Assessing the Impact of Prenatal Medication for opioid use disorder on Discharge Home with Parents among Infants with Neonatal Opioid Withdrawal Syndrome

Abstract

Objectives:

The number of women with opioid-related diagnoses in the United States has significantly increased in recent decades, resulting in concomitantly higher rates of infants born with neonatal opioid withdrawal syndrome (NOWS). Addressing prenatal opioid exposure is a priority for Alaska health systems. The objectives of this study were to: 1) identify maternal and neonatal factors associated with receipt of Medication for opioid use disorder (MOUD) and 2) determine the impact of prenatal MOUD on discharge to parents among infants with NOWS in three Alaska hospitals.

Methods:

A retrospective chart review using a standard abstraction form was conducted to collect data on neonatal and maternal characteristics, neonatal treatment, and infant discharge disposition for infants with NOWS born at the three hospitals between July 2016 and December 2019. A multivariable logistic regression model was used to determine factors associated with discharge to parents.

Results:

There were 10,719 births at the three hospitals during the study period, including 193 infants (1.8%) with NOWS. Among the 193 mothers, 91 (47.2%) received MOUD during pregnancy. Among infants with NOWS, 136 (70.5%) were discharged to parents, 51 (26.4%) were discharged to a relative or foster care. Infants were significantly (OR 3.9) more likely to be discharged to parents if the mother had received prenatal MOUD.

Conclusions:

MOUD among pregnant women with opioid use disorder furthers the goal of keeping families together and is a critical step towards reducing the impact of the ongoing opioid epidemic on Alaska families, communities, and the child welfare system.

1. Introduction

Opioid use during pregnancy has steadily increased since the early 2000s throughout the United States (US). From 1999 to 2014, the number of women with opioid use disorder (OUD) at delivery quadrupled¹. In 2019, nearly 7% of women in the US reported using prescription opioids during pregnancy, with 1 in 5 reporting opioid misuse². In addition to increasing maternal adverse events³, opioid use during pregnancy increases the risk of infants developing Neonatal Opioid Withdrawal Syndrome (NOWS), which can affect short, medium and long-term health and developmental outcomes.^4–7

Infants with NOWS experience irritability, difficulty feeding and reduced sleeping, and long-term effects of NOWS include increased risk for cognitive and behavioral problems^5,6. Treatment of NOWS includes non-pharmacologic supportive care including rooming-in, breastfeeding, and caring for infants in dim, quiet rooms with careful positioning and feeding^8–10. Pharmacologic treatment with an opioid agonist such as oral morphine or methadone has been used in up to 60–80% of infants with NOWS to control withdrawal symptoms.⁷ In Alaska, the incidence of NOWS among Medicaid-enrolled births increased more than fivefold from 2.7 births per 1,000 in 2004 to 18.4 in 2015¹¹. Between 2012–2013, an estimated 87% of infants born with NOWS required pharmacotherapy, generally consisting of an opioid taper, leading to an average length of stay of 23 days¹². On average, the cost of a hospital stay for an infant with NOWS is five times more than non-affected newborns¹².

Increased opioid use and overdose directly impact child welfare systems across the United States, with unprecedented increases in children entering foster care since 2012¹³. National studies have found a direct link between overdose deaths and drug-related hospitalizations and children entering foster care¹⁴. From 2004 to 2015, 38.7% of infants with NOWs from Alaska were removed from their mother’s care within 28 days of birth¹⁵. American Indian/Alaska Native (AI/AN) people have been disproportionately affected by the removal of children from their care and placement into foster care. resulting in disruptions of traditional parenting practices, loss of culture and language¹⁶. Alaska, like many states, has a mandatory reporting system for child maltreatment and neglect. In most instances, maternal substance use which threatens the newborn infant’s health or welfare falls under these statutes. The demonstrated links between parental substance use, overdose, and foster care suggest the importance of providing evidence-based substance use treatment to stabilize families and bolster infant health and wellbeing¹⁷. Given the known protective effects of medication for opioid use disorder (MOUD) on overdose and other outcomes, access is particularly urgent and critical in the context of increasing opioid use and overdose in Alaska, which have had a disproportionate impact on Alaska Native peoples and rural populations¹⁸.

