Clinical trial registration has become more prevalent in physical therapy but it is still inadequate: A meta-research study.

ABSTRACT

BACKGROUND: A study using data from 2009 showed low prevalence and inadequate trial registration in physiotherapy. In 2013, a joint editorial recommended prospective registration in physiotherapy journals. Ten years later it is unclear whether the joint editorial achieved its intended benefit. OBJECTIVES: To investigate the proportion of randomized trials adequately registered and the extent of selective reporting of outcomes in trials of physiotherapy interventions published in 2019 and to compare these data with equivalent published data from 2009. DESIGN: Meta-research study. METHOD: A random sample of 200 trials published in 2019 was used. Evidence of registration was sought on trial registers and by contacting authors. Data from the article was compared with data from the trial registration. Data from this sample of trial published in 2019 were compared with equivalent published data from 2009. RESULTS: In 2019, the proportion of trials that were registered was 63% versus 34% in 2009 (absolute difference 29%). In 2019, 18% of the trials were prospectively registered compared to 6% in 2009 (absolute difference 12%). Unambiguous primary outcomes (i.e., method and timepoints of measurement clearly defined in the trial registry entry) were registered for 30% in 2019. Registration was adequate (i.e., prospective with unambiguous primary outcomes) for 8%, compared with 3% in 2009 (absolute difference 5%). Selective outcome reporting occurred in 73% of the trials in which it was assessable; in 2009 this proportion was 47% (absolute difference 26%). CONCLUSIONS: Registration of randomized trials in physiotherapy increased in the past decade, but it is still inadequate. More effort is still required to implement and enforce adequate registration.