Engagement with recommended developmental follow-up and supports among infants with intrauterine opioid exposure

Abstract

Objective:

Describe the engagement of opioid-exposed infants with recommended developmental surveillance and supports in the first year of life.

Methods:

We conducted a single-site retrospective cohort study of opioid-exposed infants delivered between 2016–2021, linking birth hospitalization, Developmental Follow-Up (DFU) Clinic, and Early Intervention (EI) services records. Primary outcomes were attendance at DFU clinic and evaluation by EI. We used multivariable modified Poisson regression to examine how birthing parent-, infant- and clinic-level factors are associated with service engagement.

Results:

Of 256 opioid-exposed infants, 75% engaged in at least one developmental service. Referral and attendance rates at the DFU clinic were 69% and 33%, respectively. Ninety-three percent were referred to EI, 73% evaluated and 58% enrolled in services. EI evaluation was positively associated with prolonged infant hospitalization (aRR 1.01, CI 1.002–1.01), exposure to antidepressants (aRR 1.23, CI 1.02–1.49) and cocaine (aRR 1.28, CI 1.09–1.50). Probability of attendance at DFU was higher for infants born to parents receiving care at an integrated perinatal substance use clinic (aRR 2.13, CI 1.07–4.24) and exposed to antipsychotics (aRR 1.73, CI 1.12–2.67), while those remaining in parental custody had lower probability of engagement (aRR 0.62, CI 0.39–0.97).

Conclusions:

Three quarters of the eligible infants engaged in developmental surveillance services in the first year of life. Factors relating to disease severity, location of birthing parent care, birthing parent co-exposures, and parental custody were associated with engagement. Efforts to improve engagement in recommended follow-up should elicit the perspectives of caregivers to better understand the mechanisms that drive these differences.

INTRODUCTION

The rate of opioid use disorder (OUD) among pregnant people in the United States more than doubled from 2010 to 2017, with the rate rising from 3.5 to 8.2 per 1000 birthing parent opioid-related diagnoses.¹ The incidence rates of birthing parent OUD diagnoses and infant Neonatal Opioid Withdrawal Syndrome (NOWS) diagnoses, a condition that affects between 60% to 80% of newborns exposed to opioids prenatally, vary tremendously between states and disproportionately impact individuals living in rural and low income areas.^1,2 Additionally, demographic characteristics have also been shown to be associated with the incidence of OUD and NOWS. Non-Hispanic/Latinx white individuals and those dependent on public health insurance have a higher incidence of OUD and NOWS.^1–3

Children with a diagnosis of NOWS have an increased risk of developmental delays in early childhood,^4–6 lower cognition and school performance,^7–10 and behavioral issues.^9,11 Therefore, the American Academy of Pediatrics (AAP) recommends increased developmental screening of infants with in utero opioid exposure (IOE) during the first three years of life, including with formal developmental assessments.¹² Despite these recommendations, studies have identified lower adherence to well-child care pediatric visits in the first year of life among infants with IOE compared to unexposed individuals.^13,14 For this reason, some states provide additional screening and treatment for infants with IOE beyond typical well-child care. For example, in Massachusetts, a diagnosis of NOWS is a qualifying condition for one year of Part C Early Intervention (EI) services, a federal grant program administered a the state-level that can offer services such as physical and occupational therapies, family training, counseling and psychological services through home visiting.¹⁵ Yet, even with this automatic qualification, one single-site study in Massachusetts demonstrated that the majority of eligible infants were not successfully enrolled.¹⁶ Finally, although Developmental Follow-Up (DFU) clinics are increasingly offering developmental assessments for infants born exposed to substances, information about attendance at these clinics is less well described, with data from one midwestern clinic with a high volume of opioid-exposed infants reporting less than 25% of infants referred completing a standardized assessment.⁴

While little is known about factors associated with successful engagement among infants with IOE, research shows there are several birthing parent and infant characteristics associated with engagement in the general population. With respect to EI, birthing parent non-Hispanic Black race,^17,18 lower level of education and poverty in town of residence have been associated with lower referral rates.¹⁹ In comparison, prolonged hospitalization after birth and lower birth weight,²⁰ both of which are more frequent in infants with NOWS, were associated with higher engagement.^3,21,22 Similarly, the engagement in DFU clinics for infants that required hospitalization in the neonatal intensive care unit (NICU) was lower for infants who were born to non-Hispanic black birthing parents, primary non-English speaking and from neighborhoods with a lower Child Opportunity Index (COI).²³ It is necessary to better understand the post-discharge engagement of infants with IOE with recommended developmental services and to examine individual and structural factors associated with engagement in order to better care for families impacted by perinatal opioid exposure.