MOUD (sometimes known as “medication-assisted treatment”) is a standard therapy for OUD and has been found to be effective at reducing cravings and preventing physiological symptoms of withdrawal¹⁹. Currently, methadone and buprenorphine are the two most commonly used medications for the treatment of opioid use disorder in the US, including with pregnant women.²⁰ MOUD for OUD during pregnancy has been associated with increased access to and engagement with prenatal care, increased maternal guardianship, lower rates of overdose, and reduced risk of HIV and viral hepatitis²⁰. Despite its proven benefits for both the mother and infant, there are many barriers to MOUD uptake and retention, including stigma and lack of access to care.^21,22 As such, the number of people with OUD far exceeds current treatment capacity and only a small proportion of people who enter treatment for OUD receive access to MOUD²³. While some state-level analyses have revealed increased access to MOUD during pregnancy gaps in access remain²⁴. Untreated maternal OUD is a risk factor for child placement in foster care, and pregnancy presents a unique opportunity to engage many women in substance use treatment, such as MOUD^3,9,13,25.

Despite the increasing prevalence of opioid use during pregnancy and infants born with NOWS, prenatal and newborn care vary greatly across the US²⁶. Differences in access to MOUD treatment can be explained by numerous critical gaps that exist with respect to the best practices for supporting women, particularly in rural and remote communities where there are limited specialized addiction services. Women’s health providers in Alaska hospitals have expanded their capacity to provide MOUDto prenatal women. In 2018, Alaska pediatricians, obstetricians, community members, and policy makers attended a NOWS breakout session at the AI/AN Clinical Translational Research Program’s (CTRP U54GM115371) Alaska Opioid Symposium. From this session, participants published a paper which identified prenatal OUD screening and treatment and family-based NOWS treatment as top priorities under Alaska NOWS policies and priorities²⁷. The current study was conducted in three Alaska hospitals that provide care to diverse populations across the state. The objectives of this study were to: 1) assess maternal and neonatal factors associated with receipt of MOUD, and 2) determine the impact of prenatal MOUD on discharge to parents among infants with NOWS.

2. Methods

Study setting

This study used clinical data from pregnant women who delivered an infant with NOWS at one of three Alaska hospitals: Alaska Native Medical Center (ANMC), Yukon Kuskokwim Delta Regional Hospital (YKDRH) and Fairbanks Memorial Hospital (ANMC). During the study period, July 2016-December 2019, all three hospitals used the Finnegan Neonatal Abstinence Scoring System (FINAS) to determine if infants required pharmacologic therapy for NOWS.

Alaska Native Medical Center (ANMC):

The ANMC is a 173-bed hospital, which provides inpatient, outpatient, tertiary, and specialty services annually to 175,000 Alaska Native people statewide. There are approximately 1600 births annually at ANMC among Alaska Native women from across the state.

Yukon-Kuskokwim Delta Regional Hospital (YKDRH):

The Yukon Kuskokwim Health Corporation (YKHC) is the largest rural Tribal health organization in southwest Alaska which encompasses approximately 75,000 square miles and serves a population of approximately 27,000 mostly Alaska Native people living in 50 villages. YKHC administers the YKDRH in Bethel. YKDRH has a team of family medicine doctors supported by obstetricians who oversee approximately 400–450 births per year. An additional 150–200 births of YK Delta’s infants occur at ANMC.

Fairbanks Memorial Hospital (FMH):

The FMH serves a large geographic population in central Alaska with a population of 100,000 with approximately 1,000 births per year. The FMH neonatal intensive care unit (NICU) supports births from the greater Fairbanks area.

