Trends in Prenatal Polysubstance Exposure

Abstract

Researchers are referring to polysubstance use as the “fourth wave” of the opioid epidemic. Prenatal polysubstance exposure is on the rise, leading to increased health and safety complications among infants and caregivers. This study examined the co-occurrence of polysubstance exposures and child maltreatment concerns using Child Abuse and Prevention Treatment Act (CAPTA) notifications made in Connecticut. Sixteen percent of notifications included polysubstance exposures. The average total number of substances identified among polysubstance cases was 2.39 (range 2–7). Each additional substance identified was shown to have 2.5 times higher odds of maltreatment report controlling for maternal age, infant race/ethnicity, and toxicology test type. Findings from this study suggest that the total number of substances is a primary factor contributing to maltreatment reports among infants with prenatal substance exposure.

Introduction

Researchers and news outlets are referring to the staggering recent increases in overdose deaths involving polysubstance use as the ‘fourth wave’ of the opioid epidemic (Friedman & Shover, 2023). Specifically, overdose deaths involving both stimulants (primarily cocaine and methamphetamines) and fentanyl increased from .6% in 2010 to 32.3% in 2021 (Friedman & Shover, 2023; NIDA, 2023). Concomitantly, prenatal polysubstance exposure is increasingly common with estimated rates of polysubstance use in pregnancy ranging from 5 to 10% of all pregnancies (Weerakoon et al., 2022). (Centers for Disease Control and Prevention (CDC), 2022). Attention to polysubstance exposure has increased due to recognition that health and safety complications are more common for pregnancies with multiple exposures (CDC, 2022; Deutsch et al., 2021). For instance, neonatal abstinence syndrome (NAS) and neonatal opioid withdrawal syndrome (NOWS) are more severe when the infant was exposed to other substances such as alcohol, tobacco, or cannabis, in addition to the opioid (Arter et al., 2021; Barry et al., 2021; Kelty & Hulse, 2017; Patrick et al., 2020). Alone, alcohol has been associated with increased risk of miscarriage, stillbirth, infant mortality, neurodevelopmental disorders, and fetal spectrum disorders (FASD) (Chang, 2020; Denny, 2019; Forray, 2016). Prenatal opioid exposure is associated with higher rates of hospitalization, greater healthcare utilization among infants, and decreased executive functioning (Arter et al., 2021; Barry et al., 2021; Kelty & Hulse, 2017; Patrick et al., 2020). And prenatal cannabis use has been linked to low birthweight and child development problems (CDC, 2022; Kharbanda et al., 2020). Together, multiple exposures may compound adverse effects.

Population prevalence estimates on polysubstance use in pregnancy, including the various combinations of substances used, are limited. However, extant literature suggests that certain substances are more commonly used among pregnant women with polysubstance use. A 2020 study found that among pregnant women using alcohol, 40% reported current use of one or more other substances (England et al., 2020). Another 2020 study found that more than half of pregnant women who engaged in nonmedical opioid use reported using more than five alcoholic drinks per day over a 1-month period of time (Jarlenski et al., 2017). Cannabis use has also been shown to increase the risk of other substance use during pregnancy (Goler et al., 2018; Young-Wolff et al., 2022). A more recent study in Delaware that specifically addressed polysubstance use found that despite cannabis being the most common single substance exposure type among infants, polysubstance exposures were most common with alcohol (88%), amphetamine (87%), barbiturate (89%), benzodiazepine (88%), cocaine (86%), and non-MOUD opiate/opioids (81%) exposures (Deutsch et al., 2022). Another recent study from Ohio found that pregnant women have 11.1 times increased odds of testing positive for another substance if they tested positive for nicotine, and have 22.7 times increased odds of testing positive for an opioid if testing positive for a benzodiazepine (Smith et al., 2022). In this same study, among mothers who tested positive for methadone or buprenorphine, 93% tested positive for cotinine (indicating nicotine use), 14.1% tested positive for fentanyl, 9.9% for morphine, and 8.5% for both fentanyl and methamphetamine (Smith et al., 2022).

