Association Between State-Level Criminal Justice–Focused Prenatal Substance Use Policies in the US and Substance Use–Related Foster Care Admissions and Family Reunification

Key Points

Question  What is the association between US state-level criminal justice–focused prenatal substance use policies and parental reunification among infants in foster care because of parental substance use?

Findings  Among 350 604 infants in this cohort study, a lower rate of parental reunification among infants in foster care was observed in US states that have adopted criminal justice–focused prenatal substance use policies.

Meaning  A more comprehensive approach to substance use disorder treatment for parents may be an effective strategy to promote parental reunification where it is possible.

Importance  States have enacted criminal justice–related substance use policies to address prenatal substance use and protect infants from adverse health effects of parental substance use. However, little is known about the consequences of these policies for permanency outcomes among infants in the foster care system in the United States.

Objectives  To evaluate the consequences of criminal justice–related prenatal substance use policies for family reunification and to examine differences in parental reunification by racial/ethnic group.

Design, Setting, and Participants  In this cohort study using data from the 2005 to 2017 Adoption and Foster Care Analysis and Reporting System, 13 cohorts of infants who entered the foster care system were followed up. States with criminal justice–related prenatal substance use policies were compared with states without such policies before and after their enactment using a discrete-time hazard model adjusted for individual covariates, state, and cohort fixed effects. The sample consisted of 350 604 infants 1 year or younger who had been removed from their home because of parental drug or alcohol use.

Main Outcomes and Measures  Length of time from entering the child welfare system to first reunification with a parent and hazard rates (HRs).

Results  Of the 350 604 infants 1 year or younger, 182 314 (52%) were boys, 251 572 (72%) were non-Hispanic white children, and 160 927 (46%) lived in US states with a criminal justice–focused prenatal substance use policy. Among those who were reunified, 36% of the reunifications occurred during the first year and 45% in the second year. Foster care infants who were removed from their homes because of parental substance use who live in states that have adopted criminal justice–oriented policies had a lower chance of reunification with a parent compared with states that have not adopted those policies (HR, 0.95; 95% CI, 0.94-0.96). Specifically, non-Hispanic black children who live in a state that has adopted criminal justice–oriented policies had a lower chance of reunification with a parent than non-Hispanic black children who live in a state that has not adopted those policies (HR, 0.87; 95% CI, 0.81-0.94).

Conclusions and Relevance  Given the child welfare system’s legal mandate to make every effort toward parental reunification, a more comprehensive treatment and supportive policy approach toward parental substance use might be warranted.

Since 1999, the number of individuals in the US who have died from an overdose involving prescription opioids and illicit drugs has increased 5-fold.¹ Although men have higher death rates from drug overdose than women, the age-adjusted rate of overdose deaths among women increased from 3.9 per 100 000 (standardized population) in 1999 to 14.4 per 100 000 in 2017 according to the National Center for Health Statistics.² Substance use–related mortality is a cause of concern because when these individuals are also parents, it raises the question of who will take care of their children. In addition, substance use–related morbidity is concerning given the high risk of substance use among pregnant women. In a recent national study,³ approximately 9% of pregnant women in the United States reported use of an illicit substance in the past month, 11.5% reported alcohol use, and approximately 15% reported using tobacco in the past 30 days. Furthermore, recent findings document an annual average of 2.7 million children younger than 2 years who lived in a household where at least 1 parent had a substance use disorder.⁴ These recent patterns have refocused national attention on prenatal substance use policies in the United States.

With the aim of addressing substance use among pregnant women, states have enacted a number of prenatal substance use policies (ie, treatment and supportive services, criminal justice–related initiatives, and health care professional reporting requirements) that are specifically aimed toward pregnant women with substance use disorder. Those states that have adopted criminal justice–related initiatives consider substance use during pregnancy as a form of child abuse under the civil child welfare statutes.⁵ For instance, South Carolina’s criminal child endangerment statute mentions that “maternal acts endangering or likely to endanger the life, comfort, or health of a viable fetus” are considered to be criminal child abuse.⁶

Although criminal justice–focused policies are intended to reduce prenatal substance use, they can have negative, unintended consequences for health behavior. Research has shown that in states with punitive prenatal substance use policies pregnant women are generally deterred from seeking care because of fear of criminal charges and loss of their child to the child welfare system.⁷ Women who sought care were less likely to provide information associated with their substance use, and those who received prenatal care and honestly shared their experiences with medical professionals acknowledged poor treatment.⁸ Not surprisingly, states with criminal justice policies have experienced lower substance use treatment admissions among pregnant women,⁹ and they are also less likely to provide medication-assisted treatment to pregnant women with an opioid use disorder.¹⁰ These are some of the reasons why the American College of Obstetricians and Gynecologists,¹¹ the American Medical Association,¹² and the American Psychiatric Association¹³ have advocated against the threat of criminal charges because of substance use during pregnancy. They have also mentioned that criminal prosecution may negatively alter maternal and birth outcomes because of the lack of adequate treatment.^11-13

