Christian Hansen

Although reproducibility in reading MRI images amongst radiologists and clinicians has been studied previously, no studies have examined the reproducibility of inexperienced clinicians in extracting pathoanatomic information from magnetic resonance imaging (MRI) narrative reports and transforming that information into quantitative data. However, this process is frequently required in research and quality assurance contexts. The purpose of this study was to examine inter-rater reproducibility (agreement and reliability) among an inexperienced group of clinicians in extracting spinal pathoanatomic information from radiologist-generated MRI narrative reports.

Modic changes are of increasing interest, however their age and gender prevalence are not well described. To date, the associations between Modic changes and other common vertebral pathologies have only been described in small samples (n < 100). Our aim was, in a large dataset of people with low back pain, to (1) describe the prevalence of a range of spinal pathoanatomies, and (2) examine the association between Modic changes and stages of intervertebral disc (IVD) pathology. Common pathologies were coded from the lumbar spine MRIs from 4,233 consecutive people imaged while attending a publicly-funded secondary care outpatient facility in Denmark. Prevalence data were calculated by pathology and by vertebral level. Prevalence was also calculated by age and gender categories for Modic changes. The association between stages of IVD pathology (degeneration, bulge, herniation) and Modic changes at L4/5 and L5/S1 was expressed using prevalence ratios (PR) and 95% confidence intervals. The prevalence of Modic changes and IVD pathology were greater in L4/5 and L5/S1, compared with the upper lumbar spine. There was no significant gender difference in prevalence of Modic changes (p = 0.11). The prevalence of IVD disc pathology occurring concurrently with Modic changes ranged from 11.5 to 17.5% (Type 1), 8.5 to 12.7% (Type 2) and 17.1 to 25.6% (Type 1 and/or 2) while the prevalence occurring in the absence of Modic changes ranged from 0.5 to 6.3% (Type 1), 0.3 to 4.9 (Type 2), 0.8 to 9.7% (Type 1 and/or 2). The associated PR for IVD pathology occurring concurrently with Modic changes ranged from 1.8 to 29.2 (p < 0.05). The highest PR (29.2) was between degeneration and Modic changes, indicating that it is rare for Modic changes to occur without disc degeneration. Spinal pathoanatomy was common in this population, particularly IVD pathologies, and a consistent trend of a relatively greater prevalence in the lower lumbar spine was identified. Modic changes were more likely to be present among individuals with IVD pathology than without, which may implicate mechanical factors as being one aetiological pathway for Modic changes, although other hypotheses may equally explain this association.