Intraobserver and interobserver reliability of Catterall, Herring, Salter–Thompson and Stulberg classification systems in Perthes disease

Objective To assess the percentage agreement and intraobserver and interobserver reliability among Catterall, Salter–Thompson, Herring and Stulberg classification systems with special reference to the stage of treatment. Methods Anteroposterior and frog-leg radiographs of 10 patients were evaluated by 18 reviewers on two occasions. Each patient had three sets of radiographs of different stages of disease: before treatment, 6–12 months after the initiation of treatment and at least 5 years after the treatment. Results Interobserver–intraobserver agreement rates for Catterall (set 1), Salter–Thompson (set 1), Herring (set 1), Catterall (set 2), Herring (set 2) and Stulberg (set 3) were 0.6203–0.6862, 0.6037–0.5758, 0.5955–0.4946, 0.5782–0.3864, 0.3878–0.1133 and 0.7912–0.7733, respectively. Conclusions The results of our study suggest the use of Catterall and Salter–Thompson systems prior to treatment and the Stulberg system at the end of the treatment at skeletal maturity. However, evaluation of the patients during the treatment period is still a dilemma and necessitates a new more reliable classification system.