Socioeconomic differences in use of antiseizure medication in pregnancies with maternal epilepsy: A population-based study from Nordic universal health care systems

Abstract

Objective: Research points to disparities in disease burden and access to medical care in epilepsy. We studied the association between socioeconomic status (SES) and antiseizure medication (ASM) use in pregnancies with maternal epilepsy. Methods: We conducted a cross-sectional study consisting of 21 130 pregnancies with maternal epilepsy identified from Nordic registers during 2006–2017. SES indicators included cohabitation status, migrant background, educational attainment, and household income. Main outcomes were the proportion and patterns of ASM use from 90 days before pregnancy to birth. We applied multiple imputation to handle SES variables with 2%–4% missingness. We estimated adjusted risk ratios (aRRs) and 95% confidence intervals (CIs) using modified Poisson regression with the highest SES category as reference. Results: Mothers with the highest education and the highest income quintile used ASMs least frequently (56% and 53%, respectively). We observed increased risks of ASM discontinuation prior to or during the first trimester for low SES. The risk estimates varied depending on the SES indicator from aRR = 1.27 for low income (95% CI: 1.03–1.57) to aRR = 1.66 for low education (95% CI: 1.30–2.13). Migrant background was associated with ASM initiation after the first trimester (aRR 2.17; 95% CI 1.88–2.52). Low education was associated with the use of valproate during pregnancy in monotherapy (aRR 1.70; 95% CI 1.29–2.24) and in polytherapy (aRR 2.65; 95% CI 1.66–4.21). Low education was also associated with a 37% to 39% increased risk of switching from one ASM to another depending on the ASM used. For the other SES indicators, aRRs of switching varied from 1.16 (foreign origin; 95% CI 1.08–1.26) to 1.26 (not married or cohabiting; 95% CI 1.17–1.36). Significance: Low SES was associated with riskier patterns of ASM use: discontinuation, late initiation, and switching during pregnancy. These findings may reflect unplanned pregnancies, disparities in access to preconception counseling, and suboptimal care.