Samantha Bunzli1, Emma Choong1, Cade Shadbolt1, Laura Wall2, Elizabeth Nelson1, Chris Schilling1, Helen Wilding3, L Stefan Lohmander4, Zsolt J Balogh5, Francesco Paolucci2,6, Philip Clarke7, Peter F M Choong1, Michelle M Dowsey1. 1. Department of Surgery, The University of Melbourne, St. Vincent's Hospital Melbourne, Victoria, Australia. 2. College of Human and Social Futures, the Newcastle Business School, University of Newcastle, Newcastle, Australia. 3. St Vincent's Hospital Melbourne, Library Service, Fitzroy, Australia. 4. Department of Orthopaedics, University of Lund, Lund, Sweden. 5. The University of Newcastle, School of Medicine and Public Health, Newcastle, Australia. 6. Schools of Economics and Management at the University of Bologna, Bologna, Italy. 7. Health Economics Research Centre, Oxford University, Oxford, United Kingdom.
Abstract
OBJECTIVE: To explore whether placebo surgery controlled trials achieve what they set out to do by investigating discrepancy between projected and actual design aspects of trials identified through systematic review methods. SUMMARY BACKGROUND: Interest in placebo surgery controlled trials is growing in response to concerns regarding unnecessary surgery and the societal cost of low-value healthcare. As questions about the justifiability of using placebo controls in surgery have been addressed, attention is now being paid to more practical concerns. METHODS: Six databases were searched from inception - May 2020 (MEDLINE, Embase, Emcare, APA PsycInfo, CINAHL, Cochrane Library). Placebo surgery controlled trials with a published protocol were included. Three authors extracted "projected" design aspects from protocols and "actual" design aspects from main findings papers. Absolute and relative difference between projected and actual design aspects were presented for each trial. Trials were grouped according to whether they met their target sample size ("completed") and were concluded in a timely fashion. Pairs of authors assessed risk of bias. RESULTS: Of 24 trials with data available to analyse; 3 were completed and concluded within target timeframe; 10 were completed and concluded outside the target timeline; 4 were completed without clear target timeframes; 2 were incomplete and concluded within the target framework; 5 were incomplete and concluded outside the target timeline. Trials which reached the recruitment target underestimated trial duration by 88% and number of recruitment sites by 87%. CONCLUSIONS: Trialists need to factor additional time and sites into future placebo surgery controlled trials. A robust reporting framework of projected and actual trial design is imperative for trialists to learn from their predecessors. REVIEW REGISTRATION: PROSPERO (CRD42019133296).
OBJECTIVE: To explore whether placebo surgery controlled trials achieve what they set out to do by investigating discrepancy between projected and actual design aspects of trials identified through systematic review methods. SUMMARY BACKGROUND: Interest in placebo surgery controlled trials is growing in response to concerns regarding unnecessary surgery and the societal cost of low-value healthcare. As questions about the justifiability of using placebo controls in surgery have been addressed, attention is now being paid to more practical concerns. METHODS: Six databases were searched from inception - May 2020 (MEDLINE, Embase, Emcare, APA PsycInfo, CINAHL, Cochrane Library). Placebo surgery controlled trials with a published protocol were included. Three authors extracted "projected" design aspects from protocols and "actual" design aspects from main findings papers. Absolute and relative difference between projected and actual design aspects were presented for each trial. Trials were grouped according to whether they met their target sample size ("completed") and were concluded in a timely fashion. Pairs of authors assessed risk of bias. RESULTS: Of 24 trials with data available to analyse; 3 were completed and concluded within target timeframe; 10 were completed and concluded outside the target timeline; 4 were completed without clear target timeframes; 2 were incomplete and concluded within the target framework; 5 were incomplete and concluded outside the target timeline. Trials which reached the recruitment target underestimated trial duration by 88% and number of recruitment sites by 87%. CONCLUSIONS: Trialists need to factor additional time and sites into future placebo surgery controlled trials. A robust reporting framework of projected and actual trial design is imperative for trialists to learn from their predecessors. REVIEW REGISTRATION: PROSPERO (CRD42019133296).