Interobserver and intraobserver reliability of determining the deformity angular ratio in severe pediatric deformity...

Cross-sectional reliability study.

The deformity angular ratio (DAR) is a means of quantifying magnitude of the coronal (C-DAR) and sagittal (S-DAR) plane of deformity curves to produce a total DAR (T-DAR). It has been shown to predict the risk of spinal cord monitoring alerts and actual neurologic deficits. We sought to assess the reliability of determining the C-DAR and S-DAR among pediatric spinal deformity surgeons.

Twelve preoperative anterior-posterior (AP) and lateral X-rays from the Fox multi-center pediatric deformity study were de-identified and sent to7 pediatric spinal deformity surgeons. Each surgeon measured: coronal/sagittal Cobb angles, upper/lower endplate vertebrae (UEV/LEV), apices, and number of vertebrae included in the main curve. The C-DAR and S-DAR were then calculated by dividing the Cobb angles by the number of vertebrae included in the curve. Intra- and interobserver reliability was calculated using interclass correlation (ICC).

The mean C-DAR was 14.9 (range 1.3-51.5) with a mean Cobb angle of 88.8° (range 15.0-163.0) over a mean of 7.5 (range 2.0-14.0) levels. The mean S-DAR was 8.6 (range 1.0-19.6), with a mean Cobb angle of 68.0° (range 10.0-137.0) over a mean of 7.5 (range 3-11) levels. The intraobserver reliability of the C-DAR was ICC = 0.908 (range 0.846-0.960) and the S-DAR 0.914 (range 0.815-0.961). The interobserver reliability of the C-DAR was ICC = 0.868 (range 0.846-0.938), and the S-DAR was ICC = 0.848 (range 0.815-0.961). Despite poor reliability among UEV, LEV, and apex selection (aggregated range 0.340-0.724), the C-DAR and S-DAR were demonstrated to be reliable in our study CONCLUSIONS: Reliability was good to excellent for C-DAR and S-DAR, despite poor to moderate reliability among UEV, LEV, and apex selection. These data support the use of the C-DAR, S-DAR, and combined T-DAR as a means of quantifying deformity magnitude.