Do Published Orthopaedic RCTs Match Their Registered Protocols? A 6-year Analysis of Leading Orthopaedic Journals

RCTs are a key block in the evidence pyramid, but their quality relies on detailed, consistent reporting, and one best-practice standard is prospective registration. Prospective trial registration was intended to reduce publication bias, and adherence to a prespecified protocol helps limit bias from selective reporting; any protocol or end point changes should be transparently documented and justified. However, the degree to which articles published in the leading journals on orthopaedic surgery comply with this best-practice standard has, to our knowledge, not been evaluated.

(1) Do RCTs published in leading, general-interest orthopaedic surgery journals comply with best practices regarding prospective clinical trial registration? (2) Do major discrepancies exist between registered protocols and published orthopaedic RCTs? (3) Are there specific study types that are more likely to demonstrate discrepancies?

A review was performed on RCTs published in the top five general-interest orthopaedic surgery journals, based on the 2022 scientific journal rankings (from SciMago): Journal of Bone and Joint Surgery , Bone and Joint Journal , Clinical Orthopaedics and Related Research ®, Journal of the American Academy of Orthopaedic Surgeons (JAAOS) , and Acta Orthopaedica . During the study period, all journals maintained editorial policies requiring prospective clinical trial registration as a condition of consideration for publication except for JAAOS . A systematic search on PubMed retrieved 705 potential publications, of which 324 RCTs fulfilled the inclusion criteria. For each trial, nine essential elements from the 24-item WHO minimum data set were extracted and compared between the published article and its trial registry entry, focusing on health condition, intervention, sample size, outcomes, and eligibility criteria. To answer our first question regarding compliance, we audited each article to identify the presence of a registry code. For our second question, we performed a side-by-side comparison of nine essential elements from the WHO trial registration data set (including primary outcomes, sample size, and eligibility criteria) to identify discrepancies between the registry and the final publication. Finally, to address our third question, we used chi-square tests to determine whether study characteristics, such as country of origin or subspecialty, were associated with higher rates of reporting shifts.

Most orthopaedic RCTs published in leading journals complied with registration standards, with 95% (309 of 324) having an identifiable registry entry. However, 2% (8 of 324) were published without any registry identifier or justification for its absence, and 2% (7 of 324) were identified as long-term follow-up visits that did not have unique prospective entries. Major discrepancies between registered protocols and published manuscripts were frequent. Discrepancies in the sample size occurred in 33% (102 of 309) of trials. Discrepancies in the primary outcome occurred in 25% (78 of 309) of trials. Discrepancies in the secondary outcome occurred in 60% (185 of 309) of trials. Discrepancies in the inclusion criteria occurred in 33% (102 of 309) of trials. Discrepancies in the exclusion criteria occurred in 53% (165 of 309) of trials. Trials conducted in the United States or as multicenter international collaborations were more likely to update their final results in the registry compared with single-country trials conducted outside the US (37% versus 10%; p < 0.001). No other study characteristics, including publication year or subspecialty, were associated with the presence of reporting discrepancies.

Prospective registration has become the standard for RCTs in high-impact orthopaedic journals. However, our findings suggest that a gap still exists between having a registry and the accuracy of the information contained within it. These findings suggest that registration is often treated as a procedural requirement rather than a rigorous commitment to a fixed study protocol.

The orthopaedic research community should adopt stricter standards for trial registration, reporting, and verification of registry entries to reduce undisclosed protocol changes and improve confidence in published evidence.