RT Journal Article SR Electronic T1 The quality and reporting of randomized trials in cardiothoracic physical therapy could be substantially improved JF Respiratory Care FD American Association for Respiratory Care SP respcare.02379 DO 10.4187/respcare.02379 A1 NĂ¡dia N Geha A1 Anne M Moseley A1 Mark R Elkins A1 Luciana D Chiavegato A1 Silvia R Shiwa A1 Leonardo O P Costa YR 2013 UL http://rc.rcjournal.com/content/early/2013/04/16/respcare.02379.abstract AB Background: While the number of reports of randomized controlled trials in physical therapy has increased substantially in the last decades, the quality and reporting of randomized trials has never been systematically investigated in the subdiscipline of cardiothoracic physical therapy. The primary aim was to determine the methodological quality and completeness of reporting of cardiothoracic physical therapy trials. Secondary aims were to investigate the range of clinical conditions investigated in these trials and the degree of association between trial characteristics and quality. Methods: All reports of randomized trials indexed on the Physiotherapy Evidence Database (PEDro) and coded as being relevant to cardiothoracic physical therapy were surveyed. PEDro scale individual items and total score were downloaded and some characteristics included in the Consolidated Standards of Reporting Trials (CONSORT) Statement were extracted for each trial report. Descriptive statistics were used to summarize the quality of reports. Multivariate linear regression was used to examine the association between trial characteristics and quality. Results: The mean total PEDro score for the 2970 included reports of cardiothoracic trials was 4.7 (standard deviation 1.4), with 27% being of moderate to high quality. The clinical conditions studied included chronic lung diseases (32% of trials), cardiac diseases (20%), cardiovascular surgical conditions and sleep disorders (5% each), peripheral vascular disease and acute lung disease (4% each) and critical illness and other surgical conditions (3% each). The multivariate linear regression analysis revealed that endorsement of the CONSORT Statement by the publishing journal, time since publication, evidence of trial registration, sources of funding, describing the sample size calculation, and identification of the primary outcome(s) had associations with the total PEDro score. Conclusions: There is great potential to improve the quality of the conduct and reporting of trials evaluating the effects of cardiothoracic physical therapy.