C B Chang1, S C Seong, T K Kim. 1. Joint Reconstruction Center, Seoul National University Bundang Hospital, Seongnamsi, Republic of Korea.
Abstract
OBJECTIVE: To determine whether assessments of patellofemoral (PF) joint space reliably predict the cartilage conditions in the PF joint of the patients undergoing total knee arthroplasty (TKA) for advanced knee osteoarthritis. METHODS: Radiographic joint spaces and gross cartilage conditions in the PF joint were assessed in 151 osteoarthritic knees undergoing TKA. Minimum joint space widths (MJSWs) in the medial and lateral compartments of the PF joint were measured separately on Merchant radiographs. Degrees of joint space were graded as normal, narrowed and obliterated, and expected locations of significant cartilage lesions were classified as medial, lateral, and global. Cartilage lesions were grossly assessed in terms of their severity (depth and size) and location. Cross-table analyses and kappa statistics were used to determine the level of agreements between radiographic and gross assessments and the diagnostic accuracies of the radiographic assessments. RESULTS: The agreements between the radiographic assessments and the gross assessments on the lesion severities and locations were only fair [kappa coefficient (k)=0.288 and 0.211, respectively]. The cross-table analyses demonstrated that 45 (47.4%) of 95 knees with a normal radiographic joint space had moderate or severe cartilage degeneration of the PF joint identified with gross assessments. In the radiographic assessments, the lateral compartment of the PF joint was the most frequent location of joint space narrowing (71.4%) whereas in the gross assessments, the medial compartment of the PF joint was the most frequent location of significant cartilage lesion (48.1%). Diagnostic accuracies on the lesion severities and locations were generally poor. CONCLUSION: This study demonstrates that prediction of the cartilage conditions of the PF joint by the radiographic joint space can be inaccurate.
OBJECTIVE: To determine whether assessments of patellofemoral (PF) joint space reliably predict the cartilage conditions in the PF joint of the patients undergoing total knee arthroplasty (TKA) for advanced knee osteoarthritis. METHODS: Radiographic joint spaces and gross cartilage conditions in the PF joint were assessed in 151 osteoarthritic knees undergoing TKA. Minimum joint space widths (MJSWs) in the medial and lateral compartments of the PF joint were measured separately on Merchant radiographs. Degrees of joint space were graded as normal, narrowed and obliterated, and expected locations of significant cartilage lesions were classified as medial, lateral, and global. Cartilage lesions were grossly assessed in terms of their severity (depth and size) and location. Cross-table analyses and kappa statistics were used to determine the level of agreements between radiographic and gross assessments and the diagnostic accuracies of the radiographic assessments. RESULTS: The agreements between the radiographic assessments and the gross assessments on the lesion severities and locations were only fair [kappa coefficient (k)=0.288 and 0.211, respectively]. The cross-table analyses demonstrated that 45 (47.4%) of 95 knees with a normal radiographic joint space had moderate or severe cartilage degeneration of the PF joint identified with gross assessments. In the radiographic assessments, the lateral compartment of the PF joint was the most frequent location of joint space narrowing (71.4%) whereas in the gross assessments, the medial compartment of the PF joint was the most frequent location of significant cartilage lesion (48.1%). Diagnostic accuracies on the lesion severities and locations were generally poor. CONCLUSION: This study demonstrates that prediction of the cartilage conditions of the PF joint by the radiographic joint space can be inaccurate.