Impact of Plans of Safe Care on Prenatally Substance Exposed Infants

Abstract

Objective

To assess the impact of recent federal statute changes mandating child welfare-based Plan of Safe Care (POSC) supportive programming and community-based linkages to treatment providers, resources, and services for families of infants affected by prenatal substance exposure (IPSE).

Study design

Retrospective review of Delaware’s statewide child welfare case registry data for IPSE birth notifications and subsequent hotline reports for serious physical injury/fatality concerns from 11/1/2018–10/31/20. Abstracted variables included IPSE sex, substance exposure type, family characteristics (maternal personal child welfare history or mental health diagnosis, treatment engagement), and POSC referrals.

Results

Of 1436 IPSE, 1347 (93.8%) had POSC support. Most IPSE (67.2%) had exposure to single substance types prenatally. Nearly 90% avoided out-of-home placement. Nearly one-fourth of mothers delivered a prior IPSE; 40% of mothers had personal histories of childhood protective services involvement. Also, 43.5% of mothers and 9.1% of fathers were referred to community-based resources, including substance use, mental health treatment, parenting classes, and home visiting nursing. Nearly 58% of IPSE were referred for pediatric/developmental assessment. Notably, 0.82% (11/1347) of IPSE with POSC sustained serious physical or fatal injury.

Conclusions

Plans of Safe Care promote supportive, potentially protective linkages to community-based programming for IPSE and their families.

The use, misuse, or abuse of substances disproportionately affects young women of child-bearing age and has led to an increasing proportion of infants born prenatally substance exposed (IPSE).^1–4 Substance misuse and abuse often negatively impacts parenting and placement stability for infants and children in the home, representing a leading risk factor for child welfare involvement, child abuse/neglect victimization, and foster care entry nationally, particularly among those younger than age 6 years.^5–10 There is evidence that infants born prenatally substance exposed may be disproportionately impacted, with the particular substance of in utero exposure unreliably predictive of future safety risk.^11–16

Within families affected by substance use, mothers parenting IPSE may struggle with active substance addiction or substance use disorders (SUD). Federal statistics in 2017 indicate 8.5% of pregnant women endorsed illicit substance use (recreationally utilizing cocaine, opioids, and marijuana), and 11.5% consumed alcohol within the past month, representing increases compared with prior years.¹ In contrast, mothers of IPSE may be stably engaged with SUD treatment or appropriate use of pharmaceuticals for treatment of chronic physical or mental health conditions that result in positive toxicological testing or infant withdrawal symptom exhibition from in utero exposure. Despite these diverse characteristics, concerns for disproportionate stigmatization, misconception, and bias have frequently impacted mother-IPSE dyads, including fears of infant/child removal, foster care placement, criminal charges, and termination of parental rights under civil-child welfare statutes in certain jurisdictions.¹⁷

Fear of societal or legal repercussions may potentially deter eligible mothers from needed medical, mental health, and SUD treatment, although engagement in supportive, family-oriented SUD treatment may in fact decrease probability of child removal and increase likelihood of reunification via the child welfare system.^18–20 Access, entry, retention, and treatment completion among mothers may also be disproportionately impacted by practical barriers and psychosocial adversities (economic disparities, lower educational attainment, need for childcare, and few social supports) compared with male peers.^21,22 Additionally, mother-IPSE dyads frequently experience co-occurring adversities like domestic violence exposure, mental health, and socioeconomic issues necessitating service provision, agency flexibility and capacity to identify and tailor interventions specifically at the individual mother-IPSE dyad level.^10,23–29

To prevent unnecessary foster care entry, facilitate ongoing postnatal mother-IPSE dyad monitoring and intervention, and promote optimization of mother-IPSE health and safety outcomes, federal statute changes to the Child Abuse Prevention and Treatment Act now specify requirements for child welfare agencies to routinely receive IPSE birth notifications and uniformly support dyads through Plans of Safe Care (POSC) programs.^30–32 Non-punitive, family-centered linkages to community-based, multi-disciplinary resources like mental health or SUD treatment providers, parenting or job skills classes, home visiting nursing, pediatric medical homes, or neurodevelopmental specialists,³⁰ POSC typically involve a comprehensive child welfare or community-based agency assessment of mother-IPSE needs during the pregnancy period or at birth hospital discharge and initiation of abuse prevention-oriented, health and safety-optimizing programming utilizing a collaborative approach with family members, health care, and child welfare professionals.^5,30–33 Although federally mandated, state-specific interpretation of statute around POSC has varied, resulting in diverse program development and implementation across the nation with little outcome or efficacy data currently available regarding best practices or the impact on IPSE-family outcomes.^18,19,32

