Intra- and Interobserver Reliability of the Cobb Angle-Vertebral Rotation Angle-Spinous Process Angle for Adolescent Idiopathic Scoliosis.

STUDY DESIGN: A reliability analysis of Cobb angle, vertebral rotation (VR), and spinous process angle (SPA) measurements in adolescent idiopathic scoliosis. OBJECTIVE: To determine the intra- and interobserver reliability of semi-automated digital radiograph measurements. SUMMARY OF BACKGROUND DATA: Cobb angle measurements on posteroanterior radiographs are commonly used to determine the severity of scoliosis. Vertebral rotation helps assess scoliosis 3-dimensionally and has a role in predicting curve progression. Recent studies have shown that the spinous process angle is a useful parameter in assessing scoliosis when using ultrasound imaging. Because the reliability of SPA measurements on radiographs has yet to be determined, it is important to compare the reliability of these 3 parameters (Cobb angle, VR, and SPA) using a computer assisted semi-automated method. METHODS: Sixty posteroanterior radiographs of patients with adolescent idiopathic scoliosis were obtained and measured twice by 3 observers who were blinded to their previous measurements, using an in-house developed program. Measurements were obtained using a semi-automated method to minimize variability resulting from observer reliability. The intra- and interobserver reliabilities were analyzed using intra-class correlation coefficients (ICCs) as well as Bland-Altman's bias and limits of agreement. RESULTS: Over 350 (intra) and 90 (inter) sets of curves with an average Cobb angle of 26° ± 9° (range, 10° to 44°) were compared for each parameter. Intra-observer reliabilities for each parameter were excellent (ICC[2,1], .82; 1.00), with mean absolute differences under 3° between most measurements. Interobserver reliability (ICC[2,1], .72; .95) was mostly good to excellent, with mean absolute differences ranging from 2.0° to 5.6°. CONCLUSIONS: Both the intra- and interobserver assessment of the Cobb, VR, and SPA from the semi-automated measurements had clinically acceptable reliability ranges and may be considered for clinical implementation. Additional studies will be conducted to determine the accuracy and sensitivity to change of these scoliosis severity measurements.