Content reporting of exercise interventions in rotator cuff disease trials: results from application of the Consensus on Exercise Reporting Template (CERT)

BACKGROUND: Exercise interventions are frequently recommended for patients with rotator cuff disease, but poor content reporting in clinical trials of exercise limits interpretation and replication of trials and clinicians’ ability to deliver effective exercise protocols. The Consensus on Exercise Reporting Template (CERT) was developed to address this problem. OBJECTIVE: To assess completeness of content reporting of exercise interventions in randomised controlled trials for patients with rotator cuff disease and the inter-rater reliability of the CERT. DESIGN: Critical appraisal. METHODS: Independent pairs of reviewers applied the CERT to all 34 exercise trials from the most recent Cochrane Review evaluating the effect of manual therapy and exercise for patients with rotator cuff disease. We used the CERT Explanation and Elaboration Statement to guide assessment of whether each of the 19-item criteria were clearly described (score 0–19; higher scores indicate better reporting). Percentage agreement and the prevalence and bias adjusted kappa (PABAK) coefficient were used to measure inter-rater reliability. RESULTS: The median CERT score was 5 (range 0–16). Percentage agreement was high for 15 items and acceptable for 4 items. The PABAK coefficient indicated excellent (5 items), substantial (11 items) and moderate (3 items) inter-rater agreement. CONCLUSION: The description of exercise interventions for patients with rotator cuff disease in published trials is poorly reported. Overall, the inter-rater reliability of the CERT is high/acceptable. We strongly encourage journals to mandate use of the CERT for papers reporting trial protocols and results investigating exercise interventions.