Table 1.

Summary of Studies on Opioid Use and Overdose in the Year Following Delivery (n=7)

Observational Studies
Author (Year)	Study Design & Time Period	Sample	Time Measured Post-delivery	Opioid Outcome	Relevant Independent Variables Assessed	Results
Ellis et al. (2019)	Retrospective chart review of methadone treatment clinic from 2010–2017	Pregnant women enrolled in methadone treatment during pregnancy (n=72)	30 days and 90 days	Opioid misuse, measured via toxicology	Age, race, injection drug use history, delivery type, methadone dose at delivery, days in treatment during pregnancy, opioid prescription at delivery discharge	Among the sample, 38.9% had opioid misuse within 30 days of delivery and 52.8% within 90 days. Opioid misuse at 30 days was significantly associated with days in treatment during pregnancy (p=.008), having a C-section (p=.045), and having an opioid prescription at discharge (p=.028). Age, race, IDU history, and methadone dose at delivery were not significantly associated with 30-day opioid misuse, and no variables were associated with misuse at 90 days. In adjusted models, days in treatment during pregnancy was negatively associated with opioid misuse at 30 days postpartum (b=-0.01, p=.011).
Kern-Goldberger et al. (2020)	Cross-sectional study of Nationwide Readmissions Database (NRD) from 1998–2014	Women aged 15–54 (n=1,063,845)	Not reported; used ICD-9-CM codes to identify postpartum	OUD-related hospitalization; Opioid overdose hospitalization	Age, race, health insurance type, residential ZIP-code income quartile, hospital teaching status, hospital urbanicity, hospital bed size, region of USA	From 1998–2014, postpartum OUD hospitalizations increased from 0.8% to 2.1% and postpartum overdose-specific admissions increased from 18.8 per 100,000 admissions to 65.2. In adjusted models, risk of postpartum OUD-related admissions was significantly higher among women aged 30–34 vs. aged 25–29 (aRR=1.20, p<.001), those with Medicaid vs. private insurance (aRR=6.56, p<.001), and those at an urban vs. rural hospital (aRR=1.66, p<.001). Postpartum OUD admissions were significantly lower among women aged 15–17 (aRR=0.09, p<.001) and 18–24 (aRR=0.62, p<.001) vs. 25–29, those with residential ZIP-code incomes above the first quartile (p<.001), and those in all other race/ethnicity categories compared to whites (aRR=0.12–0.73, p<.001). Hospital bed size and teaching status were not associated with postpartum OUD admissions.
Nielsen et al. (2019)	Retrospective cohort study of a statewide-linked administrative database from 2011–2015	Women who had a delivery of a live born neonate between 2012–2014 (n=174,517 deliveries)	12 months	Fatal and nonfatal opioid overdose	Age, race, education, marital status, health insurance type, OUD diagnosis, homeless in year before delivery, incarceration in year before delivery, depression, anxiety, number of ED visits in year before delivery, opioid treatment enrollment, MOUD at delivery, Opioid prescriptions during pregnancy, nonfatal overdose in year before delivery, delivery type, infant NAS diagnosis, infant low birth weight, infant receipt of breastmilk before discharge, prenatal care	There were 189 women who experienced at least one overdose in the year following delivery (11 per 10,000 deliveries). Of the overdoses, 93% were nonfatal and 58% occurred 7–12 months post-delivery. OUD diagnosis was associated with a substantially higher overdose rate compared to those without OUD (349.3 vs. 5.9 per 10,000 deliveries). In adjusted analyses among women with OUD, women who had a nonfatal overdose in the year before delivery (aOR=2.48, 95% CI=1.19–5.17), 3 or more ED visits in the year before delivery (aOR=2.28, 95% CI=1.8–3.75), an anxiety diagnosis (aOR=1.92, 95% CI=1.16–3.17), and an infant with NAS diagnosis (aOR=2.03, 95% CI=1.26–3.27) were at increased odds of an overdose in year following delivery. Demographics, homelessness, incarceration, depression, other opioid and drug treatment variables, delivery type and prenatal care, and other infant conditions were not significantly associated with opioid overdose among women with OUD.