Our objectives were to assess IPSE-family characteristics involved with child welfare; type and distribution of POSC referrals to community-based programs for mothers, fathers, and IPSE; and trends among IPSE who sustained serious physical or fatal injury within the first 12 months of life subsequent to receipt by the agency of their birth notification.

Methods

Hotline notifications regarding IPSE birth and any subsequent hotline reports for serious physical injury/fatality concerns received by Delaware’s statewide child welfare agency and entered into an IPSE-case registry between November 1, 2018 and October 31, 2020 were retrospectively reviewed. From the case registry, abstracted variables included IPSE characteristics (sex, prenatal substance exposure type), maternal characteristics (personal childhood child welfare history, comorbid mental health concerns, treatment engagement, substance use/abuse, confirmation of valid prescription, prior IPSE delivery), and POSC characteristics (maternal, paternal, and IPSE referrals to SUD, mental health, home visiting nursing, parenting classes, and intimate partner violence counseling). For each mother-IPSE dyad, POSC provision involved screening, intake, and family-specific program development in accordance with 4 potential POSC pathways previously categorized by the state child welfare agency³⁰ (Figure; available at www.jpeds.com). Assessments involved identification of support persons for IPSE-family units, existing treatment providers and upcoming appointment dates, IPSE needs/risk assessment made at birth hospital discharge (such as exposure/withdrawal symptoms, developmental concerns, medical conditions, sleeping arrangements), maternal needs/risk assessment (including SUD or mental health, parenting skills/attachment/bonding concerns, family planning needs, basic housing/food/transportation needs), paternal needs/risk assessment (SUD, mental health parenting skills), and need for other referrals (home visiting nursing, employment/job training, intimate partner violence counseling, financial assistance). Assessments were performed by either child welfare or hospital-based social workers or by MAT providers depending on the eligibility pathway and timing of POSC preparation.³⁰ Substance exposure was ascertained from maternal/infant toxicological test results pursuant to a universal testing protocol upon admission to labor/delivery across the seven birthing centers in Delaware and from mother’s verbal admission of prenatal substance use during screening. Implementation of child safety plan agreements and out-of-home placement were also assessed, as IPSE could have had both preventive programming (the POSC) and additional safety interventions based on risk assessment, case circumstances and agency structured decision-making tools.³⁷ Cases involving serious physical or fatal injury sustained by IPSE were classified according to existing state statute and investigation protocols.³⁸

The institutional review board of our institution approved this study.

Statistical Analyses

Descriptive statistics (frequencies, percentages) for IPSE, family, and POSC characteristics were reported. A univariable logistic regression analysis was performed to explore the association between characteristics and POSC implementation, and odds ratios with 95% confidence intervals were reported. A backward stepwise multivariable logistic regression model was used to select the variables that are simultaneously associated with POSC. A p value cutoff point of 0.05 was used to retain a variable in the model. Cases involving serious physical injury/fatality were analyzed qualitatively for trends.

Results

During the study period, 1436 unique IPSE births were reported to child welfare (Table I). Most IPSE were male (n = 775, 54.0%), and most mothers identified as white race (n = 789, 54.9%). Most mothers had comorbid mental health concerns (n = 798, 55.6%), and 40.3% (n = 578) were involved with child welfare personally during childhood or reported childhood trauma. Nearly 23.7% (n = 340) of mothers had delivered a prior IPSE. Most mothers (n = 924, 64.3%) did not possess a valid prescription for substances identified on toxicology, and most cases (n = 883, 61.5%) involved recreational substance abuse concerns versus appropriate medicinal use (n = 154, 10.7%).

Table I.

