Substance use in pregnancy remains common in the US with approximately 1 in 5 pregnant individuals having past-month use, 15% of whom meet DSM diagnostic criteria for substance use disorder (SUD). There has been an associated increase in the overdose mortality among pregnant and postpartum individuals, an increase of 81% from 2017-2020. The highest rates of overdose mortality are found in the late postpartum period. Research has repeatedly demonstrated that treatment of SUD in pregnancy improves perinatal outcomes including higher antenatal maternal retention in treatment, higher attendance at routine prenatal care visits, and fewer children removed from maternal custody and placed in the foster care system. However, there is a paucity of data regarding optimization of continuation of these positive effects in the postpartum period. The objective of this study was to determine if there are predictive factors associated with maternal retention in treatment and child custody retention that can be identified at an individual patient level.

An IRB approved retrospective cohort study was performed involving Substance use Treatment And Recovery Prenatal (STAR) Clinic patients from December 2020 – February 2024. Data sources included the electronic medical record, recovery data from the Oklahoma Department of Mental Health and Substance Abuse Services, service logs from STAR Clinic patients, and Department of Human Services and Child Protective Services. Descriptive statistics for the cohort including maternal demographic information, rates of treatment retention, and rates of child custody retention were performed. Logistic regression analyses were then performed to evaluate predictors of treatment retention at 1 month, 6 months, 12 months, and 24 months postpartum.

50 STAR patients were included in the study. 26 patients had postpartum recovery service data and/or medication assisted therapy (MAT) data available for analysis from the data sources described above. The median age was 30 years. Of these patients, 77% were Caucasian, 17% were black, and 6% identified as unknown or 2 or more races. Of the 26 patients with available postpartum recovery and/or MAT data, 77% remained in treatment at 1 month postpartum, 42% at 6 months postpartum, 35% at 12 months postpartum, and 7.7% at 24 months postpartum. Of the population studied, 9 patients have data available at the 24 month mark based upon when they delivered. Logistic regression analysis was then performed to determine predictive factors for retention in treatment postpartum. The results were not statistically significant, however notable trends included number of STAR Clinic prenatal visits and number of prenatal MAT visits.Number of STAR Clinic prenatal visits had an inverse relationship with retention in postpartum treatment with an OR of 0.39 (0.07, 0.87; p = 0.12). Patients with increased MAT visits antenatally had increased postpartum retention with an OR of 1.72 (1.17, 3.87; p = 0.06). Of our patient sample, 6 children were removed from maternal custody with 3 of the 6 children reunited at a later date. 1 child that was reunited was subsequently removed from custody again approximately 6 weeks later. Given the rarity of this outcome in this cohort of patients, a logistic regression analysis was deferred for predictive factors for child retention.

While the results of this study were not statistically significant, this descriptive analysis highlights areas for future research to optimize postpartum outcomes for the maternal-infant dyad. With continued data collection and a larger sample size, statistically significant predictors of maternal treatment retention and child custody retention may be elucidated.