Reliabilities of and Correlations Among Five Standard Methods of Assessing the Sagittal Alignment of the Cervical Spine

Abstract

The reliabilities of and correlations among 5 standard methods of assessing cervical sagittal alignment were evaluated. To investigate the reliabilities of and correlations among 5 standard methods of assessing cervical sagittal alignment. Although various cervical sagittal alignment assessment methods are widely used, their relative reliability and intercorrelation have not been reported. From 442 lateral cervical radiographs, 40 with lordotic, 40 with straight or sigmoid, and 40 with kyphotic alignment were selected. Two orthopedic surgeons independently evaluated the sagittal alignment in each group twice using CCL, C1-C7 Cobb, C2-C7 Cobb, sagittal tangent, and the Ishihara methods. Intraobserver and interobserver reliabilities were confirmed and the correlations among the 5 methods were measured. Intraobserver and interobserver reliabilities for all 5 methods were good. In the lordotic group, the correlations among all 5 methods were consistently strong (r = 0.731 to 0.922). In the straight or sigmoid group, the correlations were weak to moderate among the CCL, C2-C7 Cobb, sagittal tangent, and Ishihara methods but tended to be weak between these 4 methods and the C1-C7 Cobb method (r = -0.245 to 0.777). In the kyphotic group, the correlations were also weak to moderate among the same 4 methods, and were statistically insignificant between them and the C1-C7 Cobb. The correlations among the CCL, C1-C7 Cobb, C2-C7 Cobb, sagittal tangent, and Ishihara methods are strong when lordosis is retained; otherwise, they are moderate to poor. In the kyphotic group, C1-C7 Cobb has no significant correlation with the other 4 methods.