Interrater and intrarater reliability of the Kuntz et al new deformity classification system

Kuntz et al recently introduced a new system for classifying spinal deformities. This classification of spinal deformity was developed from age-dependent deviations from the neutral upright spinal alignment.

To determine the interobserver and intraobserver reliabilities of the new Kuntz et al system for classifying scoliosis.

Fifty consecutive patients were evaluated. Three observers independently assigned a major structural curve, minor structural curve, curve type, apical vertebral rotation, spinal balance, and pelvic alignment to each curve following the guidelines described by Kuntz et al. Assignment of the curves was repeated 4 weeks later, with the curves presented in a different blinded order. The Kendall W and Holsti agreement coefficients were used to determine the interobserver and intraobserver agreement.

The intraobserver value of agreement for all parameters was 0.85 (range, 0.28-1.0), and the mean Kendall W coefficient was 0.89 (range, 0.5-0.97), demonstrating perfect reliability. The interobserver agreement averaged 0.7 (range, 0.251-1.0). The mean Kendall W coefficient was 0.67 (range, 0.19-1.0), demonstrating substantial reliability. The average time for classification of 1 curve was approximately 8.4 minutes.

The new Kuntz et al deformity classification system is comparable to the Lenke et al system in terms of reliability. However, the Kuntz et al classification system provides no recommendations for surgical interventions. It is more complex and time-consuming and therefore may be of limited value in daily clinical practice.