In addition to health effects, attention is paid to infants with prenatal polysubstance exposure (IPPSE) due to increased risk of child maltreatment. Data has shown increased child protective service (CPS) involvement and child removals among families affected by substance use over the past twenty years (Children’s Bureau, 2018). Studies have shown that 61.2% of infants with prenatal substance exposure, inclusive of single and polysubstance exposures, are reported to CPS before the age of one (Prindle et al., 2018). Caregivers with ongoing substance use issues may be unable to provide reliable care to their infant and respond inconsistently to their infant’s cues and needs (Wells, 2009). Research has shown there is reduced neural responsiveness in caretakers using substances, impacting their ability to form healthy attachments with infants (Landi et al., 2011).

It is also important to consider the broader range of risk factors impacting child maltreatment, including socioeconomic and health factors (Nair et al., 2003). Cumulative environmental risk has been found to be a more significant predictor of poor child development and health outcomes in comparison to prenatal substance exposure (Carta et al., 2001). The cumulative risk model suggests that child maladjustment is predicted by the total number of risk factors a child is exposed to (Yumoto et al., 2008). These risks may be related to the health consequences of prenatal substance exposure as well as dangers of ongoing substance use after delivery. Maternal prenatal and postnatal polysubstance use poses increased risk of overdose, with half of drug overdose deaths involving multiple drugs (O’Donnell et al., 2017). This is likely due to the dangers of taking different substances with synergistic effects, mixing substances from different sources, and potential for drugs from the street to be laced with fentanyl (Compton et al., 2021; O’Donnell et al., 2017).

Other risk factors impacting these families may include, co-occurring mental health issues and trauma, socioeconomic barriers, domestic violence, or other serious relationship problems (Deutsch et al., 2021; Greenfield et al., 2006; Wells, 2009). For example, studies have shown women struggling with substance use disorder (SUD) have experienced higher rates of trauma and adverse childhood experiences, which can negatively impact parenting practices, and subsequently lead to child maltreatment (LeTendre & Reed, 2017; West et al., 2020; Winstanley & Stover, 2019).

Responsive to increasing prevalence and child maltreatment concerns, federal child welfare policy was amended in 2016 to require that states develop family care plans (FCP) that address the health and substance use treatment needs of infants with prenatal substance exposure and their caregivers. Understanding the prevalence of polysubstance exposure among families at risk for child maltreatment is needed to identify appropriate treatment supports for the infant and caregiver. The identification of common polysubstance exposures may provide healthcare and substance use providers the opportunity to utilize specific treatment interventions to address the unique and severe effects of polysubstance use among infants and caregivers. The purpose of the current study is to examine the state prevalence and variations in polysubstance exposures and document safety concerns among this population.

Methods

Sample

Data for this study were extracted from the Connecticut CAPTA notification system. The notification system captures 19 data elements including the date and location of the notification, type of substance exposure from a list of 11 substance types, infant toxicology test (if used) and finding, race/ethnicity data on the mother and infant, and FCP information. During the study time frame (March 15, 2019, to August 22, 2022), 7,920 notifications were submitted. Of these, 1,070 (13.5%) were excluded due to missing data. Each excluded notification included only the date and three completed data elements: (1) the reporting hospital, (2) indication that the infant was tested for substances, and (3) indication that the infant did not have intrauterine exposure. CT CAPTA policy does not require a notification for every infant who is tested for substances. Therefore, these notifications should not have been submitted in the first place. Excluded notifications were submitted in statistically similar percentages each month of data collection indicating that this is a random policy implementation failure, rather than systematic (i.e., occurred only at the beginning of the policy implementation period due to the implementation learning curve).

We used Department of Public Health (DPH) birth data from March 15, 2019 to August 22, 2022 to calculate percent and rate of total births. During this timeframe, 115,352 births were recorded in CT. Notifications were made for approximately 59 infants per 1,000 live births (5.9% of births).

