The accuracy of anterior-posterior measurements in the assessment of spinal canal compromise in burst fractures.

Abstract

The amount of spinal canal compromise is one of the primary criteria employed to specify treatment of thoracolumbar burst fractures. Canal compromise has been calculated utilizing measurement of anterior-posterior canal or by calculating the area of the involved spinal canal on CT scans. Since these measurements are critical to surgical decision making, we have compared anterior-posterior measurements with area measurements to assess the variability between these two techniques. Three mm axial CT scans from ten patients with thoracolumbar burst fractures were studied. The axial CT images at the midpoints of the vertebral bodies at the levels above and below the fracture level, and the axial CT image demonstrating the most significant canal compromise at the fracture level were digitized using an optical scanner. Using the image 1.43 software, the anterior-posterior dimensions and area dimensions of each axial CT scan were obtained and the amount of canal compromise calculated. Linear measurements consistently overestimate the extent of spinal canal compromise as compared to the true area of canal compromise (6-46%, p < 0.005). This amount of variation did not correlate with the level, patient age or the extent of canal compromise. This preliminary study suggests that the two measurement techniques used to guide surgical decision making are not comparable. Further evaluation comparing the use of linear and area measurements is warranted.