Femoral tunnel reaming method in anterior cruciate ligament reconstruction cannot be determined from plain radiographs alone

Background: The precise location of tunnel placement significantly influences the outcome of anterior cruciate ligament (ACL) reconstruction, and an accurate understanding of tunnel placement remains paramount during postoperative assessment. Despite this, surgeons commonly estimate tunnel position and the method used to ream the tunnel based on plain radiographs alone. This study aims to determine the accuracy and reliability with which orthopaedic surgeons and musculoskeletal specialists can identify femoral tunnel (FT) reaming technique using standard postoperative radiographs alone, and to evaluate whether perceived tunnel malposition influences assumptions about reaming method. Methods: This cross-sectional diagnostic study reviews postoperative radiographs from 40 ACL reconstructions (ACLRs) performed by the senior author. Twenty transtibially-reamed knees were matched with 20 anteromedially-reamed knees according to age, gender and body mass index (BMI). Surgical technique was identical for both groups, with the exception of FT reaming method, and postoperative radiographs were routinely obtained two weeks after surgery. Fifteen blinded orthopaedic or musculoskeletal trained reviewers evaluated radiographs for tunnel position and assumed reaming method. A kappa analysis was used to determine agreement among reviewers and assumed reaming method error rates were assessed parametrically. Results: Reaming method was chosen correctly 64% of the time. However, the overall kappa value was low at 0.26. Transtibial error rate correlated with negative judgements of FT placement with a Pearson correlation coefficient of 0.648 (P=0.01). Conclusions: These results suggest that the reaming method cannot be reliably determined using standard postoperative radiographs, and that evaluators who judge the tunnel placement to be poor may be biased toward assuming the tunnel was reamed transtibially. Surgeons should be careful not to associate failed ACLR with particular reaming methods without additional data.