Comparison of Cobb angle measurement of scoliosis radiographs with preselected end vertebrae: traditional versus digital acquisition.

Abstract

Prospective study. To compare variability in Cobb angle between digitally and traditionally acquired scoliosis radiographs. Previous studies have shown that the 95% confidence interval for Cobb angle can vary from 2.6 degrees to 8.8 degrees. No study directly comparing Cobb angles measured from traditional and digitally acquired radiographs has been reported. A spine model simulating 25 single right thoracic curves (range, 20 degrees-60 degrees) was imaged using traditional and digital techniques. Traditional films and miniaturized printed digital films were each measured twice manually. Digital films were also measured twice using computer imaging software. Overall mean angle and (95% confidence interval) were 41.7 degrees (39.1 degrees-44.3 degrees) for traditional, 40.6 degrees (37.4 degrees-43.8 degrees) for digital, and 39.7 degrees (36.3 degrees-43.1 degrees) for computer measurements. Overall correlation was 0.994 for traditional and digital, 0.987 for traditional and computer, and 0.993 for digital and computer. Fixed effect model analysis demonstrated that, although a statistically significant difference existed between the 3 methods of measuring the Cobb angle (P < 0.0001), the difference between methods was less than 2 degrees. Any of the 3 evaluated methods of measurement can be used to measure Cobb angles. Additionally, the methods can be used interchangeably.