Infants prenatally substance exposed involved with child welfare

	IPSE n = 1436 (%)	With POSC n = 1347 (%)	Without POSC n = 89 (%)	Odds ratio* (95% CI)	P value
Sex
Female	660 (46.0)	620 (46.0)	40 (44.9)	Ref†
Male	775 (54.0)	727 (54.0)	48 (53.9)	0.98 (0.63, 1.51)	0.92
Unknown	1 (0.1)	0 (0)	1 (1.1)	-	-
Race
White	789 (54.9)	746 (55.4)	43 (48.3)	Ref†
Black	602 (41.9)	559 (41.5)	43 (48.3)	0.75 (0.48, 1.16)	0.20
Hispanic	35 (2.4)	34 (2.5)	1 (1.1)	1.96 (0.41, 35.24)	0.51
Other/unknown	10 (0.7)	8 (0.6)	2 (2.2)	0.23 (0.06, 1.56)	0.07
Substance Exposure
Single	965 (67.2)	913 (67.8)	52 (58.4)	Ref†
Polysubstance	471 (32.8)	434 (32.2)	37 (41.6)	0.67 (0.43, 1.04)	0.07
Valid Prescription
No	924 (64.3)	870 (64.6)	54 (60.7)	Ref†
Yes	508 (35.4)	473 (35.1)	35 (39.3)	0.84 (0.54, 1.31)	0.43
Unknown/NA	4 (0.3)	4 (0.3)	0 (0)	-	-
Substance Use
Abuse	883 (61.5)	823 (61.1)	60 (67.4)	Ref†
Medicinal use	154 (10.7)	151 (11.2)	3 (3.4)	3.67 (1.34, 15.17)	0.03
Unknown/NA	399 (27.8)	373 (27.7)	26 (29.2)	1.05 (0.66, 1.71)	0.85
CPS Screening
Out	30 (2.1)	18 (1.3)	12 (13.5)
In	1406 (97.9)	1329 (98.7)	77 (86.5)	11.51 (5.23, 24.54)	<0.001
Mother Already Engaged in Treatment Prior to POSC
No	392 (27.3)	392 (29.1)	0 (0)
Yes	480 (33.4)	479 (35.6)	1 (1.1)	-	-
Unknown/NA	564 (39.3)	476 (35.3)	88 (98.9)	-	-
Safety Plan Instituted
No	1002 (69.8)	954 (70.8)	48 (53.9)	Ref†
Yes	424 (29.5)	392 (29.1)	32 (36.0)	0.62 (0.39, 0.99)	0.04
Unknown/NA	10 (0.7)	1 (0.1)	9 (10.1)	0.01 (0, 0.03)	<0.001
Infant Out-of-Home Placement
No	1267 (88.2)	1214 (90.1)	53 (59.6)	Ref†
Yes	162 (11.3)	133 (9.9)	29 (32.6)	0.2 (0.12, 0.33)	<0.001
Unknown/NA	7 (0.5)	0 (0)	7 (7.9)	-
Maternal Delivery of Prior IPSE
No	1095 (76.3)	1034 (76.8)	61 (68.5)	Ref†
Yes	340 (23.7)	312 (23.2)	28 (31.5)	0.66 (0.42, 1.06)	0.08
Missing	1 (0.1)	1 (0.1)	0 (0)	-
Maternal Mental Health Concerns
Yes	798 (55.6)	741 (55.0)	57 (64.0)	Ref†
No	638 (44.4)	606 (45.0)	32 (36.0)	1.46 (0.94, 2.3)	0.10
Maternal CPS History as a Child
Yes	578 (40.3)	544 (40.4)	34 (38.2)	Ref†
No	855 (59.5)	800 (59.4)	55 (61.8)	0.91 (0.58, 1.41)	0.67
Missing	3 (0.2)	3 (0.2)	0 (0)

Note:

^*Odds ratio of association with POSC;

^†reference group.

CI (confidence interval), CPS (child protective services), IPSE (infants born prenatally substance exposed), NA (not available), POSC (Plan of Safe Care)

Most IPSE (n=965, 67.2%) were exposed to a single type of substance versus polysubstance (n = 471, 32.8%) exposures (2 or more different substances) (Table I). The most common exposure categories were marijuana, medication assisted treatment (MAT) exposures, non-MAT opiate exposures (heroin, morphine), and cocaine (Table II). Most IPSE (n = 670, 74.4%) were exposed to only marijuana; 16.6% (n = 150) exposures were marijuana plus one other substance, versus 9% (n = 81) exposures that were marijuana plus two other substances. Of the exposures to only one type of substance, MAT (methadone, subutex, suboxone) represented 18% (n = 173) (i.e., those on MAT only, out of those with only one substance exposure). Whereas most marijuana exposures were not combined with other substances (Table III), in contrast, most alcohol (88.0%), amphetamine (87.3%), barbiturate (88.9%), benzodiazepine (88.1%), cocaine (85.9%), and non-MAT opiate/opioid (81.0%) exposures were in combination with other substances (polysubstance exposures).

