The association between registration status and reported outcomes in physiotherapy randomised controlled trials

Abstract

Background/Aims Clinical trial registration has been proposed as a method of mitigating selective reporting in scientific research. It remains unknown whether trial registration is associated with reported outcomes in physiotherapy trials. This study aimed to analyse the association between registration status and outcome (the rejection or acceptance of a primary null hypothesis) for physiotherapy randomised controlled trials. Methods All randomised controlled trials reporting a physiotherapy intervention in publications listed in PubMed between 1 January 2017 and 30 June 2017 were included. Trial registration was determined based on the reporting of a registration number in the primary article or by identifying trials through trial registries. Results Of the 291 trials analysed, 176 (60.5%) were registered; 115 (39.5%) were not. There was no significant association between trial registration and outcome on multivariate analyses (Odds Ratio 1.65; 95% Confidence Interval (0.92–2.96); P=0.09). Only 22% of trials were prospectively registered. Conclusions Registration status and trial outcome are not associated in randomised controlled trials of physiotherapy interventions. The rate of physiotherapy trial registration remains low.