Abstract

In patients with Legg-Calvé-Perthes (LCP), the Catterall and Herring classifications based on standard hip radiographs are frequently used. In this study, the compatibility of classification systems in hip radiographs and MR images of LCP patients was examined among specialist working in the departments of Orthopedics and Radiology, and the reliability was investigated. In this study, which included 37 Perthes patients, two separate presentations were prepared in which the radiological imaging of the patients was determined by drawing lots. Six experienced specialists were selected as observers, three of whom were radiologist and three of pediatric orthopedists. Each presentations were sent to the observers one month apart. They were asked to categorize visual data. Data were evaluated with percent agreement (PA) and Gwet's AC1 method. In the Catterall classification, agreement among pediatric orthopedists decreased in MRI examination, while an increase was observed among radiologists. In the Herring classification, while pediatric orthopedists remained at a similar level, radiologists' inter-observer reliability increased. When the intra-observer reliability was examined, it was observed that the radiologists in the Catterall classification and the orthopedists in the Herring classification were more stable in the evaluations made with MRI. The usability of both classification systems, which provide prognostic prediction, in MRI examinations is a crucial unknown. It is fruitful that both radiologists and pediatric orthopedists work together in novel classification systems to be created using MRI.

Key words: Hip radiograph, MRI, legg-calvé-perthes disease, reliability