Comparison of arthroscopic and MRI findings of osteochondral damage in knees.

Magnetic resonance imaging (MRI) scans are widely used in the assessment of knees, often prior to arthroscopic procedures. The reporting of osteochondral damage on MRI scans can be variable. The correlation between MRI reports of osteochondral damage and that found at arthroscopy is often inconsistent. A retrospective case-note review of a single-surgeon series of 175 arthroscopic procedures was performed. Eighty-three patients were included in the study. The remainder were excluded if an MRI scan had not been performed or had been performed more than 3-months before surgery. The condition of the articular cartilage demonstrated by MRI was compared to that found at arthroscopy. Data was analysed for presence and extent of osteochondral damage. Comparison between MRI and arthroscopy findings showed high specificity (90%) and negative predictive values (89%) for osteochondral damage but low sensitivity (46%). Cohen's kappa values < 0.2 revealed very poor correlation for the extent of damage. This study demonstrates MRI as a good identifier of osteochondral damage but an unreliable descriptor for such change.