Table II.

Substance Exposure Types

Categories	n	%
Marijuana	901
Marijuana only	670	74.4
Marijuana plus 1 other drug	150	16.6
Marijuana plus 2 or more other drugs	81	9.0
Opiates and opioids (excluding medication-assisted treatment)	357
Opiates only	68	19.0
Opiates plus 1 other drug	124	34.7
Opiates plus 2 or more other drugs	165	46.2
Opiates medication assisted treatment (MAT)	464
MAT only	173	37.3
MAT plus 1 other drug	135	29.1
MAT plus 2 or more other drugs	156	33.6
Cocaine	191
Cocaine only	27	14.1
Cocaine plus 1 other drug	61	31.9
Cocaine plus 2 or more other drugs	103	53.9
Benzodiazepines	59
Benzodiazepines only	7	11.9
Benzodiazepines plus 1 other drug	20	33.9
Benzodiazepines plus 2 or more other drugs	32	54.2
Amphetamines	71
Amphetamines only	9	12.7
Amphetamines plus 1 other drug	30	42.3
Amphetamines plus 2 or more other drugs	32	45.1
Barbiturates	9
Barbiturates only	1	11.1
Barbiturates plus 1 other drug	6	66.7
Barbiturates plus 2 or more other drugs	2	22.2
Methamphetamine	3
Methamphetamine only	1	33.3
Methamphetamine plus 1 other	0	0
Methamphetamine plus 2 or more other drugs	2	66.7
Alcohol	25
Alcohol only	3	12.0
Alcohol plus 1 other drug	10	40.0
Alcohol plus 2 or more other drugs	12	48.0
Antipsychotics and selective serotonin reuptake inhibitors (SSRIs)	4
Antipsychotics and SSRIs only	4	100.0
Antipsychotics and SSRIs plus 1 other drug	0	0
Antipsychotics and SSRIs plus 2 or more other drugs	0	0
Other	6
Other only	2	33.3
Other plus 1 other drug	2	33.3
Other plus 2 or more other drugs	2	33.3

Table III.

Single vs polysubstance exposure types

Substance	Single exposure (% of total with exposure)	Poly exposure (% of total with exposure)
Alcohol	12.0	88.0
Amphetamines	12.7	87.3
Barbiturates	11.1	88.9
Benzodiazepines	11.9	88.1
Cocaine	14.1	85.9
Opiates/opioids/non-MAT	19.0	81.0
Opiates MAT	37.3	62.7
Marijuana	74.4	25.6
Methamphetamines	33.3	66.7
Antipsychotics/SSRIs	100.0	0.0
Other*	33.3	66.7

Note:

^*kratom, phenylcyclohexyl piperidine

MAT (medication assisted treatment), SSRIs (selective serotonin reuptake inhibitors)

For 1347 (93.8%) IPSE, a POSC was developed and implemented. No POSC was developed for 89 (6.2%), due to either being placed out of the home at birth hospital discharge, active family involvement with child welfare at the time of infant birth resulting in either safety plan implementation or treatment service provision instead (i.e., the agency was already involved due to safety concerns involving another child in the home), or due to the infant being born in Delaware but residing out of state warranting external child welfare involvement. Most infants (88.2%, n = 1267) avoided out-of-home placement. Maternal substance use, screened in status, no safety plan institution, and infant non-removal were associated with increased odds of POSC development (Table I). Also, no prior IPSE deliveries and no maternal mental health concern tended to be associated with increased odds of POSC development. In multivariable logistic regression, maternal race, substance abuse, child protective services screen, and infant placement remained in the model indicating a significant simultaneous association of these variables with POSC (Table IV).

Table IV.