Schiff et al. (2018)	Retrospective cohort study of a statewide-linked administrative database from 2011–2015	Women with OUD who had a delivery of a live born neonate between 2012–2014 (n=4,154 deliveries)	12 months	Fatal and nonfatal opioid overdose	Receipt of MOUD (methadone or buprenorphine), 3-month time interval from 0–12 months postpartum	Overdose rates were at the highest from 10–12 months postpartum at 12.4 per 100,000 person-days (95% CI=9.1–16.4), compared to 9.7 (95% CI=6.9–13.3) the year before conception and 3.3 (95% CI= 1.6–6.1) in the third trimester. While higher, differences in overdose rates 7–12 months postpartum compared to pre-conception were not statistically significant, and only significantly differed from the third trimester. Compared to those who did not receive MOUD, those with MOUD had a lower overdose rate at each 3-month postpartum interval but was only statistically significant at 4–6 months postpartum (1.3, 95% CI=0.2–4.7 vs. 10.7, 95% CI=6.8–15.9).
Wen et al. (2019)	Retrospective study of Nationwide Readmissions Database (NRD) linked per year from 2010–2014	Women aged 15–54 who had a hospital delivery record between Jan. to Oct. of each year from 2010–2014 (n=15,701,149 deliveries)	60 days	OUD-related (primary diagnosis) hospitalization	Age, health insurance type, OUD diagnosis at delivery, delivery type, delivery postpartum hemorrhage, pregestational and gestational diabetes, hypertensive disorders during pregnancy, chronic hypertension, residential ZIP-code income quartile, hospital teaching status and urbanicity, hospital bed size	There were 1,039 postpartum readmissions for OUD. Those with an OUD diagnosis at delivery had over a 100-fold increased risk of a postpartum OUD readmission than those without an OUD diagnosis (RR=109.61, 95% CI=96.2–124.9). In adjusted models where OUD diagnosis was controlled, OUD-related readmissions were positively associated with maternal ages 20–24 vs. 25–29 (aRR=1.16, 95% CI=1.00–1.36), having Medicaid (aRR=7.57, 95% CI=5.06–11.31) or Medicare (aRR=4.41, 95% CI=3.64–5.34) vs. private insurance, and having a C-section (aRR=1.28, 95% CI=1.12–1.45). Women aged 15–19 vs. 25–29 (aRR=0.66, 95% CI=0.50–0.87) and who experienced gestational diabetes (aRR=0.29, 95% CI=0.17–0.48) or hypertensive disorders during pregnancy (aRR=0.63, 95% CI=.48–0.83) were at lower risk of OUD readmissions. OUD readmissions were not significantly associated with postpartum hemorrhage, pregestational diabetes, chronic hypertension, residential ZIP-code income quartiles, hospital teaching status, or hospital bed size.
Intervention Evaluations
Author (Year)	Study Design & Time Period	Sample	Time Measured Post-delivery	Opioid Outcome	Intervention Description	Results
Cochran et al. (2018)	One-group repeated measures from 2015–2016	Pregnant women with OUD diagnosis within two weeks of buprenorphine induction (n=21)	8 weeks	Self-reported abstinence from illicit opioid abuse	Patient Navigator (PN): ten prenatal and four postnatal sessions with a trained PN to address barriers to healthcare and drug treatment	Past 30-day abstinence from illicit opioid use increased from 68% at baseline (during pregnancy before 25 weeks gestation) to 96% at 8 weeks postpartum. After adjusting for treatment exposure, abstinence from illicit opioids increased over time (b=0.15, p<.001).
Guille et al. (2020)	Nonrandomized control trial from 2017–2018	Pregnant women seeking treatment for OUD in obstetrics office (n=98)	6–8 weeks	Opioid toxicology	Telemedicine: Standardized OUD treatment during pregnancy delivered via telemedicine; compared to patients where treatment was delivered with office-based in-person care	At 6–8 weeks postpartum, 9.8% and 20.8% of telemedicine and in-person patients, respectively, had a positive opioid toxicology (p=.24). After propensity score and covariate adjustment, patients with treatment delivered via telemedicine and in-person did not differ on postpartum opioid screens (p=.66).