Institutional Approvals

Alaska Native Tribal Health Consortium (ANTHC) and Southcentral Foundation (SCF), co-owners of ANMC, participated through the NIH Environmental Child Health Outcomes Program’s IDeA States Pediatric Clinical Trials Network (ECHO ISPCTN) in a study funded through the National Institutes of Health (NIH) Helping to End Addiction Longterm℠ (HEAL) Initiative²⁸ to evaluate the current experience in NOWS (Advancing Clinical Trials in Neonatal Opioid Withdrawal Syndrome (ACT NOW) Data Collection Study) through chart review of infants born 2016–2017^26,28. We obtained approval from the network, the Alaska Area IRB, and the three Tribal health organizations, ANTHC, SCF, and YKHC and from FMH, for a local NOWS Data Collection study to extend chart reviews for this local Alaska study.

Study Procedures

We conducted a retrospective chart review of the hospitals’ electronic medical record to extract data on neonatal and maternal characteristics, neonatal treatment, and discharge disposition for opioid-exposed term (≥ 36 weeks gestation) infants born between July 2016 and December 2019 at the three Alaska hospitals. The inclusion criteria were a) NOWS scoring obtained within the first 12 hours of life, b) Maternal history of prenatal opioid use in the medical record, or c). Maternal and/or infant toxicology screen positive for opioids. Infants were excluded if they were <36 weeks gestation, had major birth defects, respiratory support after 72 hours , hypoxic ischemic encephalopathy, neonatal seizure disorder, major surgical intervention, or infant exposure to opioids not related to NOWS treatment.

Data abstraction was conducted using a standard abstraction form developed through the ACT NOW study.

Measures

Dependent variables:

MOUD: MOUD included methadone, buprenorphine, and buprenorphine/naloxone (Suboxone®) treatment anytime during pregnancy and was retrieved from the maternal electronic medical record.

Discharge home to biological parents:

Discharge home to biological parents was determined as Yes/No and was retrieved from the newborn electronic medical record.

Independent variables:

Maternal characteristics:

Demographic data and information on maternal health status were extracted from clinical records retrieved from review of the newborn electronic health record in all three hospitals. Collected demographic data included: age of mothers at delivery, race (Black, White, American Indian (AI)/Alaska Native (AN), Native Hawaiian/Pacific Islander, Asian, other) and the rural and remoteness of the mother’s home community using the Rural-urban commuting area (RUCA) code of 1 to 10 (RUCA ≥4 is defined as rural). Data were also gathered on the number of previous pregnancies (parity), prenatal substance use (i.e. opioids, stimulants) determined from either self-report or toxicology (i.e. prenatal urine drug screens), adequate prenatal care (defined as ≥ 3 visits and prenatal care starting prior to third-trimester)²⁹.

Perinatal Care Characteristics:

Data on the care received by pregnant women and infants with NOWS were extracted from the mother’s and newborn infant’s electronic medical records. Information was gathered on supportive (non-pharmacologic) and pharmacologic treatments provided to infants with NOWS.

Newborn Discharge Disposition:

Data on the presence of biological parents caring for the infant during the newborn hospitalization (presence of biological parents in the room or visiting and caring for the infant) were extracted by review of newborn nursing records as part of the chart review. Data on the length of newborn hospital stay and discharge disposition of infants was also captured as part of the review.

Data Analysis

The ANTHC REDCap Survey platform (a secure web application for building and managing online surveys and databases) was used to house data for this project³⁰. Using R statistical software, we analyzed the characteristics of mothers and opioid-exposed infants at ANMC, FMH, and YKDRH hospitals from July 1, 2016 to December 31, 2019.

