Abstracts

Rationale: Pregnancy planning is vital in women with epilepsy (WWE) due to the risks of teratogenicity and harmful cognitive effects to the fetus of some antiepileptic medications (AEDs), such as valproic acid.  However, WWE are at risk of unintended pregnancy, as some AEDs increase the metabolism of hormonal contraceptives.  Prior studies using self-selected WWE who respond to surveys have found a high incidence of unintended pregnancy in WWE (65%; Herzog et al, Neurology 2017, 88:728-733).  The goals of this study are to use rigorously sampled population data to assess whether the proportion of unintended pregnancies are higher in WWE than in women without epilepsy (WWoE), and to compare prenatal care and breastfeeding in WWE and WWoE. Methods: The Pregnancy Risk Assessment Monitoring System (PRAMS) is an annual survey of randomly sampled post-partum women administered by the Centers for Disease Control (CDC).  Thirteen states query whether the respondent has active epilepsy, and these self-reports were used to categorize the women.  We used data from the PRAMS from 2009-2014 to assess the incidence of unintended pregnancies in WWE compared to WWoE. We adjusted the data for maternal age, race, and socioeconomic status, and calculated odds ratios for unintended pregnancy using logistic regression.  We also compared breastfeeding, prenatal care, and prenatal vitamin use in WWE to those in WWoE. Results: This study included 75,580 women, of whom 548 reported having a diagnosis of epilepsy in the 3 months prior to their pregnancies. WWE had a higher proportion of unintended pregnancies (60.26%) than did WWoE (51.23%), p < 0.001.  WWE were younger (p=0.001) and of lower socioeconomic status (SES; p < 0.001) compared to WWoE.  Among WWE, unintended pregnancies were more likely in those with younger age (odds ratio (OR) 1.21, 95% confidence interval (CI) 1.01-1.44), lower SES (OR 3.84, CI 2.31-6.38), and African American race (OR 2.31, CI 1.28-4.16); p < 0.05.  After adjusting for age, race, and SES, epilepsy was not a significant predictor of pregnancy intention (OR 1.08, CI 0.87-1.35); p=0.46.The proportion of women reporting breastfeeding was lower in WWE than WWoE after adjusting for covariates (68% vs. 84% reported any breastfeeding, p < 0.001), with OR 0.44 (CI 0.35-0.56) for breastfeeding in WWE.After adjusting for age, race, and SES in multivariable logistic regression, epilepsy was not a significant predictor of inadequate prenatal care in WWE, defined as receiving less than 80% of expected prenatal visits.  Conversely, WWE were more likely to have “adequate plus” prenatal care, defined as receiving more than 110% of expected prenatal visits (OR 1.48, CI 1.16-1.90); p=0.002.WWE were more likely to report taking a daily prenatal vitamin during their pregnancies, with OR 1.4 (CI 1.13-1.78) for daily prenatal vitamin use (p=0.003) after adjusting for covariates. Conclusions: Although planning for pregnancy is of utmost importance for WWE, the proportion of unintended pregnancies is higher in WWE than in WWoE.  Maternal age, race, and SES differences in WWE contribute to this discrepancy.   Daily prenatal vitamin use and “adequate plus” prenatal care are higher in WWE.  The proportion of women reporting breastfeeding is lower in WWE, despite studies indicating the safety of breastfeeding in WWE.  Continuing education by caregivers about pregnancy planning and prevention and about breastfeeding in WWE is needed. Funding: Not applicable.