Status epilepticus severity score as a predictor for the length of stay at hospital for acute-phase treatment in convulsive status epilepticus. (May 2020)

• Record Type: Journal Article
• Authors: Sato, Kenichiro; Arai, Noritoshi; Takeuchi, Sousuke
• Is Part Of: Journal of clinical neuroscience. Volume 75(2020)
• Page Start: 128
• Page End: 133
• Publication Date: 2020-05
• Subjects: Convulsive seizure -- Length of stay -- Machine learning -- Status epilepticus -- Status epilepticus severity score; Brain -- Surgery -- Periodicals; Neurosciences -- Periodicals; Nervous system -- Surgery -- Periodicals; Brain -- surgery -- Periodicals; Neurosurgical Procedures -- Periodicals; Neurosciences -- Periodicals; Electronic journals; 616.8
• DOI: 10.1016/j.jocn.2020.03.004
• Languages: English
• ISSNs: 0967-5868

Abstract

Highlights: Status epilepticus severity score (STESS) for the length of stay due to seizure. Retrospectively, STESS was positively correlated with the length of stay (LOS). Predictive performance for LOS improved by incorporating STESS. Although further refinement needed, STESS may be used as a rough indicator for LOS. Abstract: To date, hospital length of stay (LOS) determinants for convulsive status epilepticus's (CSE) acute-phase treatment have not been sufficiently investigated, as opposed to those for status epilepticus's (SE) outcome predictors, such as status epilepticus severity score (STESS). Here, we aimed at assessing the significance of STESS in the LOS in patients with CSE. We retrospectively reviewed consecutive adult patients with CSE who were transported to the emergency department of our urban tertiary care hospital in Tokyo, Japan. The study period was from August 2010 to September 2015. The primary endpoint was the LOS of patients with CSE who were directly discharged after acute-phase treatment, and survival analysis for LOS until discharge was conducted. As a result, among 132 eligible patients with CSE admitted to our hospital, 96 (72.7%) were directly discharged with a median LOS of 10 days (IQR: 4–19 days). CSE patients with severe seizures, represented by higher STESS (≥3), had a significantly longer LOS after adjustments with multiple covariates ( p = 0.016, in restricted mean survival time analysis). Additionally, prediction for the binomial longer/shorter LOS achieved better performance when STESS was incorporated into the prediction model. Our findings indicate that STESS can also be used as a rough predictor of longer LOS at index admission of patients with CSE.