We assessed the characteristics and outcomes of opioid-exposed newborns and their mothers including: the proportion of infants discharged to the biological parent versus to a relative or foster care services, MOUD at any point during pregnancy, a RUCA code ≥4 or <4, maternal race, , maternal parity, adequate prenatal care (defined as ≥ 3 visits and prenatal care starting prior to third- trimester), delivery hospital, length of infant stay at hospital, parental presence (presence of parents in room or visiting caring for the infant), proportions of non-pharmacologic and pharmacologic care for infants. These characteristics were compared between mothers who had received MOUD vs. those who had not and for cases where infants were discharged to biological parents versus discharged to relatives or foster care. Categorical variables were compared using chi-square tests, and continuous variables were compared using t tests. A multivariable logistic regression model of discharge status was built to determine factors associated with discharge home to biological parents. All characteristics with a p- value < 0.1 in bivariate analyses were tested for inclusion in the logistic regression model. A two-sided p- value of <0.05 was considered statistically significant.

3. Results

Between July 1, 2016 and December 31, 2019 there were a total of 10,719 infants born at ANMC (n=5,450), FMH (n=3,694), and YKDRH (n=1,575) (Table 1). Of these infants, 193 (n=1.8%) were identified as full-term pregnancies with NOWS born at ANMC (n=136, 2.5%), YKDRH (7 including 2 who were transferred to ANMC, 0.4%) and FMH (n=50, 1.4%). Infants who were born prematurely (< 36 weeks gestation) were excluded to reduce sample heterogeneity. In total, 40 infants were excluded due to prematurity (ANMC n=29; FMH n=9; YKDRH n=2). The annual incidence of opioid-exposed infants at ANMC was 2.5% with no variations (p=0.17, Mann Kendall Trend test)) between July 2016 and December 2019, while the incidence at FMH was 1.35% with no significant trend (p=0.73, Mann Kendall Trend test).

Table 1.

Births of infants with NOWS at Alaska Native Medical Center, Fairbanks Memorial Hospital and Yukon Kuskokwim Delta Regional Hospital between July 1, 2016 and December 31, 2019 (n=193)

	N (%) N=193
Maternal Characteristics
Parity, mean ± SD	3.2 ±1.8
Race
Alaska Native	164 (85)
Black	2 (1)
White	24 (12)
Unknown	7 (4)
RUCA: rural	88 (46)
Maternal use of any opioid during pregnancya	192 (99)
Maternal heroin use	66 (34)
Maternal stimulant use	56 (29)
Any MOUD during pregnancya	89 (47)
Buprenorphine treatment	36 (19)
Methadone treatment	23 (12)
Buprenorphine + Naloxone treatment a	30 (16)
Hospital
Alaska Native Medical Center	138 (72)
Fairbanks Memorial Hospital	50 (26)
Yukon Kuskokwim Delta Regional Hospital	5 (3)
Adequate Prenatal Careb	139 (71)
Neonatal Characteristics and Care
Parental Presence, yesc	149 (77)
Sex of infant, male	95 (49)
Infant pharmacologic treatment	46 (24)
Newborn Discharge Disposition
Length of Stay, median, (min – max)	5 (1–55)
Length of Stay ≥ 5 days	98 (51)
Discharged to Parent	136 (70)

Footnote: All data were obtained from clinical records retrieved from review of the newborn electronic health record for infant and mother in all three hospitals; RUCA: rural = 4 or higher on RUCA scale

^(a)Maternal substance use was determined via clinical records as stated above, with source being either self-report, physician or social worker note, or toxicology data during pregnancy or at birth in mother or infant. Maternal use of any opioid includes: heroin, hydrocodone, oxycodone, hydromorphone, methadone, buprenorphine, buprenorphine + naloxone, fentanyl, other prescription opioids (i.e. Codeine, orphine, naloxone, Norco (hydrocodone), Norbuprenorphene, Norhydrocone, Noroxycodone, Oxymorphone,Percocet, Suboxone,T3, Tramadol, Vicodin) and opioid-like (Kratom). One newborn had NOWS scoring but the mother did not have documented opioid use.

^(b)Adequate prenatal care was defined as ≥ 3 visits and prenatal care starting prior to third-trimester.

^(c)Presence of parents in room following birth or visiting and caring for the infant.