Recent literature has documented a substantial increase in foster care admissions because of parental substance use.^14-16 Children of parents with substance use are at higher risk of experiencing physical abuse, neglect, and domestic violence.^17,18 According to estimates, these children are more likely to stay longer in foster care, experience re-referral to Child Protective Services,¹⁹ and are less likely to be reunified with their families.²⁰

Despite the well-documented consequences of children’s exposure to a parent’s substance use, along with the increase in the number of them being taken out of the custody of their parents and their low reunification rates, little has been done to understand the repercussions of criminal justice–focused prenatal substance use policies for family reunification rates of children who have been removed because of parental substance use. Although state-level criminal justice–focused prenatal substance use policies have been shown to be associated with an increase in the proportion of substance use–related foster care admissions among infants,²¹ the association of these policies with family reunification is understudied. Because the United States seeks to promote child protection through family reunification and stable permanent placement for children with their biological parents as much as possible,²² it is important to examine the association of current state-level punitive policies and the chance of family reunification. The aim of this study was to assess the association of these policies with family reunification. In addition, this study sought to examine differences in parental reunification by racial/ethnic group because previous studies^23,24 have documented racial/ethnic differences in the chance of parental reunification regardless of the reason for removal of the child.

The primary source of data for this cohort study was the 2005 to 2017 Adoption and Foster Care Analysis and Reporting System (AFCARS), a federally mandated data collection activity for all children covered by the protections of Title IV-B/E of the Social Security Act (§427) in the United States. Data are collected annually by the Children’s Bureau, Administration on Children and Families, US Department of Health and Human Services, and distributed by the National Data Archive on Child Abuse and Neglect. The AFCARS provides information on children in foster care who have been placed, cared for, or supervised by a state child welfare agency. The data are organized by child, with information on the child’s sociodemographic characteristics, including age, sex, birth date, and race/ethnicity, as well as the foster parent’s and adoptive parent’s sociodemographic characteristics. Data were deidentified, and based on Hofstra University policy, this study did not require approval by the Hofstra University Institutional Review Board.

The sample was restricted to infants 1 year or younger who had been removed from their home because of parental drug or alcohol abuse (this term is used in the variable we selected from the AFCARS codebook, but hereafter we refer to parental substance use). By focusing on this age group, we aimed to capture a substantial proportion of cases of child welfare admissions of infants with neonatal abstinence syndrome associated with their mothers’ substance use. This demographic group is essentially the population that is most associated with this study’s objective of assessing the consequences of criminal justice–related prenatal substance use policies for family reunification.

Children can be removed from their families by Child Protective Services for a variety of reasons associated with various kinds of abuse or neglect. However, in this study, our analysis was limited to cases in which children were removed because of parental substance use, parental alcohol use, or both. We were interested in children who were newly admitted to the child welfare system because it allowed us to accurately assess the length of time that they stayed in foster care (350 604 observations and 832 160 observation-years) and hazard rates (HRs).

The dependent variable in this analysis was the length of time (in years) from entering the child welfare system to first reunification with a parent. Although there are different kinds of permanency outcomes, such as adoption, relative care, and guardianship,²⁵ our focus was restricted to parent reunification because criminal justice–focused policies emphasize parents rather than other caregivers.

The independent variables of interest were state-level prenatal substance use policies that are focused on the criminal justice system. We identified states that have a law that classifies prenatal substance use as a criminal offense, such as child abuse.⁵ Current state policy information is updated monthly by the Guttmacher Institute (New York, New York) and is publicly available online²⁶; historical data were provided separately by them. States with policies oriented toward the criminal justice system were identified as those that either had or have a law that explicitly criminalizes prenatal substance use or those that consider prenatal substance use to be child abuse under child welfare laws. In 2017, these include Washington, DC, and the following 24 states: Alabama, Arizona, Arkansas, Colorado, Florida, Illinois, Indiana, Iowa, Louisiana, Maryland, Minnesota, Missouri, Nevada, North Dakota, Ohio, Oklahoma, Rhode Island, South Carolina, South Dakota, Texas, Utah, Virginia, Washington, and Wisconsin. The remaining 26 states did not have regulations that considered prenatal substance use as a criminal offense.

Individual covariates were the child’s sociodemographic characteristics. These included sex and race/ethnicity.

