Objective: The purpose of this study was to evaluate the usefulness of magnetic resonance imaging (MRI) in detecting elbow articular cartilage injuries through comparison of preoperative MRI and magnetic resonance arthrography (MRA) with arthroscopic findings.
Design: Retrospective case analysis.
Setting: Tertiary care orthopedic private practice.
Patients: Consecutive series of 31 patients presenting with elbow pain and diagnosed at arthroscopy with articular cartilage defects of the elbow. All patients had a preoperative MRI or MRA using a 1.5 T magnet.
Interventions: Each patient had a systematic elbow arthroscopy, with careful inspection and recording of chondral injuries in 4 anatomical regions: capitellum, radius, trochlea, and ulna. Each MRI/MRA was then independently reviewed by 2 radiologists blinded to the arthroscopic findings.
Main outcome measures: The sensitivity, specificity, negative predictive value, positive predictive value, and accuracy were calculated for MRI and MRI compared with arthroscopy as the gold standard, for each of the anatomical regions.
Results: The accuracy of MRI was 45% for chondral injuries of the radius, 65% for the capitellum, 20% for the ulna, and 30% for the trochlea. The accuracy of MRA was 45% for chondral injuries of the radius, 64% for the capitellum, 18% for the ulna, and 27% for the trochlea.
Conclusion: We conclude that the ability of MRI and MRA using a 1.5 T magnet to detect articular cartilage lesions is limited. Neither MRI nor MRA demonstrates the intraarticular surface as accurately as direct visualization with the arthroscopy. This may be improved with the use of 3 T MRI.
Clinical relevance: This study demonstrates that MRI and MRA with a 1.5 T magnet, as used in community practice, have limited ability to detect cartilage lesions of the elbow.