SD= standard deviation RUCA= rural urban commuting area; Parity= number of prior pregnancies;

MOUD= medication for opioid use disorder; Hep C= hepatitis C.

Medication for opioid use disorder (MOUD)

Among 193 mothers, 91 (47%) received MOUD during pregnancy (Table 1). In univariate analysis, MOUD during pregnancy was more common in women identified as receiving adequate prenatal care (p<0.001), and less common in women with prenatal stimulant use (p=0.001). MOUD was associated with increased parental presence during newborn hospitalization (p=0.001). MOUD was more common in mothers of infants who were discharged home to parents (p<0.001) (Table 2).

Table 2.

Maternal and neonatal factors associated with receipt of Medication for Addiction Treatment among mothers who gave birth at Alaska Native Medical Center, Fairbanks Memorial Hospital and Yukon Kuskokwim Delta Regional Hospital between July 1, 2016 and December 31, 2019 (n=191)

	MOUD N=91	No MOUD N=100	Total N=191
	N (%)	N (%)	N (%)	p-value
Maternal Characteristics
Race: Alaska Native	77 (86)	85 (87)	162 (86)	0.98
RUCA: rural	35 (39)	52 (52)	87 (46)	0.08
Maternal heroin use during pregnancya	29 (32)	37 (37)	66 (35)	0.55
Maternal methamphetamine use during pregnancya	14 (15)	32 (32)	46 (24)	0.01
Maternal stimulant use during pregnancya	16 (18)	40 (40)	56 (29)	0.001
Hospital
Alaska Native Medical Center	62 (68)	76 (76)	138 (72)	0.41
Fairbanks Memorial Hospital	27 (30)	22 (22)	49 (26)
Yukon Kuskokwim Delta Regional Hospital	2 (2)	2 (2)	4 (2)
Adequate prenatal careb	81 (89)	56 (56)	137 (72)	<0.001
Neonatal Characteristics and Care
Parental Presence, yes	80 (88)	67 (67)	147 (77)	0.001
Infant pharmacologic treatment	25 (28)	21 (21)	46 (24)	0.38
Newborn Discharge Disposition
Length of Stay ≥ 5 days	54 (60)	44 (44)	98 (52)	0.04
Discharged to Parentc	78 (87)	58 (58)	136 (72)	<0.001

Footnote: Data on MOUD were missing for 2 participants, thus analyses of MOUD are conducted on 191 participants. All data were obtained from clinical records retrieved from review of the newborn electronic health record for infant and mother in all three hospitals; RUCA: rural = 4 or higher on RUCA scale

^(a)Maternal substance use was determined via clinical records as stated above, with source being either self-report, physician or social worker note, or toxicology data during pregnancy or at birth in mother or infant.

^(b)Adequate prenatal care was defined as ≥ 3 visits and prenatal care starting prior to third trimester.

^(c)Other included discharge to treatment program with mother (n=2), adoptive parent (n=1) and discharge to case worker (n=1). Discharge status was missing for 1 participant.

MOUD= medication for opioid use disorder; Hx= history; Tx= treatment; ANMC= Alaska Native Medical Center; YKDRH= Yukon Kuskokwim Delta Regional Hospital; FMH= Fairbanks Memorial Hospital; Tox= toxicology; RUCA= rural urban commuting area; Parity= number of prior pregnancies

MOUD was also associated with increased length of stay for the infant (p=0.04). (Table 2). Median hospital length of stay of opioid-exposed newborns was 5 days (range 1–55 days). Length of stay ≥5 days was more common at FMH than ANMC and YKDRH. Length of stay was longer (mean:20.5 days; SD 10.2 days) in infants who received pharmacologic therapy for NOWs in comparison to infants who did not receive pharmacologic therapy (p= <0.001; data not shown in tables).