Variables

Dependent variable

Within the CAPTA notification system individuals making a deidentified notification are directed to a separate maltreatment reporting system if a safety concern exists. Safety concerns included: (1) infant tested positive as a result of maternal substance misuse, (2) there is a concern that mother’s substance use will impact parental functioning, and (3) family presents with suspicions of abuse or neglect. For our analysis, if any of these three items were answered affirmatively, we recoded the case as 1=yes, safety concern, 0=no safety concern. We chose to create a dichotomous variable assessing the presence or lack of a safety concern given that a single report is made to CPS regardless of whether a family is noted as having one or all three safety concerns. Functionally speaking, the variable “safety concern (yes/no)” is measuring whether the family will be reported to CPS or not.

Independent variable

Number of substance exposures was developed from the notification data. Personnel making a notification can indicate any number of 11 substance categories (alcohol, buprenorphine, cocaine, methadone, prescription opioids, PCP, prescription benzodiazepine, non-prescription opioids, other illegal/non-prescribed medication, the misuse of prescription drug or over-the-counter medication, and cannabis). For bivariate descriptive analyses, these substance categories were recoded into 12 mutually exclusive dummy categories reflecting single substance and polysubstance types. To assess prevalence of various substance types among polysubstance exposures, we also created dichotomous variables capturing any exposure to the 11 types (e.g., 1 = any alcohol exposure, 0 = no alcohol exposure). To answer our primary research question polysubstance exposures, we created a variable summing the number of substances identified in the notification (range: 1–6).

Control variables

Mother’s age at birth was recorded as a whole number. Maternal race/ethnicity was coded into five dummy variables for each of the race/ethnicity categories: non-Hispanic (NH) white, NH African American, NH Multi-race/other race, Hispanic (any race), unknown/not disclosed, with the largest, NH white, as the reference group. Type of infant toxicology was coded into seven dummy variables for each category (1) meconium, umbilical cord; (2) meconium, urine; (3) umbilical cord; (4) urine; (5) urine, umbilical cord; (6) others; (7) infant not tested. FCP type was recorded either as “verified” (developed during pregnancy) or “developed” with a hospital provider (i.e., hospitalist, social worker) at the time the notification was provided. These FCP types were re-coded as (1) prenatal or (2) postnatal.

Data Analysis

Univariate and bivariate statistics were used to describe the sample and compare cases with single substance exposures to cases with polysubstance exposures. For comparing continuous variables (e.g., maternal age) across exposure groups, we used one-way ANOVA. For comparing categorical variables (e.g., type of substance) across exposure groups, we used crosstabs with Chi square.

We developed a logistic regression to identify the factors associated with a safety concern. For the reasons noted above, we opted to measure the dichotomized variable safety concern (yes/no), which warrants an analytic approach such as logistic regression designed to test the association between independent variables and a binary outcome variable. We included all independent and control variables mentioned above, and findings were reported with robust standard errors at the hospital level.

Results

Sample Characteristics

Table 1 presents sample characteristics. In total, close to six percent of total births received notifications (n=6850). The average age of mothers was about 28 years old. Non-Hispanic white mothers were most common (44%), followed by non-Hispanic Black (21%) and Hispanic mothers (21%). Approximately 14% of notifications did not identify the mother’s race/ethnicity. Compared to the state population of births, mothers with notifications were younger on average. Non-Hispanic white and Hispanic mothers were proportionally under-represented among notifications compared to state population data, and non-Hispanic Black mothers were over-represented.

