MRI Spondylodiscitis Severity Index: A Novel Scoring System Based on Contrast-Enhanced MRI of the Spine for Predicting the Clinical Course and Severity of Pyogenic Spondylodiscitis

Abstract

ObjectiveThis study aimed to validate a new MRI-based scoring system for pyogenic spondylodiscitis (PS) using contrast-enhanced MRI at admission. The system was designed to classify disease severity and provide a quantifiable tool for risk stratification and treatment planning.Materials and methodsWe retrospectively analysed 60 patients with clinically and laboratory-confirmed PS at a single tertiary hospital between September 2016 and September 2024. A novel scoring system, the MRI-SSI, evaluated vertebral destruction, signal abnormalities, endplate erosion, and epidural or paravertebral extension. The scores (0-9, 10-19, >= 20) categorised disease as mild, moderate, or severe. Two experienced radiologists scored the images, and we recorded hospitalisation duration and the need for surgery. Statistical analysis included independent samples t-tests, ANOVA, and interobserver agreement was assessed using Cohen's kappa and intraclass correlation coefficients (ICCs).ResultsOf the 60 patients, 17 (28.3%) required surgical intervention. Patients needing surgery had significantly higher MRI-SSI scores than those managed conservatively (18.58 vs. 13.86; p = 0.004). The scoring system had a strong predictive value for surgical requirement (AUC = 0.720), with 58% sensitivity and 88% specificity. Interobserver agreement for the total MRI-SSI score was excellent (ICC = 0.906).ConclusionThe MRI-SSI accurately predicted hospitalisation duration and surgical needs in PS patients. This tool, based solely on admission MRI findings, can refine initial treatment decisions, leading to more targeted resource utilisation and improved overall patient outcomes.