Early Predictors of Poor Neurological Outcomes in Pediatric Status Epilepticus: Initial Glasgow Coma Scale (GCS) Score, Full Outline of UnResponsiveness (FOUR) Score, and Seizure Duration.

Background and objective Status epilepticus (SE) is one of the most critical medical emergencies that may result in significant morbidity and mortality if not addressed in a timely and effective manner. Early identification of prognostic predictors is essential for determining outcomes and guiding management. Clinical tools such as the Glasgow Coma Scale (GCS) score, Full Outline of UnResponsiveness (FOUR) score, and seizure duration are commonly used, but their independent prognostic value in pediatric SE remains uncertain. This study aimed to evaluate the association between initial GCS, FOUR score, and seizure duration and neurological outcomes in children with status epilepticus. Methods This prospective analytical study included 175 children aged one month to 14 years who fulfilled the inclusion criteria and were admitted to a tertiary care pediatric ICU between August 2022 and May 2024. We prospectively enrolled pediatric patients aged one month to 14 years who presented with convulsive SE (seizure duration of more than five minutes) and whose clinical scores were evaluated before sedative administration. Patients with post-cardiac arrest or trauma-related seizures were excluded. Initial GCS and FOUR scores were recorded at admission before midazolam administration. Seizure duration was documented and categorized as 5-30 minutes or more than 30 minutes. Outcomes were classified as recovery, neurological deficit, or death at discharge. Specifically, a "poor neurological outcome" was defined strictly as in-hospital mortality. A "good outcome" was defined as overall survival, which included both patients who completely recovered and those discharged with a neurological deficit. Multivariate logistic regression analysis was performed to identify independent predictors of poor outcome (defined as death). Results Of the 175 children, 152 (86.4%) recovered fully at discharge, eight (4.5%) developed neurological deficits, and 15 (8.5%) died. Mean GCS (10.70) and FOUR (12.20) scores were significantly lower in patients with poor outcomes compared to those who recovered (p < 0.01 for both). Seizure duration of more than 30 minutes was strongly associated with adverse outcomes (p < 0.001). On univariate analysis, a lower GCS score (p = 0.01, 95% confidence interval (CI): 0.40-0.71), a lower FOUR score (p = 0.04, 95% CI: 0.42-0.74), and prolonged seizure duration (p < 0.01, 95% CI: 9.12-77.42) independently predicted poor neurological outcome. On multivariate analysis, only seizure duration of more than 30 minutes independently predicted poor neurological outcome (adjusted odds ratio (AOR) = 12.14, 95% CI: 3.14-43.26; p < 0.001), while GCS and FOUR scores lost significance. Conclusions Lower GCS and FOUR scores, along with prolonged seizure duration, were significantly associated with poor neurological outcomes on univariate analysis. However, after multivariate adjustment, only seizure duration remained an independent predictor. Based on our findings, these simple bedside tools can help with early prognostication and guide management.