Abstracts

Rationale: Status epilepticus (SE) may lead to neurologic sequelae and medical complications, including deaths, and affects quality of life and healthcare costs. The limited literature on hospital readmissions suggests that patients hospitalized for seizures present with a high rate of readmission. We aimed to estimate the hospital readmission rate for SE after a prior admission for SE among epilepsy patients in the USA, and study potential risk factors for readmission. Methods: Retrospective descriptive study using the US Healthcare Cost and Utilization Project’s 2016 Nationwide Readmissions Database (HCUP-AHRQ). Outcomes. Primary outcome: Describe the number of patients who had more than one admission for SE in 2016 using a retrospective cross-sectional analysis. Secondary outcomes: Determine the cumulative probability of undergoing a second admission for SE within the following 6 months after the index event admission, and study potential risk factors related to this readmission. Inclusion/exclusion criteria. Inclusion criteria: (1) patients with epilepsy of all ages with (2) nonelective hospital admission in 2016 due to SE as first diagnosis code. For the secondary outcome, we added (3) discharge month of index event between January and June included (to assure 6 months of follow up within the HCUP database). Exclusion criteria: (1) admissions without basic demographic/clinical information and (2) discharge against medical advice. For the secondary outcome, we only considered patients who survived the index admission and the first readmission if a patient had >1. Transfer to another hospital was not considered readmission. To define SE, we used the US ICD-10 codes within the epilepsy category with mention of SE.Predictors analyzed. Related to patients’ demographic and clinical characteristics (age, gender, type of epilepsy [refractory or non-refractory], type of SE in the index admission [non-refractory, refractory or super-refractory SE, based on mechanical ventilation], primary payer, median household income, length of hospital stay) and hospital characteristics (ownership/control, bed size, teaching status, and location). Statistical analysis. Descriptive statistics, and time-to-event analysis (time censored at 6 months of follow up; Cox proportional hazards regression model). We calculated times from the first day of admission of the index event. Results: We included ~31,500 epilepsy patients admitted due to SE in 2016 (study population detailed in Table 1). Among these patients, 91.7% had 1 admission for SE (2.8% died), and the others were readmitted: 6.5% had 2 admissions (1.9% died), 1.3% had 3 admissions (2.9% died), and 0.5% had ≥ 4 admissions (unable to report <10 deaths) (Figure 1a). The cumulative probability of having had a readmission at 6 months from the index admission was approximately 11% -mostly within the first 3 months, 7.5% (Kaplan-Meier curve, Figure 1b). Older patients were less likely to have a readmission for SE than younger patients (HR 0.99, 95%CI 0.988-0.997, p adjusted=0.02). Patients with refractory epilepsy were more likely to have a readmission for SE than patients with non-refractory epilepsy (HR 1.44, 95%CI 1.18-1.75, p adjusted=0.007). The predictor refractory epilepsy remained statistically significant after stratifying for pediatric patients, adults, and elderly. Conclusions: Hospital readmission for SE among epilepsy patients is considerably high, which may affect healthcare costs and may highlight a need for intervention. Resources focused on patients with risk factors, such as refractory epilepsy, may help decrease readmissions for SE. Funding: ERF. MAG by FAME.