Our aims in this retrospective cohort study were to (1) describe the engagement of infants with IOE in the recommended EI program and DFU clinic visits; and (2) evaluate the extent to which birthing parent-, infant- and healthcare level factors influence the engagement in both these services.

METHODS

Study Design and Participants

We conducted a retrospective cohort study of birthing parent-infant dyads with perinatal opioid exposure with deliveries between April 2016 and October 2021 at a single tertiary academic medical center in our state. During the study period, a specialty clinic was opened in 2018 to provide prenatal, postnatal and pediatric care to families impacted by parental substance use. All deliveries resulting in a live birth where the infant was observed or received treatment for NOWS were included. Observations for NOWS were identified using the Partners Electronic Data Warehouse by identifying infants with evaluations using the modified Finnegan Neonatal Abstinence Syndrome Scoring Tool or the Eat, Sleep, Console assessment tools.^24,25 We excluded newborns who developed NOWS secondary to medical intervention in the setting of critical illness and infants who had opioid exposure only during labor and delivery.

The hospital-based Developmental Follow-Up clinic offered 3 visits in the first year of life: at 5 months, 8 months and 12 months of age.

Outcomes

Our primary outcomes were divided into Early Intervention and Developmental Follow-Up clinic engagement. First, we assessed the several steps of involvement in EI services: 1) referral to EI; 2) evaluation by an EI provider; and 3) enrollment with an individualized family service plan. In order to assess these outcomes, we linked electronic medical record (EMR) information with data from Partners Department of Public Health (DPH) EI database, combining information from both manual chart review and the DPH EI dataset. In the DPH dataset, dates of referral, evaluation, and individualized family services plan (IFSP) were used to determine completion of referral, evaluation and enrollment before 1 year, respectively.

Similarly, for each study participant we categorized DFU clinic engagement in the first year of life, as follows: 1) referral placed in patients’ EMR in the first year of life; 2) completion of at least one visit at DFU clinic and 3) completion of ≥2 visits.

Primary Exposure and Birthing Parent-Infant Characteristics

Our primary exposure of interest was the type of prenatal opioid exposure, classified as prescribed opioids, including medication for Opioid Use Disorder (MOUD), and non-prescribed or illicit opioids. We manually extracted birthing parent characteristics and co-exposures from prenatal and delivery EMR, including: birthing parent age (years, and advanced birthing parent age >35 years), race/ethnicity, marital status (partnered or single), public insurance coverage (yes/no) as a proxy for low socioeconomic status, household zip code was used to determine state-adjusted COI, a multidimensional index that relies on 3 domains (education, health and environment, social and economic) - to summarize the resources that promote healthy development (categorized into very high, high, moderate, low and very low)²⁶, gravidity and parity, use of other non-prescribed substances [self-report and birthing parent toxicology: cannabis (yes/no), cocaine (yes/no)], tobacco and alcohol consumption during pregnancy (self-report yes/no)]. From birth hospitalization records, we collected the following: birthweight (kilograms), gestational age (weeks and days, prematurity), sex (male/female), NOWS diagnosis, defined as present on the EMR problem list (yes/no), pharmacological treatment (yes/no), second pharmacological agent (yes/no), length of hospitalization (days), meconium and urine toxicology (binary for cannabis, cocaine, amphetamines, fentanyl, heroine), highest level of care required (categorized as: level 1 Nursery, level 2/ Special care nursery (SCN), level 3/ NICU), reporting to child welfare services (yes/no), discharge disposition (parental custody or binary guardian or foster placement). From EMR for the first year of life, we also obtained information on location of pediatric providers (categorized as: affiliated providers, outside providers and clinic specialized in the care of families affected by substance use disorders [SUD]).

Statistical analysis

Data were analyzed using Stata version 16.1 (StataCorp LLC) software. We first conducted summary statistics analysis of individual-level variables to examine differences between those engaged in EI or DFU and those who were not, using two-sided t-test for continuous variables and Fisher’s exact test for categorical variables, with a level of significance of p<0.05. Additionally, we described the proportion of infants completing each stage of involvement in EI services (referral, evaluation, enrollment) and DFU engagement (referral, completion of at least one visit, completion of ≥2 visits). For multivariable analysis, we used completion of evaluation for EI and completion of at least one visit for DFU engagement as our primary outcomes.