Factors significantly associated with Plans of Safe Care in a multivariable stepwise backward logistic regression model for variable selection

Variable	Odds ratio (95% CI)	P value
Maternal race
Caucasian	Ref
African American	0.57 (0.35, 0.93)	0.02
Hispanic	1.11 (0.22, 20.16)	0.92
Unknown	0.32 (0.06, 2.5)	0.21
Substance exposure
Abuse	Ref
Use	3.9 (1.34, 16.97)	0.03
Unknown	0.96 (0.57, 1.67)	0.88
CPS screening
Out	Ref
In	9.79 (3.6, 24.81)	<0.001
Infant out-of-home placement
No	Ref
Yes	0.18 (0.11, 0.3)	<0.001
Unknown	-	-

CPS (child protective services), CI (confidence interval)

For 43.5% (n = 586) IPSE, mothers were referred to community-based programming through POSC (Table V; available at www.jpeds.com). Over one-fourth (27.1%, n = 365) of mothers were newly referred to SUD treatment and mental health treatment providers (13.1%, n = 176), 10.8% (n = 145) mothers were referred to parenting classes, and 8.6% (n = 116) were referred for home visiting nursing services. Of mothers, 34.7% (n = 498) were known to be engaged in services prior to POSC. Of these mothers, 83.3% (n = 415) were engaged in SUD, 35.5% (n = 177) were engaged in mental health, and 7.6% (n = 38) were engaged in other programs like DV or WIC. For 122 IPSE cases with POSC (or 9.1%; Table V), fathers were referred to community-based services; 4.8% (n = 64) of fathers were referred for SUD treatment, whereas 1.0% (n = 13) were referred for mental health treatment. Plans of Safe Care involving both mothers and fathers were implemented for 109 IPSE. Most IPSE (57.8%, n = 779) were referred for services, including nearly half (48.7%, n = 656) referred to a pediatric medical home for longitudinal follow up and nearly one-fifth (19.7%, n = 265) to Developmental Pediatrics/Early Intervention services for neurodevelopmental evaluations related to in utero substance exposure.

Table V.

Plans of Safe Care referrals for mothers and fathers

Total Plans of Safe Care, n =1347	Mother, n (%)	Father, n (%)
Total persons with POSC referrals	586 (43.5)	122 (9.1)
Total number of POSC referrals	860	140
Home visiting nursing	116 (8.6)	7 (0.5)
Mental health	176 (13.1)	13 (1.0)
Substance use disorder	365 (27.1)	64 (4.8)
Parenting class	145 (10.8)	43 (3.2)
Intimate partner violence counseling	13 (1.0)	0
Housing	8 (0.6)	0
Special supplemental nutrition program for women, infants, and children	9 (0.7)	0
Other*	28 (2.1)	13 (1.0)

Note:

^*Other: Cribs for Kids, Medicaid, Prenatal home visiting, post-natal Parents as Teachers early childhood educational support.

POSC (Plans of Safe Care)

In qualitative analysis, 11 IPSE with POSC (0.82%, n = 11/1347) sustained serious or fatal injury within the first year of life³⁸ (Table VI; available at www.jpeds.com). Of those, 45.5% (n = 5) involved exposures to single substances prenatally, more than half of which (60.0%, n = 3) were marijuana-only exposures. Fractures were involved in 45.5% (n = 5) of cases; 27.3% (n = 3) involved bed-sharing fatalities. Most mothers (63.6%, n = 8) were involved in treatment services prior to POSC implementation. Polysubstance exposures were overrepresented, accounting for 54.5% of all exposures in this cohort.

Table VI.

Qualitative analysis of serious physical injury and fatality cases (online)

