To determine 30-day injury readmissions risk amongst persons with epilepsy compared to individuals without epilepsy using a nationally representative US database.

Injury risk is higher in persons with epilepsy than in persons without epilepsy. To our knowledge, this is the first study to use the Nationwide Readmissions Database (NRD) to assess risk of injury readmissions after an epilepsy index hospitalization. 

Individuals of all ages with epilepsy as the primary diagnosis were identified using validated ICD-9-CM codes in the 2014 NRD. 1:4 matching on propensity scoring (by age, sex, Elixhauser comorbidity index) was performed to compare persons with epilepsy versus controls. Primary outcome was 30-day readmission for an injury defined by ICD-9-CM diagnosis codes following discharge from the index hospitalization. Logistic regression was conducted to examine factors associated with readmissions.

There were 66,339 unique persons with epilepsy (mean age 42.2 years, female 48.9%) and 331,695 without epilepsy (mean age 43.9 years, female 48.0%). 177 (0.25%) persons with epilepsy and 639 (0.23%) persons without epilepsy were readmitted for an injury. Statistically significant differences (p-value <0.05) in baseline demographics in persons with vs. without epilepsy included: Medicaid insurance (30.7 vs. 27.1%), discharge to a facility (16.4 vs. 12.4%), and hospital bed size (63.1% vs. 58.0%). In the multivariate analysis the following factors were associated with injury readmissions in persons with epilepsy: 1) Medicare (OR 2.8 95%CI 1.8-4.4) and Medicaid (OR 2.0 95%CI 1.2-3.6) vs. other insurance status; 2) discharge to a facility other than home (OR 1.5 95%CI 1.0-2.1).

Higher proportions of 30-day injury readmissions were observed in persons with epilepsy vs. without epilepsy. Our findings suggest that it may be important to further optimize discharge planning in persons with epilepsy on Medicare, Medicaid or being discharged to a facility, as their risk of injury appears to be greater.