Purpose: To evaluate intra- and interobserver agreement for the interpretation of lumbar 1.5-T magnetic resonance (MR) images in a community setting.
Materials and methods: The study design was approved by the Institutional Review Board of the Ramón y Cajal Hospital. According to Spanish law, for this type of study, no informed consent was necessary. Five radiologists from three hospitals twice interpreted lumbar MR examination results in 53 patients with low back pain, with at least a 14-day interval between assessments. Radiologists were unaware of the clinical and demographic characteristics of the patients and of their colleagues' assessments. At the second assessment, they were unaware of the results of the first assessment. Reports on Modic changes, osteophytes, Schmorl nodes, diffuse defects, disk degeneration, annular tears (high-signal-intensity zones), disk contour, spondylolisthesis, and spinal stenosis were collected by using the Spanish version of the Nordic Modic Consensus Group classification. The kappa statistic was used to assess intra- and interobserver agreement for findings with a prevalence of 10% or greater and 90% or lower. kappa was categorized as almost perfect (0.81-1.00), substantial (0.61-0.80), moderate (0.41-0.60), fair (0.21-0.40), slight (0.00-0.20), or poor (<0.00).
Results: Endplate erosions and spondylolisthesis were observed in less than 10% of images. Intraobserver reliability was almost perfect for spinal stenosis; substantial for Modic changes, Schmorl nodes, disk degeneration, annular tears, and disk contour; and moderate for osteophytes. Interobserver reliability was moderate for Modic changes, Schmorl nodes, disk degeneration, annular tears, and disk contour; fair for osteophytes; and poor for spinal stenosis.
Conclusion: In conditions close to those of clinical practice, there was only moderate interobserver agreement in the reporting of findings at 1.5-T lumbar MR imaging.
Supplemental material: http://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.09090706/-/DC1.
(c) RSNA, 2010