OBJECTIVE: To determine diagnostic performance statistics of extra-labral magnetic resonance (MR) findings for detection of labral tears in a population of patients with clinical suspicion of this diagnosis. MATERIALS AND METHODS: Seventy-nine patients clinically suspected of having a labral tear (who underwent arthroscopy) had their MR studies retrospectively reviewed to determine the presence of lateral acetabular oedema-like marrow signal, ganglia, dysplastic femoral bumps, synovial herniation pits and geodes. These findings were then correlated with the arthroscopic presence (or absence) of a labral tear. RESULTS: All findings (lateral acetabular oedema-like marrow signal, ganglia, dysplastic femoral bumps, synovial herniation pits and geodes) had a specificity and positive predictive value (PPV) of 100%. Lateral acetabular oedema-like marrow signal had a sensitivity of 35% and 20% negative predictive value (NPV). This was the only statistically significant finding (P < 0.05). The sensitivity and NPV of ganglia were 12% and 16%, dysplastic femoral bumps (12%, 16%), synovial herniation pits (4%, 14%) and geodes (6%, 15%) respectively, (P > 0.05). CONCLUSION: Lateral acetabular oedema-like marrow signal is a useful sign (100% PPV) in the MR diagnosis of a labral tear, if one is clinically suspected. The other findings (ganglia, dysplastic femoral bumps, synovial herniation pits and geodes) were not statistically significant. Further studies are required to evaluate these.
OBJECTIVE: To determine diagnostic performance statistics of extra-labral magnetic resonance (MR) findings for detection of labral tears in a population of patients with clinical suspicion of this diagnosis. MATERIALS AND METHODS: Seventy-nine patients clinically suspected of having a labral tear (who underwent arthroscopy) had their MR studies retrospectively reviewed to determine the presence of lateral acetabular oedema-like marrow signal, ganglia, dysplastic femoral bumps, synovial herniation pits and geodes. These findings were then correlated with the arthroscopic presence (or absence) of a labral tear. RESULTS: All findings (lateral acetabular oedema-like marrow signal, ganglia, dysplastic femoral bumps, synovial herniation pits and geodes) had a specificity and positive predictive value (PPV) of 100%. Lateral acetabular oedema-like marrow signal had a sensitivity of 35% and 20% negative predictive value (NPV). This was the only statistically significant finding (P < 0.05). The sensitivity and NPV of ganglia were 12% and 16%, dysplastic femoral bumps (12%, 16%), synovial herniation pits (4%, 14%) and geodes (6%, 15%) respectively, (P > 0.05). CONCLUSION: Lateral acetabular oedema-like marrow signal is a useful sign (100% PPV) in the MR diagnosis of a labral tear, if one is clinically suspected. The other findings (ganglia, dysplastic femoral bumps, synovial herniation pits and geodes) were not statistically significant. Further studies are required to evaluate these.