Reducing Abuse and Neglect Recurrence Among Young Foster Children Reunified With Their Families

Of 400 total children (the roundness of this number is coincidental) who participated in the SYNCHRONY project (see below) by the St Louis County Family Court, 272 children in 223 families were reunified with their birth parents or placed in guardianship with close relatives. Original reasons for placement in protective custody (N = 383, 17 missing) included neglect only (42%), physical abuse only (33%), drug-exposed infant only (10%), neglect and physical abuse (5%), drug-exposed infant and neglect (3%), sexual abuse only (1%), sexual abuse and neglect (1%), physical and sexual abuse (0.5%), drug-exposed infant and physical abuse (0.3%), and missing (4%). Neglected children were less likely to be reunified with their families than those who had been physically abused (Table 1). There was no difference in case disposition by gender or age. For most of the 223 reunified families (n = 183) the index child was the only living child of both biological parents. In our cohort, African-American children were more likely to be reunified with their families than were white children (P = .02) (Table 2).

The 400 referrals to the SYNCHRONY Project over the program interval from 2011 to 2022 were derived from a total of 2934 children under age 6 years who were placed in protective custody of the St Louis County Court over that period. At any given juncture, cases were selected for referral on the basis of (1) concurrent availability of intervention slots within the program, which were limited by budgetary constraints; and (2) presumption by Court officers that: (a) there existed unmet mental health or educational (parent-training) needs of the family; and (b) the estimated time-to-disposition was not shorter than the time required to conduct evaluation and intervention recommendations within the program. There is no data available on children who were not referred to the program, but the children enrolled were representative of State and national child welfare samples for this age group, with respect to race, time to reunification, and the profile of maltreatment incidents for which children in this age group are referred for placement in protective custody.

At time of referral to the SYNCHRONY project, reunified children were on average 33.6 months old (2.8 years old). The children ranged in age from birth to 72 months (average 33.3 ± 34.1); 26.5% were non-Hispanic white, 67.3% were African-American, 5.1% multiracial, 0.7% Asian, and 0.4% Hispanic. The racial and ethnic composition of St Louis County according to the US Census Bureau¹⁰  is 64.7% non-Hispanic white, 25.1% African-American, 4.9% Asian, 3.1% Hispanic, and 2.4% multiracial. The 2021 demographics of children in Social Services custody in St Louis County¹¹  were 28.7% non-Hispanic white, 63.6% African-American, 7.7% Hispanic, 7.3% multiracial or race unknown, 0.2% Asian, and 0.2% Native American (total of 1463 children), so although African-Americans are significantly over-represented in the general child welfare population in the County, referrals to SYCHRONY are generally consistent with the racial and ethnic make-up of the County’s child welfare population, with the exception that Hispanic children are under-represented.

Because there are severe ethical constraints on enrolling separated families-in-crisis into research studies, we examined the aggregate proportion of reunified children who re-entered foster care (into any Missouri Court, as of March 2022, systematically tracked by the St Louis County Family Court using Case.net [https://www.courts.mo.gov/cnet]) for all children referred to the SYNCHRONY Project from 2011 to 2021. We compared re-entry statistics to contemporaneous Missouri and national administrative data on CM recidivism.

The United States Children's Bureau is a federal agency organized under the United States Department of Health and Human Services' Administration for Children and Families that monitors re-entry proportions into protective custody. In 2019, the Children’s Bureau CM re-entry proportion after a first episode of protective custody ranged by State from an average of 12% to 29% for children of all ages.¹²  Missouri’s re-entry proportion was 16.2%. Nationally, children under 5 have substantially higher re-entry proportions than older children, with one 20-state report documenting re-entry proportions after a first episode at 36% (under 1 year old), 33% (1–2 years old), and 30% (3–5 years old).¹³ 

