Conflict of interest in sports medicine: does it affect our judgment?

The American Academy of Orthopaedic Surgeons (AAOS) and other orthopedic societies require authors to disclose conflicts of interest (COIs). We conducted a study to evaluate how a hypothetical research team's reported COI would influence the perceived value of its data. Using a hypothetical prospective study, we asked orthopedic surgeons and nonoperative sports medicine specialists to rate the value of the data, given different study designs, statistical significance, and research institutions (academic vs private). The fictional research team disclosed the project was funded by a pharmaceutical company and all team members received consulting compensation. Eighty percent of 522 respondents thought COI disclosure is important in the interpretation of study results, 41% reported always using this information when interpreting data, and 24% reported that a case series with significant positive results at an academic center was likely trustworthy (this percentage decreased to 5% when the study was set in a community hospital). When no significant difference was found in results, 42% thought the study was trustworthy. When the study design yielded level I evidence (randomized controlled trial) at an academic center, 57% thought the study was trustworthy (when the study was set in a community hospital, this percentage decreased to 39%). When the results of the design showed no difference among groups, the majority of respondents (62%) thought the study was trustworthy. Although the majority of respondents thought disclosure is important, fewer than half reportedly used this information when interpreting study results. Randomized controlled trial status improved the perceived reliability of the data over a case series but was not as important as reporting "negative" results.