A systematic review identifies shortcomings in the reporting of crossover trials in chronic painful conditions

Objectives: To investigate the reporting of study features of interest in abstracts and full texts of journal publications of crossover trials in chronic painful conditions. Study Design and Setting: Systematic review based on a MEDLINE (PubMed) search (January 1990-August 2014). Results: Ninety-eight publications on crossover studies with 3,513 study participants were eligible for inclusion. Double-blind status and randomized allocation to treatment groups are commonly reported in both abstracts and full texts (90 of 98 publications and 82 of 98 publications, respectively). Adverse events are reported in both abstract and full text in 49 of 98 publications and in the full text only in 44 of 98. A breakdown of results by treatment period is provided only in 23 of 98 publications, and if so, is reported only in the full text, never in the abstract. There is a time trend for the reporting of randomization in abstracts; it is more likely to be reported in recent studies (P = 0.0094). No time trends are detected in the reporting of double-blind status (P = 0.1087) and adverse events (P = 0.6084). Conclusion: The reporting of adverse events in the abstract and the reporting of results specified by crossover period in the full texts of journal publications on crossover pain trials should be improved. (C) 2015 Published by Elsevier Inc.