Diagnostic Accuracy of MRI in Rotator Cuff Tears

Abstract

Rotator cuff tears are a common cause of shoulder pain and disability. MRI is widely used for diagnosis, but performance varies by tear type and tendon involved. Objective: To evaluate the diagnostic accuracy of conventional MRI for detecting rotator cuff tears using arthroscopy as the reference standard. Methods: An observational cohort study was conducted from June 2024 to December 2024. Adults with clinically suspected rotator cuff pathology underwent standardized shoulder MRI followed by arthroscopy within the study window. Consecutive sampling yielded 58 participants (mean age 52.79 ± 9.37 years; symptom duration 7.4 ± 4.93 months; 55.2% male). MRI reports documented the presence/absence of a tear, its thickness (partial vs. full), size category, and tendon involvement; intraoperative findings served as the Gold standard. Diagnostic metrics (sensitivity, specificity, PPV, NPV, accuracy) were calculated from 2×2 tables. ROC analysis assessed overall discrimination. Results: MRI identified a tear in 26/58 shoulders (44.8%); arthroscopy confirmed tears in 25/58 (43.1%). The diagnostic contingency table yielded 18 true positives, eight false positives, 25 true negatives, and seven false negatives. Corresponding performance metrics were: sensitivity 72.0%, specificity 75.8%, PPV 69.2%, NPV 78.1%, and accuracy 74.1%. ROC analysis showed fair discrimination (AUC 0.739; p < 0.002). Among MRI-positive cases, 57.7% were partial-thickness and 42.3% full-thickness; the supraspinatus was most frequently involved. Conclusion: Conventional MRI demonstrated moderate accuracy in detecting rotator cuff tears compared to arthroscopy, with performance influenced by the higher proportion of partial-thickness disease. These findings support MRI as a reliable first-line test, while highlighting potential gains from optimized 3-T protocols, selective MR arthrography in cases of equivocal partial tears, and targeted ultrasound for subscapularis-predominant pathology. Further, larger, standardized prospective studies are warranted to refine estimates and guide imaging pathways.