External Validation of the Spinal Orthopedic Research Group Index for Spinal Epidural Abscess 90-day Mortality in a Geographically Remote Population.

Abstract

Retrospective cohort study. To externally validate the Spinal Orthopaedic Research Group (SORG) index for predicting 90-day mortality from Spinal Epidural Abscess (SEA) and compare its utility to the 11-item modified frailty index (mFI-11) and Charlson Comorbidity Index (CCI). Providing a mortality estimate may guide informed patient and clinician decision-making. A number of prognostic tools and calculators are available to help predict the risk of mortality from SEA, including the SORG index, which estimates ninety-day post discharge mortality. External validation is essential before wider use of any clinical prediction tool. Patients were identified using hospital coding. Medical and radiological records were used to confirm the diagnosis. Mortality data, and data to calculate the SORG index, mFI-11 and CCI was collected. Area under the curve (AUC) and calibration plots were used to analyse. 150 patients were included: 58 female (39%), with median age 63 years. Fifteen deaths (10%) at 90-days post discharge and 20 (13%) at one-year. The mean SORG index was 13.6%, mean CCI 2.75, and mean mFI-11 was 1.34. The SORG index (P=0.0006) and mFI-11 (P<0.0001) were associated with 90-day mortality. AUC for SORG, mFI-11, and CCI were 0.81, 0.84, and 0.49, respectively. The calibration slope for the SORG index showed slight overestimation in the middle ranges of the predicted probability, more so than mFI-11, and was not well-calibrated over the higher ranges of predicted probability. This study externally validated the SORG index, demonstrating its utility in our population at predicting 90-day mortality, however, it was less well calibrated than the mFI-11. Variations in algorithm performance may be a result of difference in socioethnic composition and health resource between development and validation centres. Continued multicentre data input may help improve such algorithms and improve their generalisability.