Interobserver reliability of scoliometer measurements of rotational deformity in adolescent idiopathic scoliosis

Assessment of rotational deformity of spinal curvature is an important method of evaluating curves and in following the progression of adolescent idiopathic scoliosis. An evaluation of the interobserver reliability of scoliometer measurements was performed. We also iinvestigated whether interobserver reliability varied with the magnitude of the Cobb angle. Three clinicians, measured the axial rotation of thoracic and lumbar curves in degrees on fifty patients with adolescent idiopathic scoliosis during the same office visit. The Cobb angles were measured on the same day. Correlations were made by determining a Pearson correlation coefficient. Agreement between the raters was good with coefficients ranging from.83 to.89 (average=.86). The interobserver variability was 1.3 degrees and was not dependent on the magnitude of the Cobb angle. Correlation between degrees of rotation and Cobb angle was poor for each of the raters with average coefficient.55. These results demonstrate that interobserver reliability in use of the Scoliometer is good, not excellent. They also confirm that rotational deformity has no relation to the sagittal curve. The Scoliometer is a useful augmentation device in evaluating and planning treatment in scoliosis, but is not reliable in long term tracking of progression when used by multiple observers.