Abstract
Background: The degree of cervical foraminal stenosis on MRI scans may be measured and categorised using the Kim or modified Kim methods. These grading scales have not previously been validated in a cohort of patients awaiting surgery.
Objectives: To establish the normal foraminal and root diameters as well as the consistency of inter and intra-rater grading using the Kim and modified Kim grading systems in pre-operative surgical patients.
Methods: Asymptomatic cervical nerve roots and foramina demonstrated on the pre-operative MRI scans of adult surgical patients with cervical radiculopathy were measured and categorised by six raters using the Kim and modified Kim grading methods. Repeat “second pass” measurements were made by the same assessors on the same images a minimum of one month later.
Results: Foraminal diameters (mm) in asymptomatic foramina were C2/C3 (mean ± SD): 4.18 ± 1.44, C3/C4 2.96 ± 1.23, C4/C5 3.02 ± 1.19, C5/C6 3.15 ± 1.33, C6/C7 3.53 ± 1.36, C7/T1 3.93 ± 1.34. Nerve root diameters were C3 3.11 ± 0.87, C4 2.95 ± 0.77, C5 2.56 ± 0.73, C6 2.26 ± 0.76, C7 2.56 ± 0.82, C8 3.83 ± 0.86.
Inter-rater consistency was kappa [95% CI]: Kim 0.01 [0.00, 0.03], modified Kim 0.08 [0.05, 0.10]. Intra-rater consistency was kappa [95% CI]: Kim 0.81 [0.77, 0.86], modified Kim 0.69 [0.62, 0.76].
Conclusion: There was poor inter-rater consistency but good intra-rater consistency when assessing the severity of foraminal stenosis on axial T2 MRI scans. Foraminal diameter was narrowest at C3/C4 and C4/C5, whereas the smallest root diameter was C5/C6. Volumetric or oblique MR may improve consistency.