Discharge to parent

Among the 193 infants with NOWs, 136 (70%) were discharged home to parent, while 25 (13%) were discharged to a relative, 26 (13%) were discharged to foster care, 3 were transferred to another hospital for possible pharmacological treatment, 1 was transferred to a rehabilitation unit with the biological mother, 1 was discharged to adoptive parents, and 1 was unknown. The proportion of infants discharged to parent did not change over time and was not associated with rurality, hospital, or race. Discharge home to biological parent was positively associated with maternal MOUD (p<0.001, adequate prenatal care (p<0.001), parental presence (p<0.001), and previous pregnancies (parity) (p=0.005) while discharge home to parent was negatively associated with maternal heroin (p<0.001) and stimulant use (p<0.001) during pregnancy (Table 3). Infants who were discharged to parent had a shorter length of hospital stay (p=0.004). In a multiple logistic regression analysis, maternal MOUD (OR. 3.9; 95% CI. 1.5–10.9; p=0.007), adequate prenatal care (OR. 3.7; 95% CI 1.5–9.2; p=0.005) and parental presence (OR. 5.0; 95% CI 1.9–13.8, p=0.002) were associated with higher likelihood of discharge home to parent, while maternal heroin use during pregnancy was associated with lower likelihood of discharge home to parent (OR 0.11; 95% CI 0.04–0.28; p<0.0001) (Table 4).

Table 3.

Maternal and neonatal factors associated with infants with NOWS discharged home to biological parent(s), Alaska Native Medical Center, Fairbanks Memorial Hospital and Yukon Kuskokwim Delta Regional Hospital between July 1, 2016 and December 31, 2019 (n=190).

	Infant Discharged to biological Parent(s)	Infant Discharged to Relative or Foster Care or Other
	N (%) N=136	N (%) N=54	Unadjusted ORs (95%CI)	p-value
Maternal Characteristics
MOUD, yes	78 (57)	12 (22)	4.9 (2.46–10.54)	<0.001
Race Alaska Native	115 (86)	49 (89)	1.4 (0.56–4.04)	0.64
RUCA: rrural	65 (48)	23 (41)	1.2 (0.65–2.26)	0.49
Maternal Heroin use during pregnancya	28 (21)	39 (70)	0.11 (0.05–0.22)	<0.001
Maternal Stimulant use during pregnancya	20 (15)	38 (70)	0.08 (0.04–0.17)	<0.001
Buprenorphine treatment during pregnancy	32 (24)	4 (7)	4.0 (1.4–14.0)	0.02
Methadone treatment during pregnancy	14 (10)	9 (16)	0.60 (0.25–1.5)	0.33
Buprenorphine + Naloxone treatment during pregnancy	25 (18)	5 (9)	2.3 (0.89–7.1)	0.13
Hospital
Alaska Native Medical Center	93 (67)	45 (33)		0.22
Fairbanks Memorial Hospital	39 (80)	10 (20)	1.9 (0.89–4.30)
Yukon Kuskokwim Delta Regional Hospital	4 (80)	1 (20)	0.65 (0.14–3.39)
Adequate prenatal careb	117 (86)	21 (38)	10.3 (5.0–21.7)	<0.001
Neonatal Characteristics
Parental Presence, yesc	118 (87)	30 (54)	5.7 (2.8–11.9)	<0.001
Infant Pharmacologic Treatment	27 (20)	19 (34)	0.47 (0.24–0.94)	0.059
Newborn Disposition
Length of Stay ≥ 5 days	59 (44)	38 (68)	4.2 (2.6–7.3)	0.004
	mean ± SD	mean ± SD
Parity	3.4 ± 1.9	2.7 ± 1.4	1.3 (1.1–1.7)	0.005
Length of stay	7 ± 7.1	11.6 ± 11.6	0.95 (0.91–0.98)	0.003

Footnote:

Note= Data on discharge status was missing for 1 participant, thus analyses of discharge are conducted on N=190 participants. RUCA: rural = 4 or higher on RUCA scale

All data were obtained from clinical records retrieved from review of the newborn electronic health record for infant and mother in all three hospitals

^(a)Maternal substance use was determined via clinical records as stated above, with source being either self-report, physician or social worker note, or toxicology data during pregnancy or at birth in mother or infant.