Birth parents of children age birth to 6 years are referred to the SYNCHRONY Project (SP) by the St Louis County Family Court to address unmet mental health or educational needs of children or their families. Families are typically referred in a matter of weeks following placement in protective custody; the specific timing varies case to case as a function of when the Court team deems families ready to benefit from referral to the program. In all cases this is before disposition of the Case. Families can continue to engage with Synchrony services after disposition as long as desired. SP is a voluntary clinical program; families are not court-ordered, rather encouraged to participate by the Court when referred. The court referred families with a 0 to 6 year old child with possible untreated mental health needs, or a parent in the same family who may have untreated mental health needs and/or need for parenting education. Eighty percent (N = 319) of the 400 families referred enrolled in SP, and time to disposition (reunification, guardianship, or adoption) is shortened, on average, for referred versus nonreferred families.⁶  Translators and/or Spanish-speaking staff are available to the small number of Spanish-speaking parents who have engaged in the clinical program. The program maintains clinical data only on those families who participated clinically, and regular participation of birth parents was characteristic of—but by no means a prerequisite for—reunification. When referred, parents meet with a lead clinician assigned to them for the duration of their participation—a board-certified MD or DO child and adolescent psychiatrist or a clinical psychologist (PhD) with specialized training and expertise in infancy and early childhood. This disciplinary background represents potential “value-added” in decision-making responsive to the needs of young developing children, as elaborated in the Tulane University model for court-based intervention for infants and toddlers,¹⁴  which served as a prototype for the SP.⁶  In an initial clinical encounter, the lead clinician clarifies the premises of the program: (1) that reunification is the objective; (2) that the program’s services, delivered by members of a trained, transdisciplinary clinical team are available to the parent; and (3) that the clinical team will participate as members of the family support team and will communicate medical recommendations on behalf of the child or children in Court custody, based on continuous appraisal of what would parameterize safe visitation and ultimately safe reunification of birth parent and child. Funding for services was covered by a County government subsidy, and nonallowable costs for this fund, including transportation, were supported by donor funds; families were not charged.”

Typically, these recommendations were based on 2-generation assessment, ie, ascertainment of clinical characteristics of (1) the child, as relevant to specification of his or her developmental needs; (2) the parent-child relationship (historic and observed), as relevant to the viability of safe reunification, and any indication for parenting training or family therapy; and (3) clinical evidence of significant unmet mental health needs of birth parent(s) or children that would have implications for the safety and viability of reunification. For each family, a comprehensive set of recommendations for clinical support was a primary endpoint of initial assessment.

In all cases, recommendations were communicated to the Court and the Department of Social Services, and whenever clinically-indicated services of appropriate quality were not accessible to the family, they were provided by SP. Each family was scheduled for a minimum of quarterly visits for assessment of clinical progress, titration of clinical services, direct clinical observation of progress, and documentation of medical judgment regarding the safety of visitation and/or reunification at each juncture. In most cases, the results of serial assessment comprised evidentiary support for the Court’s decision-making regarding the appropriateness and viability of reunification.

The SP Parenting Education program used the Incredible Years curriculum for toddlers,¹⁵  a group-based curriculum keyed to video vignettes that are viewed, discussed, and role-played by parents. This was recommended to SP families whenever logistically feasible; a majority of reunified families attended 12 to 24 Incredible Years sessions, either in person, or in a virtual format during the COVID-19 pandemic. “Hands-on practice” sessions with the parents’ own children (whenever possible) supplemented the curriculum following every third session—these involved semistructured play activities that could be re-implemented at home; master’s level clinicians supported families during hands-on practice sessions; their observations helped inform clinical appraisal of parents’ progress toward safe reunification.

Data linkage was achieved in 3 successive steps. First, the St Louis County Family Court reviewed the listing of families of children age birth to 6 years referred to the SYNCHRONY Project following initial placement in protective custody since the program’s inception in 2011. Next, for those cases known to have been closed with a disposition of reunification or guardianship, identifiers were crossreferenced with Court records for the entire State of Missouri for the period from the date of disposition through March 2022 to determine whether they were subsequently re-referred for placement in protective custody. This list included families that were referred to SYNCHRONY but did not enroll (N = 37) in an intent-to-treat analysis. We exclude cases that died unrelated to maltreatment from re-entry analyses. Reunification occurs when the child is returned to the home of either the mother, father, or both. Guardianship typically refers to a return to the home of a close relative and is considered a positive permanent option for exiting care for children when they cannot return to a nuclear family of origin (exit to guardianship could be to a nonrelative, though kin guardianship is prioritized as a permanent exit for children in foster care in most states).¹⁶  It is customary for child welfare statistics to merge reunification and guardianship into a single category. Finally, aggregate statistics for the sample (see below) were calculated from an anonymized data set containing dates of rereferral of the children who were identified in the data linkage, along with nonidentifying individual-level demographic variables. We compared statistics on cumulative 10-year follow-up of the children enrolled in the SYNCHRONY Project to Missouri and National data (described above) for proportions of re-entry into protective custody following reunification. The data linkage procedure and analysis were reviewed by the Washington University Human Research Protection Office and deemed not to involve activities subject to Institutional Review Board oversight.

