Discontinuation of treatment for alcohol use disorder during pregnancy and postpartum in the United States

Abstract

Background

The degree of alcohol use disorder (AUD) treatment utilization during the perinatal period is unknown. We report the prevalence of preconception receipt of medications for AUD (MAUD) and psychosocial interventions (PSY), discontinuation during pregnancy, and postpartum resumption in a multi-state sample, comparing pregnant and nonpregnant people with AUD.

Methods

Using MarketScan combined commercial and Medicaid claims (2016–2019), we identified individuals with AUD who had continuous insurance coverage throughout pregnancy, classifying those with a live birth as pregnant, and compared their MAUD and PSY patterns to nonpregnant peers matched by age, insurance type, and calendar time. All individuals had ≥1 claim for: (a) AUD diagnosis and (b) MAUD or PSY in the year preceding the study. Outcomes—filled MAUD prescriptions (naltrexone, acamprosate, and disulfiram) and receipt of PSY—were identified via claims. We computed rates of MAUD and PSY receipt, stratifying by five observation windows for pregnant individuals (12-week preconception; first, second, and third trimesters; 12 weeks postpartum) and nonpregnant peers (by corresponding windows). We assessed time to treatment discontinuation using multivariable Cox regression, adjusting for sociodemographics and comorbidities.

Results

Our sample consisted of 2080 pregnant persons with AUD and 7564 matched nonpregnant AUD peers. During pregnancy, MAUD receipt declined from 12.1% (preconception) to 0.3% (third trimester) among pregnant people and from 13.5% to 8.1% in nonpregnant peers during the equivalent time period (p < 0.001). Postpartum resumption of MAUD was uncommon in the pregnant cohort (pregnant = 1.9%; nonpregnant = 7.8%, p < 0.001). PSY declined for both the pregnant and nonpregnant cohorts yet remained modestly higher in the nonpregnant cohort (postpartum 10.3% vs. 13.8%, p < 0.001). In adjusted analyses, pregnant people were more likely to discontinue MAUD than nonpregnant peers (HR = 2.11 [1.71–2.60]) yet not more likely to discontinue PSY (HR = 1.01 [0.87–1.17]).

Conclusions

Among pregnant people with preconception AUD receiving treatment, MAUD utilization is low and discontinuation is widespread, persisting postpartum.