A prospective analysis was conducted of 13 cohorts of infants who entered the foster care system from 2005 to 2017. Discrete-time hazard models²⁷ with a complementary log-log specification were used to estimate the association between state-level criminal justice–oriented policies and the chance of family reunification. The primary timescale was time from entering the child welfare system to first reunification with a parent. Censoring events were any other permanency outcome (ie, adoption or guardianship). For those experiencing reunification, duration was calculated as the length of time in years from entering the child welfare system to first reunification with a parent. The first family reunification was considered as an absorbing state, which means that repeated spells were not allowed in the model. In other words, only the infant’s first placement was considered. To explore potential differences within race/ethnicity in the association between criminal justice–oriented prenatal substance use policies and parental reunification, a test for interaction was conducted. Then, the association was estimated between state-level criminal justice–oriented policies and the chance of family reunification for non-Hispanic white children and non-Hispanic black children in stratified analyses. All models were adjusted for state and cohort fixed effects, and SEs were clustered at the individual level. All analyses were conducted using Stata, version 14 (StataCorp).

Of the 350 604 infants 1 year or younger, 182 314 (52%) were boys, 251 572 (72%) were non-Hispanic white children, and 160 927 (46%) lived in states with a criminal justice–focused prenatal substance use policy. Table 1 lists summary statistics of the variables used for the analysis. Thirteen cohorts were followed up from 2005 to 2017. During this period, 32% (n = 113 323) of children were reunified with a parent. In non–criminal justice states, 33% (n = 62 917) of children were reunified with a parent, 52% (n = 98 632) were boys, and 69% (n = 130 877) were non-Hispanic white children. In criminal justice states, 31% (n = 50 406) of children were reunified with a parent, 52% (n = 83 682) were boys, and 75% (n = 120 695) were non-Hispanic white children. Among those who were reunified, 36% of the reunifications took place within the first year and 45% in the second year.

Table 2 summarizes the results of the discrete-time hazard model. The findings show a statistically significant negative association of criminal justice–oriented prenatal substance use policies with chance of parental reunification. Specifically, living in a state with criminal justice–oriented policies reduced the chance of reunification with parents (HR, 0.95; 95% CI, 0.94-0.96). In terms of race/ethnicity, non-Hispanic black children had statistically significantly lower rates of reunification than non-Hispanic white children (HR, 0.87; 95% CI, 0.86-0.88). No statistically significant difference in rates of reunification between male and female infants was observed in the analysis.

There was evidence of a difference between races/ethnicities in the association between state-level criminal justice–oriented policies and the chance of family reunification. A model stratified by race/ethnicity was estimated (Table 2). Foster care non-Hispanic black children who live in states that have adopted criminal justice–oriented policies had a lower chance of reunification compared with non-Hispanic black children who live in states that have not adopted those policies (HR, 0.87; 95% CI, 0.81-0.94). Likewise, the analysis showed that foster care non-Hispanic white children who live in states with criminal justice–oriented policies had a lower chance of reunification with a parent compared with those in states that have not adopted those policies (HR, 0.91; 95% CI, 0.88-0.95). Figure 1, Figure 2, and Figure 3 show the corresponding estimated survival rates considering the main covariate of interest of prenatal substance use policies.

This study examined the consequences of criminal justice–related prenatal substance use policies for family reunification. The analysis found that infants who entered the foster care system because of parental substance use who live in states that have criminal justice–oriented prenatal substance use policies had a lower chance of reunification with a parent compared with states that have not adopted those policies. In addition, the study found that non-Hispanic white children experienced higher parental reunification compared with non-Hispanic black children, although both non-Hispanic black and non-Hispanic white children had a lower chance of parental reunification if they lived in states with criminal justice–focused prenatal substance use policies.

Child welfare systems prioritize reunification with parents over other permanency outcomes because of policies that require them to make all reasonable efforts to intervene to address the reason for child removal so that the family can be reunified.²⁸ When parents are provided with treatment for substance use disorders and other support mechanisms, they may successfully return to their roles as parents.²⁹ However, criminal justice–focused substance use policies might be a deterrent to such outcomes by lowering the use of substance use disorder treatment, especially among mothers.⁹ This might explain the study’s finding of lower chance of parental reunification in states with criminal justice–focused prenatal substance use policies. The finding of lower parental reunification, coinciding with the rise and the peak period of the opioid epidemic, further highlights the importance of this issue because the number of children living in households led by caregivers with opioid use disorder experienced a sharp increase during this period,³⁰ including a substantial increase in overall foster care admissions because of parental substance use.¹⁶ The study’s findings also point to the importance of adopting a multifaceted comprehensive policy approach toward parental substance use because a criminal justice policy focus might not only reduce the rate of substance use treatment⁹ but also result in a higher number of infants entering the foster care system.²¹ Subsequently, these infants may have a lower chance of reuniting with their parents.

