Radiographic assessment of degenerative lumbar spinal stenosis: is MRI superior to CT?

To determine the reliability and dependability of magnetic resonance imaging (MRI) and computerized tomography (CT) in the assessment of lumbar spinal stenosis and correlate the qualitative assessment to both a quantitative assessment and functional outcome measures. Multiple studies have addressed the issue of CT and MRI imaging in lumbar spinal stenosis. None showed superiority of one modality.

We performed a standardized qualitative and quantitative review of CT and MRI scans of 54 patients. Intra-observer and inter-observer reliability was determined between three reviewer using Kappa coefficient. Agreement between the two modalities was analyzed. ODI and SF-36 outcomes were correlated with the imaging assessments.

Almost perfect intra-observer reliability for MRI was achieved by the two expert reviewers (κ = 0.91 for surgeon and κ = 0.92 for neuro-radiologist). For CT, substantial intra-observer agreement was found for the surgeon (κ = 0.77) while the neuro-radiologist was higher (κ = 0.96). For both CT and MRI the standardized qualitative assessment used by the two expert reviewers had a better inter-observer reliability than that between the expert reviewers and the general reporting radiologist, who did not utilize a standardized assessment system. When the qualitative assessment was compared directly, CT overestimated the degree of stenosis 20-35 % of the time (p < 0.05) while MRI overestimated the degree of stenosis 2-11 % of the time (p < 0.05). No correlation was found between qualitative and quantitative analysis with functional status.

This study directly demonstrates that MRI is a more reliable tool than CT, but neither correlates with functional status. Both experience of the reader and the standardization of a qualitative assessment are influential to the reliability.