Hip Magnetic Resonance Arthrography Using Normal Saline Is Less Accurate Than Using a Gadolinium-Based Contrast Agent

Abstract

To evaluate the diagnostic performance of hip magnetic resonance (MR) arthrography with a gadolinium-based contrast agent (GBCA) or normal saline (NS) for intra-articular pathologies. This retrospective study was approved by the institutional review board of our hospital, and the requirement for informed consent was waived. The study included 37 consecutive patients (38 hips; 21 right hips; 24 males; mean age, 33 years) who underwent GBCA-hip MR arthrography from July 2011 to January 2020 and 30 consecutive patients (30 hips; 20 right hips; 21 males; mean age, 40 years) who underwent NS-hip MR arthrography from January 2018 to June 2020. All images were evaluated twice independently by two radiologists blinded to the arthroscopic findings for the presence of labral tears, cartilage abnormalities, or ligamentum teres tears. Intra- and inter-reader reliabilities were determined by kappa values (k) using the Chi-squared test, and diagnostic performance was evaluated based on the arthroscopic findings. A p-value less than 0.05 was considered to indicate statistical significance. Intra-reader reliability in the detection of intra-articular pathologies ranged from moderate to almost perfect (k =0.510-0.840) and inter-reader reliability ranged from moderate to substantial (k =0.441-0.695) for GBCA- and NS-hip MR arthrography. The diagnostic accuracy of GBCA- and NS-hip MR arthrography was as follows: 81.6-92.1% and 66.7-73.3% for labral tears, respectively; 68.4-86.8% and 63.3-70.0% for cartilage abnormality, respectively; and 68.4-76.3% and 50.0-56.7% for ligamentum teres tears, respectively. NS-hip MR arthrography may be less accurate than GBCA-hip MR arthrography. Level II of Development of diagnostic criteria (consecutive patients with consistently applied reference standard and blinding).