To examine independent factors associated with engagement in EI and DFU, we conducted separate multivariable modified Poisson regressions using forward and backward stepwise selection of variables. For each of these outcomes, candidate variables were derived from univariate modified Poisson regression, where those that reached a level of significance of p<0.1, were included in the stepwise selection phase. We controlled for a priori variables associated with EI and DFU engagement described in literature (advanced birthing parent age, birthing parent race and ethnicity, prematurity) and for varying trends in practices by including the birth year. We then proceeded with combined stepwise selection of variables into the model, using p<0.05 for inclusion, in subsequent steps. We adjusted for multiple comparisons using the Holm-Bonferroni method and reported adjusted confidence intervals for risk ratios. Missingness in categorical variables was controlled for by considering missingness in a specific variable the reference category and forcing a composite dummy variable for individuals that had any missing information to the Poisson regression models. For the continuous variable of length of hospitalization, we addressed missingness by calculating the mean length of hospitalization for subjects without missing information and imputed that value. We examined our models for collinearity, used goodness-of-fit tests and assessed receiver-operator curves.

We conducted multiple sensitivity analyses to evaluate the strength of our findings. First, we analyzed the influence of the same birthing parent being represented more than once for multiple pregnancies during the study period, by utilizing mixed-effects logistic regression models. Secondly, we recognize that infants who are sicker and have more comorbidities have a longer hospitalization after birth with specialized attention in a SCN or NICU. Therefore, we assessed differences in estimates adjusting for SCN/NICU admission, regardless of whether this step up in care was related to NOWS diagnosis.

We implemented a data quality check of 5% of the charts to review the accuracy of the extracted information. This study was reviewed and approved by the Partners Institutional Review Board.

RESULTS

We identified 256 birthing parent-infant dyads, predominantly non-Hispanic/Latinx white birthing parents(79%), who were publicly insured (87%) and had received any treatment with MOUD (81%) during pregnancy. Eight percent had only non-prescribed opioid use and 30.5% reported both prescribed and non-prescribed opioids. More than half the birthing parents used tobacco or nicotine products (69%), with a lower proportion endorsing other substances, including cocaine (29%) and cannabis (29%). On average, birthing parents had 1.3 previous live births (SD 1.3). Infants had a mean gestational age of 37.7 weeks (SD 2.6 weeks) and mean birthweight of 2.9kgs (SD 0.7kgs). The average length of hospitalization was 17.9 days (SD 16.6 days), during which 129 infants received pharmacologic treatment for NOWS with at least one medication (50%) and under a quarter (22%) requiring a second pharmacological agent. One-hundred and eighty-six infants (73%) required care in a higher acuity than the level 1 nursery for medical treatment of NOWS by hospital policy or treatment of other medical conditions. At the time of discharge home or transfer to another facility, 91.4% of dyads had been reported to child protective services per state guidance, with three quarters of the parents retaining custody of their child (76%). The majority of the infants lived in an area with very low (27%) and low (30%) COI, compared with almost 19% living in areas with a high or very high score. Most of the infants (57%) received their pediatric care outside of the network of the hospital where their NOWS management took place, while 18% were engaged in care in a clinic specialized in the care of families affected by SUD (Table 1).

Table 1 -.

Characteristics of the study population by engagement in the developmental services

