The quality and reporting of randomized trials in cardiothoracic physical therapy could be substantially improved.

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 have 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.

RESULTS

The mean ± SD total PEDro score for the 2,970 included reports of cardiothoracic trials was 4.7 ± 1.4, with 27% being of moderate to high quality. The clinical conditions studied included chronic lung diseases (32% of the trials), cardiac diseases (20%), cardiovascular surgical conditions (5%), sleep disorders (5%), peripheral vascular disease (4%), acute lung disease (4%), critical illness (3%), and other surgical conditions (3%). 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, description of 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.