External validation of newly modified status epilepticus severity score for predicting mortality in patients with status epilepticus in a regional hospital in Taiwan

Abstract

PurposeTo evaluate and compare the performance of the status epilepticus severity score (STESS), modified STESS (mSTESS), and the newly modified STESS (nSTESS) in predicting mortality in patients with status epilepticus (SE) at a regional hospital in Taiwan. MethodsData were collected from 81 patients with SE, aged over 18 years at a regional medical hospital in Tainan from January 2012 to December 2022. SE were treated following the standard treatment protocol. Exclusion criteria included missing data, lack of adherence to the treatment protocol, and transfer to tertiary medical centers. Outcome measures included differences in characteristics between survivor and non-survivor groups, the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of STESS, nSTESS, mSTESS. Receiver operating characteristic (ROC) curve and area under curve (AUC) of scales were generated. Calibration with Hosmer–Lemeshow test was built as well. ResultsThe study found significant differences in seizure types (p = 0.015) and undergoing intubation (p = 0.017) between survivor and non-survivor groups. No significant differences were observed in age, gender, underlying diseases, or the category of antiseizure medications (ASMs) usage. The ROC curve for STESS, nSTESS and mSTESS showed similar predictive values of around 0.75, indicating moderate performance of prediction. The Hosmer-Lemeshow test showed no significant difference between real-world predictions and these three clinical scales. In the aspect of accuracy, sensitivity and specificity, nSTESS has similar overall accuracy as in STESS and mSTESS. ConclusionThis external validation study demonstrates the moderate performance of nSTESS in predicting mortality in SE patients at a regional hospital in Taiwan. These outcomes underscore the practical utility of these scales in clinical practice, with nSTESS demonstrating accuracy on par with the others. Further validation in larger, multicenter cohorts and other healthcare settings is necessary to fully confirm its predictive value.