Reliability of a Magnetic Resonance Imaging-based Grading System for Cervical Intervertebral Disc Degeneration

Abstract

This was a radiographic reliability study of a novel grading system for cervical intervertebral disc degeneration. This study aimed to develop and test the reliability of a reproducible grading system for cervical intervertebral disc degeneration on the basis of the routine magnetic resonance imaging (MRI). Cervical disc degeneration is common after middle age, and the morphology of cervical disc degenerative disease has often been studied using MRI. There are few specific MRI-based grading systems for cervical intervertebral disc degeneration despite the clinical importance of this problem. This study proposes a novel reproducible grading system for cervical disc degeneration and demonstrates the reliability of this classification scheme. A grading system for cervical intervertebral disc degeneration was developed based on relevant previous literature. MRI grading of 300 cervical intervertebral discs using T2-weighted sagittal images was performed by 4 spinal surgeons (observers) in a blinded fashion. Intraobserver and interobserver reliabilities were assessed by calculating kappa statistics. Grade I degeneration was observed in 27 discs (9.0%); grade II, in 56 (18.7%); grade III, in 124 (41.3%); grade IV, in 67 (22.3%); and grade V, in 26 (8.7%) discs. Kappa coefficients for intraobserver and interobserver agreements ranged from substantial to excellent (intraobserver, 0.907 to 0.950 and interobserver, 0.730 to 0.826). Complete agreement was obtained, on an average, in 72.1% of the discs. A difference of 1, 2, and 3 grades was observed in 22.4%, 3.3%, and 0.2% discs, respectively. This grading system is comprehensive and easily applicable with sufficient reproducibility. It can be used as a common nomenclature for research and discussions.