Determining Hinge Abduction in Legg-Calvé-Perthes Disease: Can We Reliably Make the Diagnosis?

Although hinge abduction is recognized as an important finding in children with Legg-Calvé-Perthes disease, variable diagnostic criteria exist. The purpose of this study was (1) to test the interobserver and intraobserver agreement of the current definition of hinge abduction and (2) to develop consensus regarding key diagnostic features that could be used to improve our diagnostic criteria.

Four orthopaedic surgeons with subspecialty pediatric hip interest independently assessed 30 randomly ordered cases of Legg-Calvé-Perthes disease. Each case included 2 fluoroscopic images of hip arthrograms (anteroposterior and abduction views). Surgeons graded the cases in a binary manner (hinge/no-hinge) on 2 separate occasions separated by a 4-week interval. Following reliability testing and comprehensive review of the literature, consensus-building sessions were conducted to identify key diagnostic features. Surgeons then regraded a new series of cases. Interobserver and intraobserver agreement between first/second and third/fourth readings were assessed using the Fleiss κ.

Interobserver κ for hinge abduction between the first and second surveys was 0.52 (with 0.41 to 0.60 considered moderate agreement), compared with 0.56 for the third and fourth surveys. First and second reading intraobserver agreement ranged from 0.59 to 0.83 compared with 0.75 to 1.00 for third and fourth reading. Consensus sessions identified several key diagnostic factors including: adequate visualization of the labral contour and ability of the lateral epiphysis to slip below the chondrolabral complex in abduction. Medial dye pooling, often due to asphericity of the femoral head, was not found to be a useful diagnostic criterion.

Despite a combined experience of over 70 years among the reviewers, we found just slightly better than 50:50 agreement in what constitutes hinge abduction. Consensus discussions did improve our agreement but these modest changes emphasize how difficult it is to develop reliable diagnostic criteria for hinge abduction. As a result, we caution against using hinge abduction as an inclusion criteria or outcome measure for research purposes, as the diagnostic agreement can be inconsistent.

Level III-diagnostic study.