To establish an early diagnostic method, we assessed 213 hospitalized cases of febrile status epilepticus (FSE) between January 2004 and August 2014. We categorized FSE cases into an AESD group and a non-AESD group and compared their clinical courses, laboratory data and cranial computed tomography (CT) findings.
Of 213 hospitalized FSE cases, 19 (9%) were AESD. Univariate analysis showed that the AESD group took a significantly longer time to wake after FSE, had a higher degree of respiratory acidemia, and higher levels of serum AST, ALT, LD, hyperglycemia and hyperammonemia than the non-AESD group. We developed a scoring model that predicts AESD based on multivariate analysis. Using cut-off points of 4 and more with this scoring model, we could identify the AESD cases with 93% sensitivity and 91% specificity. These scores also had a positive correlation with prognosis.
Our scoring model enables early diagnosis of AESD. Patients with high scores should be observed carefully and early intervention should be considered.