Abstracts

Rationale: The goal of this study was to identify modifiable risk factors for epilepsy-related readmissions in order to improve clinical outcomes while also reducing cost of care. Methods: The Montefiore network serves a large, multicultural urban population and is anchored by three acute care hospitals each with an emergency department (ED) and one stand-alone ED. Our study sought to identify risk factors for returning to an ED or hospital setting within a 90-day period after an initial ‘index’ discharge from an inpatient ward or an ED.We constructed an initial cohort of patients discharged from the hospital or ED in 2012 with an epilepsy or seizure-related diagnosis.  We constructed a second cohort from the initial cohort that was re-admitted to ED or hospital for any reason. We constructed a third cohort from the initial cohort that was readmitted for a seizure or epilepsy related diagnosis. We examined three 90-day outcomes for this study: epilepsy-related ED visits, epilepsy-related hospital readmission, and a combined measure of ED visits and hospital readmissions.  For hypothesized predictor variables, we examined demographic data for each patient including age, race, sex, ethnicity, preferred language, zip code, date of initial admission to hospital or ED visit, length of initial stay if admitted, date of return to either ED or hospital, whether the re-visit to ED or hospital was related to a seizure DRG or ICD9 code, and insurance type. Results: We identified 3586 patients who were discharged from a hospital inpatient unit (73%) or from ED (27%).  24% of ED discharges and 29% of hospital discharges returned to ED or inpatient care within 90 days for any reason. Among these patients, 60% and 77% of ED discharges and hospital discharges, respectively, were readmitted to hospital.  16% of ED discharges and 20% of hospital discharges returned to ED or inpatient care within 90 days for a seizure or epilepsy diagnosis. Nearly half of these patients did so within the first 3 weeks after discharge. Among patients returning for seizures, 69% of ED discharges and 89% hospital discharges were readmitted to hospital. A number of co-morbidities increased the likelihood of readmission: history of epilepsy (prevalence 63%) odds ratio (OR) of readmission 2.0 (95% confidence interval 1.6 - 2.5), history of psychiatric illness (prevalence 34%) OR 1.6 (1.4 - 2.0). Other less prevalent, but significant risk factors favoring readmission within 90 days were history of cerebrovascular disease, metastatic tumors, congestive heart failure. Conclusions: Our data shows that a high proportion of patients discharged from the either the ED or the inpatient setting (24% and 29%, respectively) are re-evaluated in the ED or re-admitted to hospital within 90 days. Approximately two-thirds of these returning patients are re-evaluated in ED or re-admitted to hospital for reasons related mainly to epilepsy (16% and 20%, respectively), while the rest are re-evaluated or re-admitted for other medical or psychiatric comorbidities. The comorbidities present at initial discharge that most predict subsequent readmission are foremost a prior history of seizures or epilepsy, or a history of psychiatric illness, and to a lesser extent any history of cerebrovascular disease, metastatic tumors, congestive heart failure.  These data indicate that patients with history of epilepsy are at high risk for re-evaluation in ED or readmission to hospital, and that this risk is further elevated by specific medical and psychiatric comorbidities. Future studies may further examine factors leading to readmission in patients at greatest risk with the goal of designing interventions that better maintain patient health are reduce need for ED and inpatient hospitalization. Funding: None