H A Bradley1, J J Rucklidge1, R T Mulder2. 1. Department of Psychology, University of Canterbury, Christchurch, New Zealand. 2. Department of Psychological Medicine, University of Otago, Otago, New Zealand.
Abstract
OBJECTIVE: Selective outcome reporting poses serious implications on our evidence base for best practice. The extent to which selective outcome reporting and trial registration occurs in the wider psychotherapy literature needs to be investigated. METHOD: Randomized controlled psychotherapy trials published between 2010 and 2014 were selected from the five highest impact factor journals in clinical psychology that publish clinical trials. Data on primary and secondary outcomes, funding, and participant numbers were extracted from the article and registry and compared. RESULTS: From 112 trials, 67 (59.8%) were registered, 27 (24.1%) were prospectively registered, and only 13 (11.6%) were correctly registered and reported. Seven of these 13 trials showed evidence of selective outcome reporting, of which four had discrepancies favoring significant outcomes. One of the remaining six trials changed their primary outcomes during participant enrollment. Overall, only five (4.5%) trials were free from selective outcome reporting. Three of these five trials had more than a 10% change between planned and achieved sample size. Funding was not associated with correct registration or reporting. CONCLUSIONS: The proportion of psychotherapy randomized controlled trials correctly registered and transparently reported is poor. Psychologists should consider the impact these results have on public confidence in reported outcomes.
OBJECTIVE: Selective outcome reporting poses serious implications on our evidence base for best practice. The extent to which selective outcome reporting and trial registration occurs in the wider psychotherapy literature needs to be investigated. METHOD: Randomized controlled psychotherapy trials published between 2010 and 2014 were selected from the five highest impact factor journals in clinical psychology that publish clinical trials. Data on primary and secondary outcomes, funding, and participant numbers were extracted from the article and registry and compared. RESULTS: From 112 trials, 67 (59.8%) were registered, 27 (24.1%) were prospectively registered, and only 13 (11.6%) were correctly registered and reported. Seven of these 13 trials showed evidence of selective outcome reporting, of which four had discrepancies favoring significant outcomes. One of the remaining six trials changed their primary outcomes during participant enrollment. Overall, only five (4.5%) trials were free from selective outcome reporting. Three of these five trials had more than a 10% change between planned and achieved sample size. Funding was not associated with correct registration or reporting. CONCLUSIONS: The proportion of psychotherapy randomized controlled trials correctly registered and transparently reported is poor. Psychologists should consider the impact these results have on public confidence in reported outcomes.
Authors: Marleine Azar; Kira E Riehm; Nazanin Saadat; Tatiana Sanchez; Matthew Chiovitti; Lin Qi; Danielle B Rice; Brooke Levis; Claire Fedoruk; Alexander W Levis; Lorie A Kloda; Jonathan Kimmelman; Andrea Benedetti; Brett D Thombs Journal: JAMA Intern Med Date: 2019-05-01 Impact factor: 21.873
Authors: Caroline Elizabeth Sherry; Jonathan Z Pollard; Daniel Tritz; Branden K Carr; Aaron Pierce; Matt Vassar Journal: Gen Psychiatr Date: 2020-02-12