Table 1:

Descriptive Characteristics of All Births and Notifications by Single and Polysubstance Exposure, Connecticut, March 15, 2019 – August 22, 2022 (n = 6,850)

	All Births (n = 115352)		All notifications (n = 6850)		Single Substance (n =5760)		Polysubstance (n =1090)
	Mean or N	SD or %	Mean or N	SD or %	Mean or N	SD or %	Mean or N	SD or %
Maternal Age***	31.19	5.53	28.09	5.79	27.63	5.75	30.52	5.35
Maternal Race/Ethnicity
White***	61716	53.5%	2980	43.5%	2302	40%	678	62.2%
Black***	15161	13.1%	1435	21.0%	1303	22.6%	132	12.1%
Hispanic***	30918	26.8%	1424	20.8%	1299	22.6%	125	11.5%
Other/Multi-race	7478	6.5%	79	1.2%	71	1.2%	8	0.7%
Declined/Not disclosed	79	0.07%	932	13.6%	785	13.6%	147	13.5%
Number of Substances***	-		1.22	.57	1	0	2.39	.65
Exposure Type
Alcohol***	-		200	2.9%	63	1.1%	137	12.6%
Cannabis***	-		5451	79.6%	4798	83.3%	653	59.9%
Cocaine***	-		567	8.3%	114	2.0%	453	41.6%
PCP***	-		83	1.2%	28	.5%	55	5.1%
Rx benzodiazepine***	-		175	2.6%	65	1.1%	110	10.1%
Rx opiates***	-		616	9.0%	211	3.7%	405	37.2%
NonRx opiates***	-		436	6.4%	114	2.0%	322	29.6%
Misuse***	-		22	0.3%	8	.1%	14	1.3%
Other illegal***	-		295	4.3%	52	.9%	243	22.3%
Buprenorphine***	-		353	5.2%	195	3.4%	158	14.5%
Methadone***	-		607	8.9%	226	3.9%	381	35.0%
Type of Infant Toxicology
Meconium	-		492	7.2%	401	7.0%	91	8.4%
Meconium, Umbilical Cord, & Urine	-		22	0.3%	16	.3%	6	.6%
Meconium & Urine***	-		2522	36.8%	2130	37.0%	392	36.0%
Umbilical Cord***	-		271	4.0%	194	3.4%	77	7.1%
Urine	-		1483	21.7%	1234	21.4%	249	22.8%
Urine & Umbilical Cord***	-		99	1.5%	66	1.2%	33	3.0%
Others	-		51	0.7%	43	0.8%	8	0.7%
Infant Not Tested	-		1910	27.9%	1676	29.1%	234	21.5%
Positive Infant Toxicology at Time of Notification			696	10.2%	509	8.8%	187	17.2%
FCP type
No FCP***	-		2192	32.0%	1671	29.0%	521	47.8%
Prenatal	-		1845	26.9%	1534	26.6%	311	28.5%
Postnatal***	-		2813	41.1%	2555	44.4%	258	23.7%
Safety Concern***	-		2353	34.4%	1595	27.7%	758	69.5%

Notes. FCP = family care plan. The racial categories used in the notification have been collapsed into the noted mutually exclusive categories. The substance exposure categories reflect single substance exposures or “polysubstance”, which includes any combination of two or more substances.

^*p < .05

^**p < .01

^***p < .001

Consistent with earlier research on CAPTA notifications (Sieger et al., 2022), the majority of notifications were single substance exposures (84%) with polysubstance exposure accounting for about 16% of notifications. Cannabis was the most common substance identified across both single (83%) and polysubstance (60%) exposures. Other common substances in polysubstance exposures were cocaine (42%), prescription opiates (37%), methadone (35%), and nonprescription opiates (30%). Seventy-eight percent of polysubstance notifications indicated that the infant was tested for substances. The most common type of test performed among polysubstance exposures were meconium & urine (36%) and urine (23%). The least common type of test performed among polysubstance exposures was meconium, umbilical cord, & urine (.6%). Polysubstance exposures were less likely to have a FCP (48%) in comparison to single substance exposures (32%). Likelihood of prenatal FCP were similar across both categories, while postnatal FCP were more common in single substance exposures (44%) in comparison to polysubstance exposures (24%). Notifications resulting in a safety concern were different across single and polysubstance exposures. Polysubstance exposures were significantly more likely to have a safety concern (70%) in comparison to single substance exposures (28%).

