A Radiological Comparison of Qualitative and Quantitative Grading Systems (Schizas Classification and Stenosis Ratio) For Lumbar Spinal Stenosis in a Low-Income Setting

Abstract

Introduction: In low-resource neurosurgical settings, the choice of a reliable and efficient imaging tool for grading degenerative lumbar spinal stenosis (DLSS) is critical. The qualitative Schizas morphological classification and the quantitative Stenosis Ratio (SR) are two established methods, but their direct correlation and practical interchangeability in such environments are not well-documented. This study aimed to determine the correlation and inter-rater reliability between the Schizas grading system and the Stenosis Ratio in patients with symptomatic DLSS in a low-income country. Methods: A retrospective cross-sectional radiological analysis was conducted on 50 consecutive patients scheduled for surgery. The researcher and two neuroradiologists independently graded the most stenotic axial T2-weighted MRI level using the Schizas system (Grades A-D) and measured the SR (dural sac area / bony canal area). Inter-rater reliability was assessed using Cohen’s Kappa and Intraclass Correlation Coefficient (ICC). Correlation between the two grading systems was evaluated using Spearman’s rank correlation (ρ). Results: The cohort had a mean SR of 0.39 (SD 0.19). Schizas Grade C was the most prevalent (44%). Inter-rater reliability was excellent for both systems: SR (ICC = 0.984, p<0.001) and Schizas grade (κ = 0.90). A very strong, statistically significant negative correlation was  found between the Schizas grade and the SR (ρ = -0.821, p < 0.001), indicating convergent  validity. Conclusion: This study demonstrates a strong radiological correlation and excellent reliability between the Schizas classification and the Stenosis Ratio. The findings validate the Schizas system as an efficient, reliable, and software-free tool for routine anatomical grading of DLSS in low-resource settings. The quantitative SR remains a precise option for complex cases, but the high concordance supports the primary use of the simpler qualitative method to optimise workflow in constrained environments.