	Total, n(%) n=256	Non-engaged, n(%) n=63	Engaged in DFU/EI, n(%) n=193	p-value
Maternal Characteristics
Age at delivery, mean (SD)	31.4 (4.9)	32.3 (5.0)	31.1 (4.8)	0.080
Advanced maternal age (AMA)	62 (24.2%)	21 (33.3%)	41 (21.2%)	0.063
Race and ethnicity				0.016
Non-Hispanic White	203 (79.3%)	46 (73.0%)	157 (81.3%)
Black or African-American	13 ( 5.1%)	6 ( 9.5%)	7 ( 3.6%)
Hispanic	8 ( 3.1%)	5 ( 7.9%)	3 ( 1.6%)
Other or multiracial	30 (11.7%)	5 ( 7.9%)	25 (13.0%)
Unknown/unavailable	2 ( 0.8%)	1 ( 1.6%)	1 ( 0.5%)
Partnered	56 (21.9%)	25 (39.7%)	31 (16.1%)	<0.001
Public payor insurance (Medicare/Medicaid) (n=253)	222 (87.7%)	51 (82.3%)	171 (89.5%)	0.18
Opioid exposure				<0.001
Only prescribed/MOUD	157 (61.3%)	52 (82.5%)	105 (54.4%)
Only non-prescribed opioids	21 ( 8.2%)	6 ( 9.5%)	15 ( 7.8%)
Prescribed/MOUD + non-prescribed opioids	78 (30.5%)	5 ( 7.9%)	73 (37.8%)
Any MOUD during pregnancy	206 (80.5%)	43 (68.3%)	163 (84.5%)	0.010
Psychiatric medications (n=255)
Antidepressants	105 (41.2%)	14 (22.2%)	91 (47.4%)	<0.001
Anti-anxiety medications	100 (39.2%)	16 (25.4%)	84 (43.8%)	0.011
Mood stabilizers	29 (11.4%)	3 ( 4.8%)	26 (13.5%)	0.067
Anti-psychotics	53 (20.8%)	4 ( 6.3%)	49 (25.5%)	<0.001
Benzodiazepines	103 (40.4%)	21 (33.3%)	82 (42.7%)	0.24
Stimulants	39 (15.3%)	12 (19.0%)	27 (14.1%)	0.42
Maternal co-exposures *
Cocaine	75 (29.3%)	10 (15.9%)	65 (33.7%)	0.007
Cannabis	75 (29.3%)	10 (15.9%)	65 (33.7%)	0.007
Amphetamines	48 (18.8%)	12 (19.0%)	36 (18.7%)	0.99
Alcohol	9 ( 3.5%)	1 ( 1.6%)	8 ( 4.2%)	0.46
Nicotine/Tobacco	177 (69.1%)	31 (49.2%)	146 (75.6%)	<0.001
Maternal care in perinatal SUD clinic	85 (33.3%)	14 (22.2%)	71 (37.0%)	0.032
Gravidity, mean (SD) (n=252)	3.7 (2.1)	3.8 (2.5)	3.7 (2.0)	0.70
Parity, mean (SD) (n=252)	1.3 (1.3)	1.5 (1.3)	1.2 (1.2)	0.20
Infant Characteristics
Gestational age (weeks), mean (SD)	37.7 (2.6)	38.5 (1.9)	37.4 (2.8)	0.004
Birthweight (kg), mean (SD)	2.9 (0.7)	3.1 (0.7)	2.8 (0.7)	<0.001
Female	117 (45.7%)	29 (46.0%)	88 (45.6%)	0.99
Treated pharmacologically	129 (50.4%)	18 (28.6%)	111 (57.5%)	<0.001
Treated with 2nd pharmacological agent	57 (22.4%)	8 (12.7%)	49 (25.5%)	0.037
Length of stay (days), mean (SD) (n=243)	17.9 (16.6)	9.8 (8.3)	20.6 (17.8)	<0.001
Highest acuity care unit
Level 1 Nursery	70 (27.3%)	33 (52.4%)	37 (19.2%)
Special Care Nursery (SCN)	119 (46.5%)	20 (31.7%)	99 (51.3%)	<0.001
Neonatal Intensive Care Unit (NICU)	67 (26.2%)	10 (15.9%)	57 (29.5%)
Parental custody at hospital discharge (n=253)	192 (75.9%)	61 (96.8%)	131 (68.9%)	<0.001
Child welfare reporting	234 (91.4%)	50 (79.4%)	184 (95.3%)	<0.001
Care and Neighbourhood Characteristics
Childhood Opportunity Index (COI) (n=254)				0.028
Very Low	68 (26.8%)	13 (21.3%)	55 (28.5%)
Low	77 (30.3%)	13 (21.3%)	64 (33.2%)
Moderate	61 (24.0%)	16 (26.2%)	45 (23.3%)
High/ Very High	48 (18.9%)	19 (31.1%)	29 (15.0%)
Pediatric Care location				0.097
Outside hospital network	146 (57.0%)	42 (66.7%)	104 (53.9%)
Same network	65 (25.4%)	15 (23.8%)	50 (25.9%)
Specialty clinic for families affected by SUD	45 (17.6%)	6 ( 9.5%)	39 (20.2%)
Engagement in developmental surveillance
No engagement	63 (24.6%)	63 (100.0%)
Only DFU	7 ( 2.7%)		7 ( 3.6%)
Only EI	107 (41.8%)		107 (55.4%)
Both DFU and EI	79 (30.9%)		79 (40.9%)