Table 2 presents descriptive sample statistics of substance type by single substance and polysubstance use. Each row is inclusive of single and polysubstance notifications that include the index substance. For example, among notifications involving any alcohol exposure (n = 200), 31.5% involved alcohol only, while 68.5% involved other substances. Notifications including cannabis had the lowest percentage of other substance use (12%). Conversely, 82% of notifications involving other illegal drugs identified other substances. Cocaine (80%), non-prescription opiates (74%), and alcohol (69%) had higher percentages of other substance use. Polysubstance use was higher for notifications involving methadone (63%) compared to buprenorphine (45%).

Table 2:

Descriptive Characteristics of Exposures by Single and Polysubstance Exposures, Connecticut, March 15, 2019 – August 22, 2022 (n = 6,850)

	All Notifications	Index Substance Only (n =5760)		Index Substance Plus Other Substances (n =1090)
	N	n	%	n	%
Index Substance
Alcohol	200	63	31.5%	137	68.5%
Cannabis	5451	4798	88.0%	653	12.0%
Cocaine	567	114	20.1%	453	79.9%
PCP	83	28	33.7%	55	66.3%
Rx benzodiazepine	175	65	37.1%	110	62.9%
Rx opiates	616	211	34.3%	405	65.8%
NonRx opiates	436	114	26.2%	322	73.9%
Misuse	22	8	36.4%	14	63.6%
Other illegal	295	52	17.6%	243	82.4%
Buprenorphine	353	195	55.2%	158	44.8%
Methadone	607	226	37.2%	381	62.8%

Factors Associated with Safety Concern

Table 3 presents factors associated with a safety concern. The largest odds ratio was observed for number of substances. For each additional substance identified in the notification, the infant had 2.5 times higher odds of a safety concern compared to notifications with fewer substances (OR=3.541, 95% CI=3.048–4.112, p<.001) controlling for other model variables.

Table 3:

Number of Substances Associated with Safety Concern, Connecticut, March 15, 2019 – August 22, 2022 (n = 6,850)

	OR	RSE	95% CI		p
			LL	UL
Number of Substances	3.541	.270	3.048	4.112	***
FCP Type
Postnatal	.132	.080	.040	.432	**
Prenatal	.253	.121	.099	.645	**
Mother Agea (years)	1.002	.006	.990	1.013
Mother Race/Ethnicity (ref. White)
Black	1.071	.179	.772	1.487
Hispanic	.821	.092	.658	1.024
Other/Multi-race	1.130	.287	.686	1.860
Declined/Not Disclosed	.742	.216	.419	1.312
Type of Substance Test (ref. Meconium & Urine)
Meconium	1.923	.688	.954	3.875
Meconium, Umbilical Cord, & Urine	4.845	1.702	2.434	9.644	***
Umbilical Cord	1.739	.620	.865	3.500
Urine	1.721	.552	.917	3.230
Urine & Umbilical Cord	2.100	1.154	.714	6.169
Others	.811	.349	.348	1.886
Not Tested	.947	.532	.315	2.847
Constant	.280	.160	.091	.860

Notes. Wald Χ² (14) = 1892.53^***, Pseudo R² = 0.1977. OR = odds ratio, RSE = robust standard error, CI = confidence interval, LL = lower limit, UL = upper limit, FCP = family care plan. The racial categories used in the notification have been collapsed into the noted mutually exclusive categories.

Interpret OR values as follows: (OR-1)^*100 = % change in odds of receiving any FCP for every unit increase in predictor variable, e.g., (.881–1)^*100 = 11.9% decrease in odds of a safety concern for Black versus White mothers.

^aMeasured in whole years (i.e., 21, 22)

^*p < .05

^**p < .01

^***p < .001

Several control variables were also significantly associated with the outcome of interest. Compared to cases with no FCP, the odds of a safety concern were about 87% lower for dyads with a postnatal FCP (OR=.132, 95% CI=.040-.432, p<.01) and 75% lower for dyads with a prenatal FCP (OR=.253, 95% CI=.099-.645, p<.05). Compared to dyads with meconium & urine testing, dyads with meconium, umbilical cord, & urine testing had 3.8 times higher odds of a safety concern (OR=4.845, 95% CI=2.434–9.644, p<.01).

