Epidural Abscess in Routine Preoperative MRI Reports Versus Documented Operative Findings, and Exploratory Mortality...

Background: Surgical decisions in spondylodiscitis rely on the preoperative magnetic resonance imaging (MRI) report, and postoperative mortality remains high. We examined how often routine MRI reports described an epidural abscess later documented at surgery and explored factors associated with mortality. Methods: Twenty-seven adults operated on for pyogenic or fungal spondylodiscitis at one centre (2017-2025) were studied retrospectively. Findings were abstracted from routine MRI reports and operative notes, with epidural pus documented at surgery as the reference. Mortality was explored with univariable Cox and receiver operating characteristic analyses. Results: Epidural pus was documented in 11 patients; none of their MRI reports described an epidural abscess (sensitivity 0%, 95% confidence interval [CI] 0-28.5%), although six reported paraspinal soft-tissue involvement. Eleven patients (40.7%) died during a median follow-up of 2.5 months. Age (hazard ratio 1.14 per year, 95% CI 1.00-1.30) and the age-adjusted Charlson Comorbidity Index (1.54 per point, 1.05-2.28) were associated with mortality and discriminated similarly (areas under the curve 0.741 and 0.736); the unadjusted index and the Pola, Homagk, Spinal Instability Neoplastic Score, and Spinal Instability Spondylodiscitis Score classifications were not. Conclusions: In this small retrospective cohort, routine MRI reports named no epidural abscess among patients in whom one was documented at surgery. Mortality was associated with age and age-adjusted comorbidity burden. These exploratory findings require prospective validation.