^*self-report, maternal and infant toxicology

Legend: CI: confidence interval; DFU: Developmental Follow-Up Clinic, EI: early intervention; MOUD: medication to treat opioid use disorder; NICU: neonatal intensive care unit; NOWS: neonatal opioid withdrawal; SCN: special care nursery; SD: standard deviation; SUD, substance use disorder

Engagement in EI and DFU

Two-hundred and thirty-nine infants (93%) were referred within the first year of life to EI, with 186 completing a formal evaluation (73%) and 149 being enrolled in services (58%) within the same time period. (Figure 1A). One hundred and ninety-seven infants were referred to the DFU clinic (77%) during the first year of life, while only one third completed at least one visit (34%) and 20% returned for subsequent visits in that time (Figure 1B). There were 7 infants (3%) that were only engaged in DFU, 107 infants (42%) were only engaged in EI and 79 infants (31%) were engaged in both services, in the first year of life (Table 1).

Figure 1.

Engagement in recommended follow-up services; A: Engagement in Early Intervention (EI) services; B: Engagement in Developmental Follow-Up (DFU) clinic (n=256)

Evaluation by EI services

In the multivariable modified Poisson regression model, exposure to antidepressants and cocaine was associated with a higher likelihood of engaging in EI, namely 23% increase for antidepressants and 28% increase for cocaine. The duration of hospitalization was associated with an adjusted risk ratio of 1.006 (CI 1.002–1.01), which reflects a 0.6% increase per day of hospitalization. (Figure 2).

Figure 2.

Multivariable modified Poisson regression for Early Intervention (EI) evaluation

Adjusted model was built using stepwise selection from base model including advanced maternal age at delivery, ethnicity/race, prematurity, year of birth and a dummy variable for missing information

Attendance at the DFU clinic

In the multivariable logistic regression, we found that infants exposed to antipsychotics had an associated 73% increase in likelihood of engaging in the DFU clinic (aRR 1.73, CI 1.12–2.67). Infants born to birthing parents who received their perinatal care in the specialty SUD family clinic had over to 2 times the probability of attending the DFU clinic (aRR 2.13, CI 1.07–4.24). Infants that remained in parental custody had a 38% lower probability of attending the clinic compared to infants placed with guardians or foster parents (aRR 0.62, CI 0.39–0.97). (Figure 3).

Figure 3.

Multivariable modified Poisson regression for Developmental Follow-Up Clinic attendance

Adjusted model was built using stepwise selection from base model including advanced maternal age at delivery, ethnicity/race, prematurity, year of birth and a dummy variable for missing information

Sensitivity analyses

In the first sensitivity analyses, we used mixed-effects logistic regression models for both primary outcomes. For engagement in EI, there no longer was an association with parental custody at hospital discharge, while birthing parent race and ethnicity, length of hospitalization and pediatric care location continued to be associated with higher odds of engagement. In the DFU engagement model, the previously observed associations did not continue to be significant.

For our second sensitivity analysis, including admission to a SCN or NICU in our model, we observed the same associations with EI and DFU engagement.

DISCUSSION

In a single-site retrospective cohort study of infants with intrauterine opioid exposure born at an academic medical center with a specialized clinic for families impacted by parental substance use, three quarters of the infants engaged in at least one of the two recommended developmental surveillance services, Early Intervention and/or a hospital-based Developmental Follow-Up clinic, during the first year of life. We observed attrition in the multiple steps that lead to engagement, with continuous decreases in rates of completion of referral, evaluation and enrollment in services or adherence to visits. Exposure to birthing parent antidepressants and cocaine use were both associated with higher likelihood of engagement in EI. Evaluation by EI was also found to be more likely in infants with longer hospitalizations. Attendance at the DFU clinic showed a positive association with birthing parent antipsychotic exposure and care at a perinatal SUD clinic. We also found that infants who remained in parental custody at the time of hospital discharge had lower probability of engaging in DFU services.

