[Correlation and diagnostic agreement between magnetic resonance imaging and arthroscopy in rotator cuff rupture: a 10-year study at the Naval Medical Center].

Introduction: arthroscopy is the gold standard for detecting rotator cuff tears, but its invasive nature limits widespread use. Magnetic resonance imaging (MRI) serves as a precise alternative. The correlation and diagnostic agreement between MRI and arthroscopy are unclear in the Mexican population.

Material and methods: adults with rotator cuff syndrome who underwent shoulder arthroscopy and MRI were included. Findings were classified as total tear (100%), partial tear grade I (< 50%), partial tear grade II ( 50%), or no tear. Concordance and correlation between MRI and arthroscopy were evaluated. The accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of MRI for diagnosing rotator cuff tears were calculated.

Results: we included 147 patients with rotator cuff syndrome. MRI identified tears in 83%, of which 91.8% were confirmed by arthroscopy. Diagnostic agreement between arthroscopy and MRI was moderate (kappa = 0.452, p = 0.0001). Similarly, arthroscopy and MRI showed moderate agreement in classifying partial tears grade I, partial tears grade II, and complete tears (kappa = 0.412, p = 0.0001). Significant correlations were found in detecting partial tears grade I, partial tears grade II, and complete tears (rho = 0.77, p = 0.0001) between MRI and arthroscopy.

Conclusions: magnetic Resonance Imaging (MRI) and arthroscopy demonstrated moderate diagnostic agreement (kappa = 0.452, p = 0.0001) along with a strong positive correlation (rho = 0.77, p = 0.0001) for detecting rotator cuff tears, regardless of severity.