have more patients on specific treatment in the second phase of the trial. McEntegart and Dawson [31] further commented that optimizing efficiency of procedures with pairwise comparisons is another reason they see unequal allocation used in dose-response studies. Unequal allocation is also common in adaptive design studies: dose-ranging studies, multi-stage studies, and studies with sample size re-estimation [12, 16].
