This article is more than 5 years old. Information may no longer be current.
Reliability, accuracy varies for radiographic assessment of knee OA prior to knee replacement
Although results from this randomized control study showed variable intra-reader and inter-reader agreement on individual radiographic characteristics between a radiologist and non-clinicians who were trained to evaluate radiographic aspects of knee osteoarthritis, there was substantial reliability regarding the Kellgren-Lawrence grade and Osteoarthritis Research Society summary score.
Researchers enrolled 309 patients who underwent primary total knee replacement and were diagnosed with osteoarthritis (OA). A radiologist trained three non-clinicians to evaluate radiographs to identify characteristics of OA. Before total knee replacement, the non-clinicians and radiologist looked at the preoperative films of 36 patients. The Kellgren-Lawrence (KL) grade and Osteoarthritis Research Society (OARSI) summary score were also calculated. Investigators assessed intra-reader and inter-reader reliability using the intraclass correlation coefficient (ICC) and the weighted kappa statistic.
Results for the three non-clinicians showed the intra-reader reliability ranged from 0.40 to 1.0 for the individual characteristics and ranged from 0.72 to 1 for the KL grade. For the OARSI score, the ICC ranged from 0.89 to 0.98. Investigators noted the inter-reader reliability ranged from 0.45 to 0.94 for the individual aspects of the radiographs, and the inter-reader reliability for the KL grade ranged from 0.66 to 0.97. For the OARSI score, the ICC ranged from 0.87 to 0.96.
According to researchers, the inter-reader reliability ranged from 0.56 to 0.85 for the KL grade between the radiologist and the non-clinicians. The ICC for the OARSI summary score ranged from 0.79 to 0.88. ‒ by Monica Jaramillo
Disclosure: The research was supported by grants from the NIH/National Institute of Arthritis and Musculoskeletal and Skin Diseases: K24AR057827, T32AR055885.