A systematic review of methods for specifying the target difference in randomised controlled trials (delta review)

Abstract

Background: Randomised controlled trials (RCTs) are widely accepted as the preferred study design for evaluating healthcare interventions. When determining sample size, a (target) difference must typically be specified. This provides reassurance that the study will be informative i.e. should such a difference exist, it is likely to be detected with the required statistical precision. From both a scientific and ethical standpoint, selecting an appropriate target difference is of crucial importance; too large or small a study is arguable unethical, wasteful and potentially misleading. While a variety of methods have been proposed in the medical field to specify a target difference, their relative merits are unclear as is whether there are methods in the non-medical field which could be used. Aim: To systematically review of medical and non-medical literature to identify methods for specifying the target difference in a randomised controlled trial. Methods: Electronic searches of biomedical and non-medical databases were performed. Clinical Trial textbooks were also reviewed. Titles and abstracts were screened prior to full-text assessment. Studies which reported a method which could be used to specify the target difference were included. Clinical trial textbooks were also reviewed. Results: The search identified 11,485 potentially relevant studies; 1,434 were selected for full-text assessment. Seven methods were identified: anchor, distribution, health economic, opinion-seeking, pilot study, reviews of the evidence base and standardised effect size (SES). The anchor, distribution and SES methods were most commonly used. Discussion: A variety of methods were identified though each had important variations in application. While no single method provides a perfect solution to a difficult question, methods are available to inform specification of the target difference and should be used whenever appropriate.