Posterior disk displacement: Morphologic assessment and measurement reliability - Lumbar spine

Background: Magnetic resonance imaging is often used to assess for disk dis placement after manipulation, but limited information about the true incide nce of iatrogenic herniations exists. To design a study that evaluates for a causal relationship, preliminary data must be obtained relating to the si ze of different types of disk displacement. The reliability of chiropractic radiologists in assessing disks and a comparison of different measuring de vices should also be evaluated. Objective: To identify average measurements for normal and displaced disks and to assess the reliability of measurements by chiropractic radiologists. Study Design: Intraobserver and interobserver reliability study assessing d isk displacement on magnetic resonance scans. Methods: Three evaluators assessed the disks on 122 magnetic resonance scan s from two imaging centers. Six categories were graded, and digitizer and r uler measurements were compared. Forty four scans were reassessed for intra observer agreement. Intraobserver and interobserver variations were measure d with intraclass correlation coefficient and kappa statistical analysis. M easurement device correlation was assessed with Pearson's r. Results: Clear size differences between different types of disk displacemen t were noted. Interexaminer measurement reliability was 0.78 to 0.84. Agree ment concerning the presence of disk displacement was 85% (kappa = 0.68), a nd the classification of disk displacements was 76% (kappa = 0.60). Intraex aminer measurement reliability was 0.40 to 0.49. Intraexaminer agreement co ncerning the presence of disk displacement was 76% (kappa = 0.52), and the classification of disk displacements was 62% to 69% (kappa = 0.38 to 0.46). Normal versus bulged disk distinctions demonstrated the most disagreement. The ruler and digitizer correlation coefficient was 0.968. Conclusions: Different disk types demonstrated distinct size averages. Inte rexaminer agreement was good concerning disk assessment and measurements. I ntraexaminer agreement was lower than expected. A millimetric ruler is an a cceptable alternative to digital measurement devices.