Agreement Between Radiologist Grading of Cervical Spinal Stenosis on Contemporaneous MRI Versus CT Myelogram and Correlations With Quantitative Measurements of Canal Stenosis and Cord Compression.

Methods: Cross-sectional study. Objective: To evaluate agreement between preoperative cervical magnetic resonance imaging (MRI) and CT myelogram (CTM) grading of cervical spinal stenosis on official radiology reports and correlation with quantitative measures.

Background: Accurate assessment of cervical spinal stenosis is critical for diagnosis and management cervical spondylotic myelopathy (CSM). Radiology reports are directly accessible by patients and used for insurance preauthorization increasing the importance of accurate grading systems.

Methods: Fifty consecutive patients with contemporaneous preoperative cervical MRI and CTM were identified. Grading (no stenosis, mild, moderate, and severe) was recorded from radiology reports. Quantitative measurements of anteroposterior (AP) diameter of the spinal canal and AP diameter of the spinal cord at the point of maximal compression for each anatomic level were performed.

Results: There were 20 males and 30 females, with a mean age of 59.6 years and a mean body mass index of 32.39 kg/m2. A total of 191 cervical levels were measured. There was poor agreement between the axial (k=0.528) and sagittal (k=0.516) MRI and CTM stenosis grades. On both the axial and sagittal MRI, the radiologist's stenosis grade was moderately correlated with the AP spinal canal diameter and space available for the cord, yet weakly correlated with the cord:canal ratio. On the axial and sagittal CTM, the stenosis grade was weakly correlated with AP spinal canal diameter, space available for the cord, and cord:canal ratio. Quantitative measures on CTM versus MRI showed moderate correlations on axial views for spinal canal (0.669) and cord (0.630) diameter, moderate correlations on sagittal views for canal diameter (0.640) and cord:canal ratio (0.601) and a strong correlation for spinal cord diameter.

Conclusions: There was poor agreement between radiologists grading of cervical spinal stenosis on contemporaneous MRI and CTM. Grading on both modalities was weakly or moderately correlated with quantitative measurements of stenosis.