Abstracts

Rationale: Peripartum seizures remain a leading cause of maternal morbidity and mortality worldwide. Therefore, prompt management and recognition of high-risk individuals is essential to reduce maternal adverse outcomes. However, it remains unclear how to best identify and monitor women at highest risk. Methods: We used the Nationwide Readmissions Database (NRD) to identify index admissions for delivery in 2013 and readmissions with a primary diagnosis of seizure. ICD-9-CM codes were used to identify index admissions for delivery as well as readmissions for seizures, defined by validated codes for epilepsy, convulsions and eclampsia in the primary diagnostic position. Logistic regression models were created to examine the association between pregnancy-related complications during the index admission and 30-day readmission for seizures. Pregnancy-related complications included gestational diabetes, gestational hypertension, pre-eclampsia, eclampsia, peripartum hemorrhage, genitourinary infections and caesarean section. In a second model, logistic regression was used to calculate the association between demographic, medical, psychiatric and pregnancy-related variables and 30-day readmission for seizures. Results: There were 1,633,391 index admissions for delivery and 295 readmissions for seizure/epilepsy at 30 days, corresponding to a rate of 19.69 per 100,000 index admissions. Approximately 80% of seizure/epilepsy readmissions at 30 days post-partum represented readmissions for eclampsia. The presence of any pregnancy-related complication increased the odds of seizure readmission by 49% (OR 1.49 (1.16-1.92), p< 0.002). Logistic regression models for pregnancy related complications noted a significant effect of gestational HTN (OR 2.86 (2.00-4.09), p< 0.005) and caesarean section (OR 1.28 (1.02-1.62, p=0.04). In a full logistic regression model using medical and demographic variables, rising income quartiles relative to the lowest income quartile were associated with a decreasing odds of seizure readmissions (OR 0.71(0.52-0.96), p=0.03 for third income quartile, OR 0.50 (0.34-0.72), p=0.03 for fourth income quartile). Pregnancy-related variables associated with seizure readmission included gestational hypertension (OR 2.78 (1.94-3.98), p< 0.001), pre-eclampsia (OR 3.23 (2.31-4.51), p< 0.001), and eclampsia (OR 19.9 (9.38-42.2), p<0.001). Medical-related variables associated with seizure readmission included pre-existing epilepsy (OR 10.63 (6.43-17.56), p <0.001). Pre-existing drug use (OR 1.7 (0.97-2.99), p=0.07) and caesarean section (OR 1.23 (0.97-1.56), p=0.09) approached but did not reach statistical significance in a full model. Conclusions: Our study is among the first, to our knowledge, to use a nationally representative, longitudinal database to examine medical and pregnancy-related exposures and their relationship to readmissions for post-partum seizures. The majority of seizures occurring at 30 days post-partum represent eclampsia rather than convulsions or pre-existing epilepsy. A variety of pregnancy-related and medical conditions were associated with post-partum seizure readmission, including gestational hypertension, pre-eclampsia, pre-existing epilepsy and eclampsia, in order of increasing magnitude. Women with these risk factors should be referred to the appropriate providers in order to optimize care during pregnancy and the postpartum period and they should be monitored closely for the development of neurologic complications. Funding: No funding