Intraobserver and interobserver reliability of Kalamchi and Macewen's classification system for evaluation of avascular necrosis of the femoral head in developmental hip dysplasia.

Abstract

This study evaluates the intraobserver and interobserver reliability of the Kalamchi and MacEwen's classification system of avascular necrosis of the femoral head. Radiographs of 48 developmentally dysplastic hips that had an average follow-up of 40.5 months (range: 36 to 52 months) and that had been treated by the same operative technique were interpreted twice by four experienced pediatric orthopaedic surgeons. When the absence or presence of avascular necrosis was taken into consideration the average intraobserver agreement percentage and kappa coefficient were 86% and 0.71, respectively. The average interobserver agreement percentage and kappa coefficient were 83% and 0.66, respectively. When the agreement on the type of avascular necrosis was analyzed, the average intraobserver agreement percentage and kappa coefficient were 85% and 0.74, respectively. The average interobserver agreement percentage and kappa coefficient were 81% and 0.66, respectively. No statistically significant difference was found between the rates of avascular necrosis of four observers. The Kalamchi and MacEwen's classification system was found to be reliable and reproducible.