Single vs polysubstance exposures	Substances identified	Screen in or out	Child safety plan	Plan of Safe Care	Plan of Safe Care referrals (maternal)	Plan of Safe Care referrals (paternal)	Plan of Safe Care referrals (infant)	Maternal mental health condition	Maternal childhood CPS history	Prior IPSE birth	Maternal treatment/service linkage prior to POSC	Event Type	Age at injury (months)
Single	Marijuana	In	Yes	Yes	No	No	Yes-pediatrician	Yes	Yes	No	No	Asphyxiation, cardiac arrest	2
Single	Methadone	In	No	Yes	Yes4	Yes4	No	No	Yes	No	Yes1	Fracture	1
Single	Marijuana	In	No	Yes	Yes1	No	No	No	No	No	No	Fracture	2
Poly	Fentanyl, methadone	In	Yes	Yes	No	No	Yes-pediatrician	Yes	Yes	No	Yes1	Fracture	6
Single	Marijuana	In	No	Yes	Yes5	No	Yes-pediatrician	Yes	Yes	Yes	Yes3	Fracture	3
Poly	Opiates, methadone, cocaine	In	No	Yes	No	No	Yes-pediatrician	Yes	No	No	Yes1	Bedsharing/unsafe sleep	1
Single	Subutex	In	Yes	Yes	No	No	Yes-pediatrician	Yes	Yes	No	Yes1,3	Fracture	2
Poly	Marijuana, methadone	In	Yes	Yes	No	No	Yes-pediatrician	Yes	No	No	Yes1,3	Fracture	.33
Poly	Cocaine, alcohol	In	Yes	Yes	Yes1,2,4	No	Yes-pediatrician	No	No	No	No	Bedsharing/unsafe sleep	3
Poly	Cocaine, fentanyl, benzodiazepines, marijuana, methadone	In	Yes	Yes	No	No	Yes-pediatrician	Yes	Yes	Yes	Yes1	Bedsharing/unsafe sleep	2
Poly	Methadone, opiates, fentanyl	In	Yes	Yes	No	Yes1	Yes-pediatrician	No	No	No	Yes1	Burn	.7

Note:

¹SUD (substance use disorder),

²HVN (home visiting nursing),

³MH (mental health),

⁴parenting classes.

CPS (child protective services), IPSE (infant prenatally substance exposed), POSC (Plan of Safe Care)

Discussion

How best to support mother-IPSE dyads affected by substance use remains an important and complex public health and policy issue, complicated by profound population heterogeneity, diverse multiplicity of family needs, and wide latitude afforded to states to define population eligibility criteria and tailor community-based POSC service provision.³⁰ Scant evidence exists around best practices or specific family referrals/interventions that positively impact IPSE-family health and safety outcomes^18,19,32; existing data indicate only that POSC enactment across jurisdictions has increased, with little case-level detail. In fiscal year 2018 (the first year of federal reporting of POSC-related child welfare statistics), 13 U.S. states reported 64.1% of IPSE received POSC support; 2019 data suggest of 38,625 IPSE reported to child welfare, most (75.4%) were provided POSC support.⁵ This study therefore describes, on a statewide level, POSC interventions, family needs and characteristics, and impact of POSC on subsequent physical/fatal injury among IPSE receiving child welfare support intended to be protective and health and safety optimizing.

Results suggest that mothers of IPSE are diverse, many struggling with substance misuse or abuse (lacking valid prescriptions for identified substances on toxicological testing), facing co-occurring psychosocial adversities (including comorbid mental health disorders), and having histories complicated by prior personal childhood involvement with the state’s child welfare system. Although approximately one-third of mothers were already engaged with either mental health or SUD treatment prior to their infant’s birth and POSC implementation, the need for POSC-facilitated treatment linkages demonstrated by our results suggests a potentially significant gap in mothers’ access to and of services prenatally; this may reflect maternal fears around societal, civil and criminal repercussions or a need to reassess efficacy of current screening practices and service provision linkages by health care professionals interacting with mothers prenatally, such as obstetricians and gynecologists. Our results highlight that POSC play a critically important prevention role, potentially mitigating future harm by bridging access to needed therapeutic services within a non-punitive framework. Although the proportion of mothers referred specifically to SUD treatment post-delivery versus during the pregnancy period was not specifically assessed (as POSC in Delaware could have been developed and implemented prior to IPSE birth),³⁰ nonetheless, results suggest opportunities exist through either clinical practice changes, education, or screening to enhance earlier linkage to SUD treatment by maternal health care professionals and should be explored.

