Inter- and Intra-observer Reliability of MRI for Lumbar Lateral Disc Herniation

BACKGROUND: The authors analyzed inter- and intra-observer agreement with respect to interpretation of simple magnetic resonance T1- and T2-weighted axial and sagittal images for the diagnosis of lumbar lateral disc herniation, including foraminal and extraforaminal disc herniations. METHODS: Forty-two patients in whom lumbar lateral disc herniation was suspected or confirmed by simple magnetic resonance imaging at one institute between May 2003 and December 2004 were included. The magnetic resonance images consisting of T1- and T2-weighted axial and sagittal images, and these were reviewed blindly and independently by three orthopaedic spine surgeons in a random manner. The images were interpreted as positive or negative for lateral disc herniation on 2 different occasions 3 months apart. Results were analyzed using Cohen's kappa statistic, and strengths of agreements were determined using the Landis and Koch criteria. RESULTS: The kappa values for inter-observer agreement averaged 0.234 (0.282, 0.111, and 0.308 respectively) on the first occasion, and 0.166 (0.249, 0.111, and 0.137 respectively) on the second occasion, with an overall mean value of 0.200. Thus, the strength of agreement was only slight-to-fair according to the Landis and Koch criteria. Kappa values for intra-observer agreement averaged 0.479 (0.488, 0.491, and 0.459 respectively), indicating moderate agreement. CONCLUSIONS: The present study indicates that simple magnetic resonance imaging is not a reliable imaging modality for diagnosing lumbar lateral disc herniation. Another imaging study with improved diagnostic values should be developed to diagnose this pathologic finding.

Diagnosis of Lumbar Lateral Disc Herniation: Value of Magnetic Resonance Imaging Revisited / 대한척추외과학회지

STUDY DESIGN: We retrospectively evaluated the value of magnetic resonance imaging (MRI) in the diagnosis of lumbar lateral disc herniations. SUMMARY OF LITERATURE REVIEW: MRI is known to be a reliable study for the diagnosis of a lumbar disc herniation. However, recent studies of its diagnostic value for lateral disc herniation have been rare. OBJECTIVES: We aimed to assess the diagnostic value of simple MRI, to determine the need for additional imaging studies and to investigate mimicking lesions. MATERIALS AND METHODS: In a lateral herniation group composed of 21 cases, including 10 foraminal and 11 extraforaminal herniations, the diagnostic value of simple MRI was evaluated, and the potential requirement for additional studies investigated. In a mimicking lesion group(5 cases), the entity of each lesion was identified. RESULTS: All 10 foraminal disc herniation cases were able to be confirmed with simple MRI, six of which were confirmed using sagittal images alone. In contrast, for the eleven extraforaminal disc herniations, sagittal MR images were not at all helpful in the diagnosis; however, six(55%) were confirmed from axial images, but the other five could not be confirmed until additional studies, such as enhanced MRI(4 cases), 1 mm-sliced CT (1) and CT-discography (3), were carried out. All 5 mimicking lesions were upper endplates of the lower vertebrae. CONCLUSIONS: Simple MRI is useful in the diagnosis of foraminal herniations, but not so helpful for extraforaminal herniations; particularly, sagittal images are of little use. Therefore, whenever a patient complaining of severe radiating pain presents with no causative finding on simple MRI, the extraforaminal regions on the axial images should be diligently scrutinized again, and additional studies considered when necessary. Conversely, mimicking lesions, such as an upper endplate, should be differentiated when a lateral disc herniation is suspected.