Measurement of vertebral rotation in idiopathic scoliosis using the Perdriolle torsionmeter: a clinical study on intraobserver and interobserver error.

This study was designed to determine the reliability and accuracy of the Perdriolle torsionmeter. Fifty-four observers were divided into three groups according to their previous experience in the field of orthopaedics and the treatment of scoliosis. Each observer, on two separate occasions, measured the apical vertebral rotation on preoperative and postoperative roentgenograms of three idiopathic thoracic scoliotic curves using the torsionmeter. For all groups, there was no statistically significant difference (P > 0.05) between the average of the first and the average of the second measurements on vertebral rotation of any of the curves. In the curve with the highest frontal Cobb angle and vertebral rotation of more than 30 degrees, there was a statistically significant difference (P = 0.03) between the average measurements of the three groups of observers. We concluded that, as both intraobserver and interobserver error risks were insignificant, the torsionmeter can be accurately used by everyone in the fields of orthopaedics provided the vertebral rotation is not greater than 30 degrees and the curve is mild or moderate. The reliability and accuracy of the torsionmeter was found to be questionable at more than 30 degrees of vertebral rotation because of increased risk of interobserver error.