AimsTo compare the diagnostic accuracy of investigators from different specialities (radiologists and orthopaedic surgeons) with varying levels of experience of 1.5 T direct magnetic resonance arthrography (dMRA) against intraoperative findings in patients with femoroacetabular impingement syndrome (FAIS).MethodsA total of 272 patients were evaluated with dMRA and subsequent hip arthroscopy. The dMRA images were evaluated independently by two non-hip-arthroscopy-trained orthopaedic surgeons, two fellowship-trained musculoskeletal radiologists, and two hip-arthroscopy-trained orthopaedic surgeons. The radiological diagnoses were compared with the intraoperative findings.ResultsHip arthroscopy revealed labral pathologies in 218 (79%) and acetabular chondral lesions in 190 (69%) hips. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy for evaluating the acetabular labral pathologies were 79%, 18%, 79%, 18%, and 66% (non-hip-arthroscopy trained orthopaedic surgeons), 83%, 36%, 83%, 36%, and 74% (fellowship-trained musculoskeletal radiologists), and 88%, 53%, 88%, 54% and 81% (hip-arthroscopy trained orthopaedic surgeons). The sensitivity, specificity, PPV, NPV and accuracy of dMRA for assessing the acetabular chondral damage were 81%, 36%, 71%, 50%, and 66% (non-hip-arthroscopy trained orthopaedic surgeons), 84%, 38%, 75%, 52%, and 70% (fellowship-trained musculoskeletal radiologists), and 91%, 51%, 81%, 73%, and 79% (hip-arthroscopy trained orthopaedic surgeons). The hip-arthroscopy trained orthopaedic surgeons displayed the highest percentage of correctly diagnosed labral pathologies and acetabular chondral lesions, which is significantly higher than the other two investigator groups (p < 0.05).ConclusionThe accuracy of dMRA on detecting labral pathologies or acetabular chondral lesions depends on the examiner and its level of experience in hip arthroscopy. The highest values are found for the hip-arthroscopy-trained orthopaedic surgeons.Level of evidenceRetrospective cohort study; III.
Read full abstract