PRP151: Receipt of preconception health services among women with and without opioid use disorder
Jennifer Bello-Kottenstette, MD, MS; Joanne Salas, MPH
Context: Use of opioids during pregnancy is associated with significant maternal morbidity and mortality. Addressing health and risky behaviors before pregnancy, in the preconception period, can improve both maternal and infant outcomes. However, the prevalence of preconception service delivery among women with opioid use disorder (OUD) is unknown. Determining the prevalence of preconception health service delivery among this high-risk group is needed to inform development of programs and policies to increase preconception health counseling in settings where women with OUD seek care. Objective: Identify the prevalence of preconception care services provided to women with OUD in the year before pregnancy and compare to those without OUD. Study Design: Cross-sectional analysis of secondary dataset using crude and adjusted logistic regression analysis. Dataset: Optum® de-identified Electronic Health Record (EHR) dataset containing data from 5 million adults, nationally distributed across the US from 2010 to 2018. Population studied: Women aged 18-55 with delivery between 2012 and 2018 (n=98,210). Outcome Measures: Main: Documentation of any preconception care service in the one year prior to pregnancy in women with OUD compared to those without OUD using ICD-9/10 codes. Results: Average age was 29; 67% of participants were Caucasian and 12% African American; 0.2% (n=209) of the sample had a diagnosis of OUD and 3.1% (n=3013) received any preconception services in the year prior to pregnancy. Women with OUD compared to those without had a higher prevalence of receiving any preconception services (6.7% versus 3.1%, p=0.002) and of having co-occurring alcohol use disorder (11.5% versus 0.3%, p<0.01) or nicotine dependence (55.5% versus 4.1%, p<0.01). Compared to women without OUD, women with OUD had an increased odds of receiving preconception services in unadjusted regression analyses (OR 2.28, [95% CI 1.32-3.92]) that was not significant after adjusting for demographics and co-occurring substance use (aOR 1.09 [95% CI 0.60-1.97]).