Intra- and interobserver reproducibility of vertebral endplate signal (Modic) changes in the lumbar spine: the nordic modic consensus group classification

Abstract

The lumbar vertebral endplate is considered a potential cause of specific low back pain. However, in relation to future research, there is need for a reliable and detailed magnetic resonance imaging (MRI) protocol to be used in the evaluation of vertebral endplate signal changes. To assess the intra- and interobserver reliability of the "Nordic Modic classification" protocol. MRI scans of 50 individuals representative of the general Danish population aged 40 were evaluated by two observers. Criteria for grading the changes were developed by the Nordic Modic Consensus Group. After consensus was established, all 50 MRI examinations were evaluated independently by each observer. Intraobserver reliability was assessed by re-evaluation of the 50 examinations by one of the observers. Kappa statistics were used to calculate agreement. Intra- and interobserver agreement of the evaluation of variables describing vertebral signal changes, i.e. Modic type, location, volume, maximum height, and endplate area, were all found to have substantial to almost perfect agreement. The evaluation of osteophytes was found to be reliable, whereas the evaluation of localized endplate defects and irregular endplates had only moderate agreement. The evaluation of development over time was found to have substantial intraobserver agreement but only moderate interobserver agreement. As expected, intraobserver agreement was generally better than interobserver agreement. In this study, we found convincing reproducibility of a detailed evaluation protocol of vertebral endplate signal changes, the "Nordic Modic Classification." The authors recommend that the evaluation protocol should be used in future studies investigating vertebral endplate signal changes.