Although the opioid crisis has mostly affected non-Hispanic white adults so far, there is evidence that the crisis is also growing in black communities.^31,32 The implementation of strategies that address prevention and early intervention and how reporting, screening, investigation, and assessment are done, as well as the availability of culturally competent child welfare services, are essential to address this disproportionality.^33,34 One of the latest attempts to address this issue is the Family First Prevention Services Act of 2018, which seeks to provide mental health services, substance use treatment, and in-home parenting skill training to parents, caregivers, and even children who are at risk of being involved with the foster care system.³⁵

This study has some limitations. First, the AFCARS data do not provide information on biological parents, which did not allow us to examine the characteristics of the parent with the substance use disorder. However, our focus on new infant foster care admissions, for whom mothers are more often the primary caregivers, most likely implies that the child has entered the foster care system because of the mother’s inability to provide care. Second, we focused only on substance use–related foster care admissions. Although we are aware that parental substance use is often associated with child neglect, we did not include foster care admissions because of child neglect in the analysis. Therefore, our approach imparts a conservative bias to our estimates. The consequences of criminal justice–focused prenatal substance use policies for parental reunification among children entering the foster care system because of maltreatment might be an important avenue for future studies to explore. Third, the AFCARS does not provide details on the type of drugs used by parents; therefore, the only information available was whether children were removed because of parental substance use. Fourth, the reduced sample size when estimating associations adjusted for individual covariates, state, and cohort fixed effects could have altered the stability of these estimates. However, the direction of the associations was consistent across analyses and model specifications. Fifth, although we were able to identify other permanency outcomes, we restricted our analysis to parental reunification because criminal justice–focused policies emphasize parents rather than other caregivers. Investigating the consequences of criminal justice–focused policies for other permanency outcomes will be an important contribution to the literature that future studies might investigate.

To our knowledge, this is the first study to provide evidence that the rate of family reunification among infant foster care admissions associated with parental substance use is lower in states that have adopted criminal justice–focused prenatal substance use policies. States without this punitive policy orientation may have more child welfare services and support mechanisms available that may make it possible for parents with substance use disorders to resume their parenting roles. Given state child welfare systems’ legal mandate to make every effort toward parent reunification, additional supports, such as comprehensive approaches to substance use disorder treatment for parents, paired with complementary child welfare services (eg, family engagement, individualized treatment, and case management), may be an effective strategy to promote reunification where it is possible.^29,36 The behavioral health issues of young foster children are well documented, and the sequelae of parental separation may be mitigated by reunification as soon as it is possible to prevent future behavioral health care costs these children may incur.^37-39 In addition, a more comprehensive treatment and supportive policy approach toward parental substance use might be warranted.

Accepted for Publication: February 13, 2020.

Published Online: May 18, 2020. doi:10.1001/jamapediatrics.2020.1027

Correction: This article was corrected on August 24, 2020, to fix an error in the Abstract and the Results section about the chance of reunification with a parent for non-Hispanic black children who live in a state that has adopted criminal justice–oriented policies vs non-Hispanic black children who live in a state without those policies.

Corresponding Author: Maria X. Sanmartin, PhD, Department of Health Professions, Hofstra University, 1000 Hempstead Turnpike, Swim Center, Room 262, Hempstead, NY 11549 (maria.x.sanmartin@hofstra.edu).

Author Contributions: Dr Sanmartin had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Sanmartin, Ali, Aktas.

Acquisition, analysis, or interpretation of data: Sanmartin, Lynch, Aktas.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Sanmartin, Aktas.

Administrative, technical, or material support: Ali.

Conflict of Interest Disclosures: Dr Sanmartin reporting receiving a Faculty Development Research grant from the School of Health Professions and Human Services, Hofstra University. No other disclosures were reported.

Disclaimer: The child removal data used in this publication were made available by the National Data Archive on Child Abuse and Neglect, Cornell University, Ithaca, New York, and have been used with permission. Data from the Adoption and Foster Care Analysis and Reporting System (AFCARS) were originally collected by the Children’s Bureau. Funding for the project was provided by the Children’s Bureau, Administration on Children, Youth and Families, Administration for Children and Families, US Department of Health and Human Services. The collector of the original data, the funder, the Archive, Cornell University, and their agents or employees bear no responsibility for the analyses or interpretations presented herein. The views expressed herein are those of the authors and do not necessarily reflect the views of the Office of the Assistant Secretary for Planning and Evaluation or the US Department of Health and Human Services.