The Modic classification: Inter- and intraobserver error in clinical practice

Abstract

<h4>Study design</h4>A reliability study of the Modic classification.<h4>Objective</h4>To determine the reliability and reproducibility of the Modic classification for lumbar vertebral marrow changes.<h4>Summary of background data</h4>In 1988, Modic et al described two degenerative stages of vertebral marrow and endplate morphology. These were type 1 (inflammatory phase) and type 2 (a fatty phase). Later in 1988, he added a third variety: type 3 where there was marked sclerosis adjacent to the endplates. No formal reliability or reproducibility studies had been performed on this Modic classification.<h4>Methods</h4>This study involved five independent observers of differing spinal experience using the Modic classification to grade 50 sagittal T1- and T2-weighted magnetic resonance imaging scans. The observers repeated the assessment at 3 weeks. Intra- and interobserver reliabilities were assessed using kappa statistics.<h4>Results</h4>There were 7 type 1, 40 type 2, 1 type 3, and 2 normal levels. The individual intraobserver agreement was substantial or excellent with kappa values ranging from 0.71 to 1. The overall interobserver agreement was excellent with a kappa value of 0.85. There was complete agreement in 78% of the levels, a difference of one type in 14% and a difference of two or more in 8% of levels. The level of experience of the observer did not correlate with a better score.<h4>Conclusions</h4>We have shown that the classification is both reliable and reproducible. It is simple and easy to apply for observers of varying clinical experience. We therefore recommend its use in clinical research and practice.