Similar to prior published literature,^27–29 results of this study support that IPSE-family units have diverse needs, including co-occurring need for parenting or job skills support, intimate partner violence counseling, home visiting nursing, and identification of a pediatric and developmental medical home. Notably, nearly 10% of fathers were also referred for treatment services and support, with nearly half of the referred population requiring linkage with SUD treatment. However, data regarding fathers in the case registry was profoundly limited, resulting from either lack of family involvement or identification; only 109 POSC involved both maternal and paternal participation. Enhancing paternal engagement in POSC programming represents an area of priority for many child welfare agencies nationally. To date, there exists a paucity of research on the role of paternal engagement and development of interventions specifically targeting fathers within families affected by substance use, and, thus, our assessment of paternal service needs contributes to this limited evidence base regarding paternal characteristics.^39,40

Nearly one-half of IPSE were referred to pediatric medical homes for longitudinal follow up, with nearly one-fifth referred for neurodevelopmental evaluations. This high referral rate likely suggests effective impact of education and recognition within the child welfare community of potential need for earlier, ongoing assessment or additional involvement of therapeutic services (physical, occupational, speech therapies) to ensure IPSE optimal development given reported associations of prenatal substance exposure with diminished cognitive performance, behavior and attention difficulties, impulsivity, and issues with executive functioning.^41,42 High referral rates to neurodevelopmental services among IPSE may provide future impetus to establish universal referral practices within Delaware’s child welfare system as standard POSC programming, although, to date, referral remains selective.

Results from prior studies suggest an association between exposure to any substance prenatally and increased likelihood of child welfare reports, substantiation, out-of-home placement, and physical abuse hospitalization,^23–25 although variability among these outcomes exist based on the specific exposure type.^43–47 In part, this may relate to risk-related misperceptions, arbitrarily associated with a particular type of substance exposure. For example, prior data suggest IPSE exposed to marijuana are less likely to have a child welfare report, substantiated report, or out-of-home placement compared with IPSE exposed to other substances, whereas IPSE exposed to opioids are more likely to have a report, substantiation, and out-of-home placement compared with IPSE exposed to other substances; however, caregiver cannabis use has been linked with child abuse, challenging presumptions about low risk or benign nature of recreational marijuana use.^44,48 Results from this study suggest that most maternal marijuana cases were screened in by child welfare (with subsequent POSC preparation) but did not result in out-of-home placement or safety plan institution. Based on thorough family assessment, marijuana use (which is illegal in Delaware) was determined to be maternal substance abuse and not medicinal use, with a very low rate of medical marijuana prescriptions confirmed (Table VII; available at www.jpeds.com). Although only a few marijuana-exposed infants sustained serious physical or fatal injury, of the total exposed to marijuana in the study population, marijuana was involved in 5 of these 11 total cases, suggesting that its use is associated with risk (Table VI). Recognition of marijuana-associated risk is critically important, given likelihood of national increases in marijuana-exposed infants in the wake of increasing, impending state-level passage of recreational marijuana use laws; both national legislative trends and our study results suggest efforts to substantially increase education and anticipatory guidance across settings (in schools, at prenatal or adolescent health care visits) on the potential negative health and safety impacts of in utero marijuana exposure are urgently needed. Referral patterns were also assessed by substance exposure type (Table VIII; available at www.jpeds.com), including for marijuana, and results indicate that identified substance type did not specifically predict need for treatment or community-based referrals. Except for exposure to antipsychotic medications and selective serotonin reuptake inhibitors, referrals were provided for all circumstances of substance exposure.

Table VII.

Population characteristics for IPSEs with marijuana-only exposures (online)

Variable	n	%
Sex
Male	361	53.9
Female	308	46.0
Unknown	1	0.1
Maternal race/ethnicity
White	265	39.6
Black	378	56.4
Hispanic	21	3.1
Other	6	0.9
Confirmed valid prescription for marijuana
No	665	99.3
Yes	5	0.7
Marijuana abuse or appropriate use
Use	82	12.2
Abuse	470	70.1
Unknown/Blank	113	16.9
Medicinal/prescription use	5	0.7
Screened in or out
In	662	98.8
Out	8	1.2
Plan of Safe Care prepared
No	24	3.6
Yes	646	96.4
Mother engaged in treatment prior to Plan of Safe Care
Yes	40	6.0
No	248	37.0
Unknown or NA	382	57.0
Safety plan instituted
Yes	74	11.0
No	590	88.1
NA, unknown, blank	6	0.9
Out of home placement
Yes	13	1.9
No	654	97.6
Unknown, blank	3	0.4
Child abuse or neglect victim
Yes	5	0.7
No	665	99.3

NA (not available)

Table VIII.