^(b)Adequate prenatal care was defined as ≥ 3 visits and prenatal care starting prior to third-trimester.

^(c)Presence of parents in room following birth or visiting caring for the infant.

OR= Odds ratio; MOUD= medication for opioid use disoeder; RUCA= rural urban commuting area; Parity= number of prior pregnancies;

Table 4.

Logistic regression of variables associated with discharge of infants with NOWS home to biological parents at Alaska Native Medical Center, Fairbanks Memorial Hospital and Yukon Kuskokwim Delta Regional Hospital between July 1, 2016 and December 31, 2019 (N=190).

Variable	OR	95% CI	p-value
Medication for Addiction Treatment during pregnancya	3.9	1.5–10.9	0.007
Maternal heroin use during pregnancyb	0.11	0.04–0.28	<0.001
Infant Length of Stay (≥ 5days)	0.93	0.87–0.98	0.009
Parental Presence	5.0	1.9–13.8	0.002
Adequate Prenatal Cared	3.7	1.5–9.2	0.005

Footnote: OR= Odds ratio; CI= confidence interval; MOUD= Medication for opioid use disorder

^(a)MOUD during pregnancy includes any of methadone, buprenorphine or buprenorphine+naloxone.

^(b)Maternal substance use was determined via clinical records as stated above, with source being either self-report, physician or social worker note, or toxicology data during pregnancy or at birth in mother or infant.

^(c)Presence of parents in room following birth or visiting and caring for the infant.

^(d)Adequate prenatal care was defined as ≥ 3 visits and prenatal care starting prior to third trimester.

4. Discussion

This retrospective chart review in three Alaska hospitals shows that receiving MOUD for opioid use disorder (OUD) during pregnancy increased the odds of infants with NOWS being discharged to a parent by four times. This finding reinforces past research that showed that maternal MOUD use greatly increases the likelihood of family unification³¹. However, despite evidence and clinical guidance supporting MOUD as a first-line treatment for OUD, more than half of the women in this study did not receive MOUD during pregnancy³². Pregnant women often experience barriers to accessing MOUD, including fear of apprehension of the child from child welfare services, racial or ethnic discrimination, limited prescribers in rural and remote communities, and a lack of accurate information about MOUD in the community, as well as among child welfare services and health care providers³³. The results of this study indicate that there is a need to expand MOUD accessibility to pregnant women in Alaska, with the potential to reduce referrals to foster care and increase family unification. Evidence supports improving access to MOUD for pregnant women through integration with prenatal and neonatal services and taking women-centered, family-based approaches to addressing complex health, social and economic factors that affect family unification³⁴. Expansion of MOUD throughout the state is supported by recently released state-wide guidance from the Alaska Department of Health and Social Services that emphasizes the importance of countering myths associated with MOUD and expanding access among high-risk groups including pregnant women and parents³⁵.

The provision of MOUD and other prenatal and social services during pregnancy represents an important opportunity to provide substance use treatment and support family unification where maternal and parental substance use is occurring. Research has found that prenatal maternal substance use alone is not predictive of foster care entry, however post-natal parental substance use has been associated with an increased likelihood of foster care placement and child maltreatment³⁶. While this study focused on maternal MOUD use in the prenatal period, additional research is needed that focuses on MOUD retention in the context of discharge to community and familial environments where parental prenatal substance use has occurred³¹.

In this study we found that MOUD use and discharge to biological parents were both lower among women who reported prenatal stimulant use (any use) and who had urine toxicology positive for methamphetamine. Prior studies have found that the concurrent use of cocaine and or methamphetamine among people with opioid use disorder can have negative impacts on access to and retention in MOUD^37,38. Stimulant use among pregnant women in the United States represents a growing public health concern, where prenatal stimulant use remains more common than opioid use³⁹. Prenatal stimulant use has been associated with psychiatric comorbidities, social isolation, histories of trauma and domestic violence, which have all been shown to negatively impact health outcomes among both children and mothers³⁹ . Our findings suggest that concurrent stimulant use was associated with lower rates of MOUD access among pregnant people. This presents an important factor to be explore in future studies of MOUD access in this population, and the unique service needs presented by people who concurrently use opioids and stimulants.

