A systematic review of the use of an expertise-based randomised controlled trial design

Jonathan A. Cook, Andrew Elders, Charles Boachie, Ted Bassinga, Cynthia Fraser, Doug G. Altman, Isabelle Boutron, Craig R. Ramsay, Graeme S. MacLennan

Under a conventional two-arm randomised trial design, participants are allocated to an intervention and participating health professionals are expected to deliver both interventions. However, health professionals often have differing levels of expertise in a skill-based interventions such as surgery or psychotherapy. An expertise-based approach to trial design, where health professionals only deliver an intervention in which they have expertise, has been proposed as an alternative. The aim of this project was to systematically review the use of an expertise-based trial design in the medical literature. We carried out a comprehensive search of nine databases-AMED, BIOSIS, CENTRAL, CINAHL, Cochrane Methodology Register, EMBASE, MEDLINE, Science Citation Index, and PsycINFO-from 1966 to 2012 and performed citation searches using the ISI Citation Indexes and Scopus. Studies that used an expertise-based trial design were included. Two review authors independently screened the titles and abstracts and assessed full-text reports. Data were extracted and summarised on the study characteristics, general and expertise-specific study methodology, and conduct. In total, 7476 titles and abstracts were identified, leading to 43 included studies (54 articles). The vast majority (88 %) used a pure expertise-based design; three (7 %) adopted a hybrid design, and two (5 %) used a design that was unclear. Most studies compared substantially different interventions (79 %). In many cases, key information relating to the expertise-based design was absent; only 12 (28 %) reported criteria for delivering both interventions. Most studies recruited the target sample size or very close to it (median of 101, interquartile range of 94 to 118), although the target was reported for only 40 % of studies. The proportion of participants who received the allocated intervention was high (92 %, interquartile range of 82 to 99 %). While use of an expertise-based trial design is growing, it remains uncommon. Reporting of study methodology and, particularly, expertise-related methodology was poor. Empirical evidence provided some support for purported benefits such as high levels of recruitment and compliance with allocation. An expertise-based trial design should be considered but its value seems context-specific, particularly when interventions differ substantially or interventions are typically delivered by different health professionals.