Abstracts

Rationale: Limited work has evaluated how demographic, health, and socioeconomic characteristics at birth, such as birth weight and family income, relate to pediatric epilepsy likelihood and severity. Specifically, existing research on this topic is restricted to small or non-US study samples and to narrow sets of potential predictors, limiting its generalizability. This literature gap is unfortunate; a better understanding of pediatric epilepsy predictors could offer practical benefits to patients, such as better targeted and tailored treatments. Methods: This study conducts retrospective database analysis, using New York State (NYS) inpatient and emergency department (ED) data from Healthcare Cost and Utilization Project (HCUP) databases for 2005-2013. This study tracks almost all births in NYS and follows this newborn population forward in time, which allows for examination of the relationship between birth characteristics and subsequent childhood outcomes, such as epilepsy diagnosis (Table), epilepsy-specific inpatient and ED health resource utilization, and overall inpatient and ED utilization. Epilepsy is defined as the occurrence of at least one 345.xx ICD-9-CM code (Table). The study considers the effect of socioeconomic characteristics such as residential zip code, demographic characteristics (e.g., gender/race), and health characteristics (birth weight). Descriptive statistics as well as Poisson multiple linear regression methodologies are employed for calculating incidence ratio rates, which are jointly adjusted for year of birth, insurance type, zip code income quartile, gender, and race. Results: The study sample consists of a population of 1,959,966 newborns. The median follow-up duration was 4.4 years (range: .1-9 years). Newborns with birth weights < 1500 grams (n=28,574) are 4.53x more likely (95% CI: 3.91-5.25) to be assigned an inpatient or ED epilepsy diagnosis in childhood, relative to newborns weighing > 3000 grams (n = 1,434,109). Looking at specific epilepsy subtypes, relative to newborns weighing > 3000 grams, newborns with birth weights < 1500 grams have 3.86x (p = .001) higher likelihood of a generalized epilepsy diagnosis, 5.78x (p = .001) higher likelihood of a focal epilepsy diagnosis, and 3.94x (p = .001) higher likelihood of an intractable epilepsy diagnosis. A childhood epilepsy diagnosis appears to be 10% (p = .048) more likely among African Americans than whites, 46% (p = .001) less likely for Asians than whites, and 20% (p = .001) less likely for females than males. A childhood epilepsy diagnosis appears to be 64% (p = .001) more likely among newborns originally on Medicaid relative to private insurance, and 18% (p = .001) more likely among babies in zip codes in the bottom relative to the top income quartile. Conclusions: Certain socioeconomic, demographic, and health characteristics from the time of birth are associated with higher likelihood of pediatric epilepsy diagnosis. Newborns with lower birth weight or lower socioeconomic status had an increased risk of developing epilepsy in childhood, as did certain demographic groups including males and African Americans. Funding: UCB Pharma-sponsored