Discussion

Prenatal polysubstance use compounds cumulative risk for child maltreatment due to the multiple dangers associated with prenatal and postnatal substance use. Polysubstance use can influence the causal mechanisms of child maltreatment, including more severe forms of SUD, health issues, and death among caregivers. Family environments where parents are using substances have often been characterized as chaotic and unpredictable with added socioeconomic stressors (Arria et al., 2012). Furthermore, individuals with SUD often have co-occurring mental health diagnoses or trauma histories, which can negatively impact emotion regulation in children (Suardi et al., 2017). Recent studies have indicated increases in polysubstance overdose deaths, particularly with stimulants and fentanyl (Friedman & Shover, 2023), suggesting increased likelihood of overdose during the postpartum period. Prior research has primarily focused on single substance prenatal exposures in child maltreatment cases; however, more attention must be given to polysubstance use among caregivers.

This study makes several contributions to the literature. First, we documented one state’s population prevalence of prenatal substance exposure including single and polysubstance exposures. Second, we documented common combinations of substances used in pregnancy and detected at birth. Third, we identified an association between the complexity of polysubstance exposures and infant safety concerns.

As noted in the introduction, population parameters on substance use in pregnancy are under-researched. National self-report data does not assess polysubstance use in pregnancy and previous birth cohort studies rely on medical record documentation of exposure, which is limited by billing codes. For example, Rebbe et al. (2019) measured the incidence of diagnosed prenatal substance exposure among newborns in Washington state from 2006–2013 but were only able to assess prevalence of the seven substance types included in ICD-9 diagnosis codes and did not distinguish between single and polysubstance exposed infants. Our study data came from a newly engineered online notification portal that documents incidence of prenatal exposure to up to 11 different substance types. With this expanded information, we documented that single substance exposure occurred in 5% of all births and polysubstance exposure occurred in 0.94% of all births, for a total of 5.94% of births identified as substance exposed.

Consistent with our results, the recent Delaware study revealed similar findings with the majority of IPSE having cannabis exposures and single compared to polysubstance exposures (Deutsch et al., 2022). Ours also found similar patterns of polysubstance exposure types compared to Delaware, although the specific substances identified differed. The Delaware study found polysubstance exposures to be common among alcohol (88%), amphetamine (87%), barbiturate (89%), benzodiazepine (88%), cocaine (86%), and non-MOUD opiate/opioids (81%) exposures (Deutsch et al., 2022). Comparatively, our study showed similarly high rates of polysubstance use among illegal drug use (82%), cocaine (80%), and non-prescription use (opiates) (74%), with lower percentages of polysubstance use among alcohol (69%), benzodiazepine (63%), and other prescription misuse (64%) exposures. While our findings might indicate fewer polysubstance alcohol exposures, it might also indicate that alcohol exposures were detected less frequently due to different testing procedures or detection methods.

Of note, our study detected differences in polysubstance exposures among caregivers using MOUD, with higher rates of polysubstance use among methadone exposures (63%) and buprenorphine (45%). The Delaware study found high rates of MOUD exposures in combination with other substances; however, it did not distinguish differences between types of MOUD (Deutsch et al., 2022). Our findings suggest the need for additional recovery supports among prenatal people receiving methadone maintenance treatment. There is limited literature addressing differences between buprenorphine and methadone treatment among perinatal people, but potential factors contributing to differences in prenatal substance use among those using MOUD can be related to the clinic setting, experiences with clinical staff, patient preferences, and other co-occurring psychiatric diagnoses. Research has found increased mortality rates among those in methadone treatment compared to buprenorphine treatment, potentially related to restrictive prescribing practices that may not be compatible with maintaining employment or frequent transportation (Gottlieb et al., 2023). Such factors may also be at play for prenatal people receiving methadone treatment.