Pregnant women who obtained MOUD had more previous pregnancies than women who did not receive MOUD, and past pregnancies were also associated with a greater likelihood of infants being discharged with parents. These findings may reflect increased familiarity with child welfare services and the foster care system, and greater acceptability and knowledge of substance use treatment options including MOUD. These results also suggest that some women who have substance-exposed pregnancies have other children who may be known to child protection authorities which could affect their relationships with case workers⁴⁰.

This study also included analysis of factors associated with discharge home to parents. Parental presence during hospitalization was significantly associated with discharge home to parents. Parental presence in the hospital could indicate greater family stability and support which could be associated with uptake of MOUD. Past studies have shown the benefits of MOUD in combination with parenting and child welfare supports to parents retaining custody of children³¹. Expanded data is needed to measure the potential impact of non-pharmacologic supports and services in rural and remote communities that could reduce foster care entry³⁷. Future studies could also consider examining differences in access to various types of MOUD (e.g. methadone vs. buprenorphine) by patient characteristics, as well as factors such as geography (e.g. rural vs. urban residence).

In this study we also explored factors associated with prolonged length of newborn hospital stay (>5 days). We found infants with NOWS who received pharmacologic treatment, such as morphine, had a longer length of stay in hospital which could be attributed to greater NOWS severity. Past research has found that NOWS infants who require pharmacologic treatment have a longer length of stay than other NOWS infants⁴¹ Infants of women who received buprenorphine in the prenatal period have been found to have lower lengths of stay in hospital than women who received methadone⁴². Similarly, the “Eat, Sleep, Console” NOWS care model NOWS can reduce reliance on pharmacologic treatment and decrease length of stay in hospital⁴¹; the three hospitals implemented this model in 2020–2021.

This study has significant considerations for clinical practice and policy. Approximately 85% of participants in the present study were AI/AN women. Given the known beneficial impacts of MOUD on maternal and infant outcomes, and the disproportionate impact of separation from children on AI/AN women, implementation of the NOWS care model has been a priority at these hospitals. This study has significant considerations for clinical practice and policy, where this model could be adapted and implemented in other North American settings. Best practice indicates that it is desirable to keep families together to promote strong family units that can support children with NOWS and reduce the intergenerational effects of substance use^5,9. The results of this study support recommendations from the 2018 Alaska Opioid Symposium which emphasized that family-based approaches to NOWS management are essential to addressing systemic social factors that influence the health and well-being of infants with NOWS.

In response to this need, during 2020–2021 all three hospitals have trained staff and implemented a family-centered “Eating Sleeping and Consoling” (ESC) care model that has potential to improve both parental bonding to and hospital-based outcomes of opioid-exposed infants^41,43.

The findings of this analysis should be considered alongside several limitations. First, this data comes from three diverse hospitals with urban and rural AI/AN infants (ANMC, YKDRH) and urban/rural infants of diverse races (FMH) which impacts the generalizability of the results; however, the lessons about MOUD are translatable to a diverse audience. The sample is diverse and relatively small, limiting sub-analyses by race or rurality. Second, this study makes use of an observational design, whereby data were derived from clinical records. As such, there might be variables that have important associations with the outcomes of interest that were not collected, such as maternal age and parity.

5. Conclusions

Use of MOUD among pregnant women with opioid use disorder can contribute to furthering the goals of keeping families together and promoting intergenerational health and wellness. Increasing availability of and uptake of MOUD among pregnant women is a critical step towards reducing the impact of the ongoing opioid epidemic on children and child welfare services.