Browse by author
Lookup NU author(s): Jenni Hislop,
Professor Luke ValeORCiD
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
Background: To determine the sample size for a randomised controlled trial (RCT), a (target) difference between treatments is typically specified, which the RCT is designed to detect. This provides reassurance that the study will likely detect this difference (where it exists) with the required level of statistical precision. One common approach is to specify a difference that is regarded as being clinically important. Determining an appropriate sample size by specifying the target difference is of crucial importance as too large or small a study could be unethical, wasteful or lead to a potentially misleading conclusion. Within the medical field, various methods have been proposed to formally determine target differences, but their relative merits are unclear, and the availability and suitability of alternative methods from non-medical fields is unknown. Objectives: To systematically review the medical and non-medical literature to identify methods to specify the target difference for use in RCTs. Methods: Electronic searches of biomedical and non-medical databases were performed. Studies reporting a method that could be used to specify the target difference were included. Titles and abstracts were screened prior to full-text assessment. Results: The search identified 11,386 potentially relevant studies; 1,274 were selected for full-text assessment. Seven methods were identified: anchor, distribution, economic evaluation, effect size, opinion-seeking, pilot study and reviews of the evidence base. Anchor, distribution and effect size methods were most commonly used. Discussion: A variety of methods were identified, though each had important variations in application. Some methods require strong assumptions to be made before application to an RCT setting. Further research evaluating usage and optimal implementation is needed.
Author(s): Hislop J, Adewuyi T, Vale L, Harrild K, Altman D, Briggs A, Fayers P, Norrie J, Ramsay C, Buckley B, Harvey I, Cook J
Publication type: Conference Proceedings (inc. Abstract)
Publication status: Published
Conference Name: Health Technology Association International (HTAi) Annual Meeting
Year of Conference: 2012
Pages: 174 abstract no. 790
Publisher: Elsevier Doyma
Library holdings: Search Newcastle University Library for this item
Series Title: Gaceta Sanitaria