Harms reporting in randomized controlled trials underpinning the American Academy of Orthopaedic Surgeons clinical practice guidelines for glenohumeral osteoarthritis

Abstract

Glenohumeral osteoarthritis is one of the most common causes of shoulder pain. As such, the American Academy of Orthopaedic Surgeons (AAOS) has developed clinical practice guidelines (CPGs) to address the management of glenohumeral osteoarthritis. These CPG recommendations stem from the findings of randomized controlled trials (RCTs), which have been shown to influence clinical decision making and health policy. Therefore, it is essential that trial outcomes, including harms data (ie, adverse events), are adequately reported. We intend to evaluate the reporting quality of harms-related data in orthopedic literature specifically relating to AAOS CPG recommendations on the management of glenohumeral osteoarthritis. We adhered to the Preferred Reporting Items for Systematic Reviews (PRISMA) as well as guidance for reporting meta-research. The AAOS CPGs for glenohumeral osteoarthritis were obtained from orthoguidelines.org, and 2 authors independently screened the guidelines for the RCTs referenced. A total of 14 studies were identified. Data were extracted from the 14 included studies independently by the same 2 authors. Adherence to the Consolidated Standards of Reporting Trials (CONSORT) Extension for Harms Checklist was assessed using an 18-item scoring chart, with 1 point being awarded for meeting a checklist item and 0 points being awarded for not meeting a checklist item. Descriptive statistics, such as frequencies, percentages, and 95% confidence intervals were used to summarize RCT adherence to the CONSORT checklist. The average score among the studies included was 7.36/18 items (39% adherence). No study adhered to all criteria, with the highest-performing study meeting 11 of 18 items (58%) and the lowest meeting 3 of 18 items (16%). A positive correlation between checklist score and year of publication was observed, with studies published more recently receiving a higher score on the CONSORT checklist (P<.05). Studies that disclosed funding information received a higher score than those that did not (P<.05), but there was no significant difference when the different funding sources were compared. Finally, double-blinded studies scored higher on the checklist than those with lower levels of blinding (single or no blinding, P<.05). Adverse events are poorly reported amongst RCTs cited as supporting evidence for AAOS Management of Glenohumeral Osteoarthritis CPGs, evidenced by a CONSORT checklist compliance rate of only 41% in this study. We recommend the development of an updated checklist with information that makes it easier for authors to recognize, evaluate, and report on harms data. Additionally, we encourage authors to include information about adverse events or negative outcomes in the abstract.