Given the dearth of literature on polysubstance exposure in pregnancy, there is a similar lack of information about relative risk of child maltreatment concern and CPS reporting among infants with polysubstance vs. single substance exposures. Our previous study documented that the type of substance predicted likelihood of maltreatment report, with illegal and polysubstance exposures significantly more likely to be reported compared to marijuana exposures (Sieger et al., 2022). This corroborated similar studies in other states that documented a relatively strong effect of substance type on CPS outcome compared to other characteristics such as race/ethnicity, maternal age, or birth characteristics cocaine (Prindle et al., 2018; Rebbe et al., 2019). The current study builds on these previous findings to illuminate a cumulative risk effect—rather than specific substance, per say, it is the total number of substances an infant is exposed to that better predicts likelihood of a maltreatment report. In our study, over two-thirds of safety reports occurred among polysubstance exposures, with each additional substance identified showing 2.5 times higher odds of receiving a safety report. Future research is needed to clarify this relationship. It is likely that providers view complex substance exposures as indicative of more severe untreated substance use disorder and, thus, higher risk for child maltreatment. However, there may be other factors such as recency of exposure that explain both the toxicology findings and perceived maltreatment risk.

Lastly, our findings reveal that the majority of infants with polysubstance exposures did not have a FCP. These results are concerning as infants with polysubstance exposures were more likely to have a safety report and would benefit from the additional treatment supports identified in a FCP. Given that FCP were specifically designed to increase access to substance use treatment and early intervention services, our findings are worrisome and demonstrate discrepancies in CAPTA implementation. Although CAPTA is a child welfare policy, FCP are most often developed in CT by substance use treatment providers, social workers, or nurses; hence, additional implementation supports from the state and hospital level are needed. Further clarification regarding the roles and responsibilities of providers developing the FCP and cross-system communications and data exchange are warranted (National Center on Substance Abuse and Child Welfare, 2021).

Our study is limited in several ways. As such, results must be interpreted cautiously. First, these data were administratively collected and de-identified. Records could not be validated against hospital discharge data, so we do not know whether all IPSE received a notification. Second, and relatedly, these are cross-sectional data collected at the time of the infant’s birth. We have no way of knowing when or for how long the substance exposure occurred or how the dyad fared after the notification was made. Third, while we could control for maternal age and race/ethnicity on likelihood of outcomes, we could not account for salient risk factors such as prenatal care, insurance status, or previous child welfare system involvement. Fourth, and similarly, we could not account for salient protective factors. For instance, we have no way of measuring substance use treatment utilization in pregnancy. Finally, these findings reflect one state’s implementation that may not generalize to other settings.

Implications for Practice

The complex needs and risks for infants exposed to polysubstance use require additional support. Despite limitations including administratively collected de-identified data, results of our study suggest the total number of substances, not necessarily the specific substance type, is a primary factor contributing to safety reports among IPSE. As CAPTA continues to be developed, the role of social work within treatment settings and the child welfare system remains important. Our findings indicate the need for continued education and support for social workers who are faced with filing maltreatment reports and developing FCP within substance use treatment programs and hospital settings. Social work practitioners warrant clarification regarding when FCP should be developed, particularly if maltreatment concerns exist at the time of birth when extra treatment supports would be beneficial.

Our results also indicate the importance of additional recovery supports for perinatal people within methadone maintenance treatment. Social workers, frequently working on the frontlines of methadone maintenance treatment, should be aware of the increased risk of polysubstance use among pregnant women using methadone. This suggests the need for earlier development of FCP in pregnancy within methadone treatment settings, prompting linkage to other service providers.

Lastly, future research using longitudinal data is needed to understand the implications of polysubstance exposures and safety reports. Future research is needed to identify specific factors contributing to social work practitioners and other providers perceiving polysubstance exposures as higher risk of child maltreatment.

Data Availability Statement:

Data subject to third-party restrictions.