Proportion of Plan of Safe Care referrals across maternal substance exposures (online)

Substance exposures	POSC referrals made: no (n, %)	POSC referrals made: yes (n, %)
Alcohol	1, 4.0	24, 96.0
Amphetamines	17, 23.9	54, 76.1
Barbiturates	1, 11.1	8, 88.9
Benzodiazepines	11, 18.6	48, 81.4
Cocaine	32, 16.8	159, 83.2
Opiate non-MAT, opioids	59, 16.5	298, 83.5
Marijuana	442, 49.1	459, 50.9
Methamphetamines	2, 66.7	1, 33.3
Antipsychotics, SSRIs	4, 100.0	0, 0.0
Opiate MAT	93, 20.0	371, 80.0
Other	1, 16.7	5, 83.3

MAT (medication assisted treatment), POSC (Plan of Safe Care), SSRI (selective serotonin reuptake inhibitor)

Historically, families facing substance use concerns with child welfare involvement have met intensive case management plans attempting to address a broad range of co-occurring issues but frequently raising concerns regarding efficacy, diminishing returns, disjointed service provision, and the need to develop more effective assessment processes and family support systems; the requisite number of services may overwhelm a parent, rendering them unable to meet the burden.^12,17–20 Growing evidence therefore supports that individualized assessments informed by existing family strengths and support networks may be more effective than traditional child welfare approaches to this heterogeneous population.^21,30 Although results suggest POSC facilitated necessary linkages between IPSE-family units and services based on the individualized family assessments, which specific POSC referrals best support families affected by SUD was not directly assessed within this study. Efficacy of specific community-based referrals or programmatic support should be informed by qualitative analysis utilizing direct family perspectives, such as through surveys or focus groups, which serve as future directions for research in this area.

Finally, regarding the proportion of seriously physically injured or fatally injured ISPE within our cohort, a previous analysis of factors associated with child welfare involvement and safety outcomes among IPSE utilizing a retrospective review of child welfare case registry in Delaware pre-POSC implementation demonstrated that 2.3% of IPSE sustained serious or fatal injury before age 1 year.¹⁰ Although polysubstance exposure and maternal mental health conditions were factors associated with child welfare involvement in this earlier study, neither substance exposure type nor mental health condition reliably predicted which IPSE would suffer later insult.¹⁰ Notably in this present study, 0.82% of IPSE supported by POSC sustained subsequent physical injury/fatality. For infants in Delaware, abuse victimization rates of 11.8 per 1000, or 11.9 per 1000 (1.2%) have been reported.^1,47 Although causal relationship certainly cannot be established and additional prospective research is needed, low rates of serious physical or fatal injury may intimate that POSC offer additional support, protection, and outcomes monitoring that reduce harm likelihood among IPSE.

Several notable study limitations exist. Retrospective review of case registry data from a single state was utilized, whose approximate population, annual birth rate and number of birthing centers (1 million residents, 11,000 live births, respectively) may limit generalizability of results to larger jurisdictions. Longitudinal engagement by families with services and efficacy of interventions after referral was not assessed due to limited data availability. Although universal toxicology screening and testing protocols across the birthing centers in Delaware exist, no single laboratory, specimen (blood, urine, meconium), or testing panel is utilized across settings, which may impact ascertained results. Despite this variability, universal toxicology assessment strengthens validity of study results as bias towards screening based on certain maternal characteristics (like race/ethnicity) is eliminated. Mother-IPSE outcomes were not specifically compared or stratified by the jurisdiction’s four potential POSC pathways; this represents an area of future research. Additional study is needed to inform impact of POSC on child abuse/neglect risk reduction.

Mother-IPSE dyads have diverse needs, but many families involved with child welfare after IPSE birth require linkage to community-based resources for treatment and support around SUD, mental health, parenting, intimate partner violence, and other psychosocial adversities. Although additional research is needed, results from a statewide assessment of POSC implementation suggest POSC promote supportive, potentially protective linkages to community-based programming for IPSE and their families.

Supplementary Material

1

Figure 1. Delaware Plan of Safe Care.³⁶

Abbreviations:

IPSE

infants prenatally substance-exposed

MAT

medication-assisted treatment

POSC

Plans of Safe Care

SUD

substance use disorder