Our findings of engagement in EI services show better rates of engagement than previously described in this population, in terms of rates of referral, evaluation and enrollment in the first year of life. This increased rate of referral likely represents hospital-focused quality improvement efforts to increase referral rates from newborn hospitalization discharge and the specialized clinic who provided primary care to many of the infants with IOE. In terms of engagement, infants in this sample who remained in parental custody were less likely to engage. In the postpartum period, birthing parents with OUD face added challenges to the stability of their recovery, with increased rates of overdose and relapse reported in this period, in addition to the challenges typically associated with newly parenting an infant.²⁷ The reasons why infants in parental custody are less likely to utilize EI services could be linked to lack of support and increased adversity faced in this period, or to negative perceptions related to these interventions being linked to child protection services. While we were not able assess reasons for engagement, previous research has identified noticed stigma and judgment from EI professionals, and guilt related to the child’s exposure to opioids as reasons for not engaging.²⁸ Similarly, the lack of perceived value or need for these programs has been described as a barrier to adherence and using these services, not only in this population but also in other children who meet eligibility criteria based on diagnosis.^29,30 Ensuring these services are seen as resources and support, instead of surveillance and scrutiny, is crucial in providing the best care. In turn, the association of longer hospitalizations with engagement in EI could be related to perception of disease severity, which has been described previously to promote engagement in both EI and DFU. ^16,19,28,29 The question remains however, why there are infants that are referred but not evaluated. Further research is needed to understand the process of referral to evaluation and possible loci for structural and systematic programmatic change, as well as ongoing research to understand parental reasons and barrier to engage in services.

Engagement in developmental follow-up services in this population has not previously been described. The rates of engagement and adherence to these visits are lower than the 40–50% range described for other conditions that require increased developmental surveillance, such as prematurity.^23,31,32 The poorer adherence to the DFU visits compared to EI engagement could be explained in part by the fact that EI is organized as a home-based intervention facilitating access for parents, whereas DFU requires transport and is located in a hospital-setting, challenges which have been described in the literature.³³ As we continue to investigate ways to improve adherence, we should focus on making this service equitably accessible, especially in neighborhoods with very low and low COI, which represent over half of our population. We identified that the exposure to antipsychotic medications, a contributor to NOWS severity,³⁴ was associated with greater adherence to DFU visits. This finding could potentially also be related to parental and caregiver views of disease severity, which has been shown to be associated with higher engagement with extremely premature and low birthweight infants. ^23,35,36 Additionally, receiving care in a multidisciplinary clinic specializing in the care of families affected by SUD helped facilitate engagement in DFU. This relationship provided a unique opportunity for birthing parent-infant dyad continuity of care in the perinatal and postnatal period, emphasizing and supporting the needs of both the infant and the birthing parent during a crucial period of vulnerability and development. Moreover, by developing trust between providers and patients, there is the possibility to remove stigma and judgment often associated with perinatal substance use and parenting. It is also possible this finding was amplified by the fact that these individuals remained engaged in the birth hospital same health care system, and thus were more easily reachable for scheduling or rescheduling DFU visits.

Limitations

This study has several limitations. First, the study population was limited to one hospital at an urban academic medical center which may limit the generalizability to other clinical settings. Second, half of the cohort received their pediatric care outside the birthing hospital network, which limited the ability to adjust for other postpartum factors impacting engagement including stability of birthing parent recovery, housing insecurity, and social supports for the first year of life, which could affect engagement in developmental screening and services. In turn, the sample is representative of the population most affected by the opioid epidemic, non-Hispanic/Latinx white people, further limiting the generalizability of our findings in settings with a more diverse population affected by opioid use. The retrospective nature of this analysis and chart review introduces biases related to ascertainment, misclassification, and differential follow-up, which were addressed by the data quality checks. Finally, we were not able to assess a family’s perceived need for the services or any individual reasons for not utilizing these services described before, all of which could be essential barriers to engaging in developmental surveillance. ^27,28

CONCLUSION

The engagement and adherence to the recommended follow-up services in infants with history of intrauterine opioid exposure is less than optimal, with only three quarters of eligible infants utilizing at least one of the two available services, Early Intervention and Developmental Follow-Up Clinic. Factors relating to disease severity, birthing parent exposures, location of perinatal care and parental custody are associated with differences in engagement in these programs. Future work investigating structured approaches to improve engagement needs to elevate the perspectives of families on factors that will promote an environment of support and equitable care, as well as identify structural and hidden barriers that perpetuate the low continued engagement of programs aimed at serving the needs of families impacted by prenatal substance exposure.