Correlation between clinical scoring systems and quantitative MRI parameters in degenerative lumbar spinal stenosis.

Background: Many quantitative MRI parameters and clinical scores have been used patients with lumbar spinal stenosis (LSS). However, the correlation between clinical scores and MRI parameters is not very clear. The objective of the study was to find out the correlation between commonly used clinical scoring systems and quantitative MRI parameters.

Methods: Eighty two patients (more than 40 years) with clinical and radiological characteristics of degenerative LSS completed 10 clinical questionnaires which included Oswestry disability index (ODI), Swiss spinal stenosis (SSS) questionnaire, Quebec pain disability scale (QPD), Visual analogue scale (VAS), modified Japanese orthopaedic association scale (mJOA), Pain disability index (PDI), Short form health survey (SF-36), Self-paced walking test (SPWT), Euro quality of life-5D (EQ-5D) and Neurogenic claudication outcome score (NCOS). Lumbosacral MRIs were performed and 8 quantitative parameters namely transverse & AP diameter of dural sac (TDD, APDD), anteroposterior diameter of spinal canal (APDS), ligamentous interfacet distance (LID), mid-sagittal diameter of thecal sac (MSDT), cross-sectional area dural sac (CSAD), lateral recess depth and angle (LRD, LRA) were measured at the maximum stenotic level at the level of the disc. The clinical and radiological parameters were then statistically analysed.

Results: There were 51 females and 31 males in the study with a mean age of 53.0253.02 ± 9.18 years. NCOS score had a moderate correlation with CSAD, LID and TDS (p<0.05,0.7>r ≥ 0.3). MSDT showed a moderate negative correlation with SSS, VAS, ODI and QPD (p<0.05,0.7>r ≥ 0.3). LRD had a moderate negative correlation with ODI and VAS score (p<0.05,0.7>r ≥ 0.3). LRA had a moderate correlation with the EQ-5D and ODI (p<0.05,0.7>r ≥ 0.3). The CSAD had a moderate negative correlation with PDI (r = -0.383, p = 0.000). For all other comparisons, there was poor or no correlation. MRI parameters showed poor or no correlation with most of components of SF-36 score.

Conclusions: A poor correlation or no correlation was noted for most of the MRI parameters when compared to commonly used clinical scores. Hence, poor MRI's don't necessarily mean poor clinical scores in LSS. The role of MRI parameters should be supplementary and overreliance on them in LSS management should be avoided.