Of the 272 children reunified or who received guardianship, 16 (5.9%) re-entered the Missouri Court system over the course of the surveillance period (average follow-up period of 4.5 years for the 2011–2021 cohort). When re-entry did occur, it was an average of 38.8 months (SD = 27 months) after case disposition (reunification or guardianship); only 3 occurred within the first 12 months. Re-entry was not significantly predicted by race (P = .99), gender (P = .29) (Figure 1), or age of the child at time of referral to SP (odds ratio [OR] = 1.005, 95% confidence interval [CI] = 0.99–1.02). Overall, proportions of adoption, reunification and guardianship for all families referred to the SP were in keeping with national averages for care-as-usual in this age group (Table 3). Of the 272 reunified children, 234 had at least 1 parent who attended at least 1 SP service appointment.

We compared these results to contemporaneous administrative data on CM recidivism, including: (1) 2019 US Children’s Bureau data for Missouri (all ages and dispositions of cases; it is not segregated by age)¹² ; and (2) a 20-state report from the Center for State Child Welfare Data¹³  specifying re-entry proportions for young children in protective custody who exited their first episode to reunification between 2003 and 2010 (Table 3). The program is a clinical project and not a randomized controlled trial with an assigned active comparison group, and as such, findings should be interpreted through this lens. In the respective categories of guardianship and reunification, 3.4% and 7.1% were rereferred to the Court by March 2022. This is a fivefold reduction in risk over care as usual: Nationally, in care as usual the rereferral proportions for this age group are 18% (OR 7.5, CI: 2.38–23.78, P < .0001) and 35% (OR 6.1, CI: 3.48–10.75, P < .0001) respectively. When restricting the analysis to State of Missouri aggregate data for re-entry of foster children of all ages, there was a threefold reduction in risk: OR 3.09 (95% CI: 1.86–5.15), P < .0001. Additionally, families that experienced re-entry experienced it later than the national average. For example, the 20-state report indicates that the risk of re-entry is highest in the first year after reunification, whereas most families in the SYNCHRONY cohort went years before the next episode occurred.¹³ 

Here systematic data captured the outcomes of a cohort of 272 children, birth to 6 years, clinically supported and consecutively reunited following placement in protective court custody. Services were at an approximate cost of $1500 per family over Medicaid reimbursement per year of service;⁶  the children experienced a threefold to fivefold reduction in the rate of re-entry into protective custody in the years following the return to their families. This finding is in keeping with the results from selected smaller, shorter-term follow-up studies of the instantiation of supplemental support to families of young children in protective custody, as recently advocated in the Safe Baby Court Approach.¹⁷  Notably, there was racial and ethnic equity in the outcomes despite minority children, especially African-American children, being overrepresented in the County’s general child welfare population and the referred population; this contrasted with the 20-state cohort, where African-American children were slightly more likely to re-enter foster care than white, Hispanic, and children of other races and ethnicities.¹³  Here, the program emphasis was on the provision of developmental, psychiatric, and educational services that were clinically indicated and otherwise difficult or impossible for families to access in care-as-usual. A trans-disciplinary team with strong experience in infancy and early childhood—as characterized the prototypic Tulane model¹⁴ —informed serial clinical recommendations regarding the safety and viability of visitation and reunification and represented a consistent vantage point of clinical observation under circumstances of frequent turnover of assigned case managers from the Department of Social Services.

A limitation of this program evaluation is that it was not possible to determine whether the clinical services themselves, or the recommendations to the Court to parameterize safe reunification (which families were deemed clinically safe to reunify when, and under what circumstances of necessary support) constituted the most salient agent of improved long-term outcome, here operationalized by absence of re-referral to any Missouri Court. What is clear is that the reduction in rereferral could not be explained as a function of more stringent gatekeeping of reunification or guardianship, since differences between SP and the comparison groups for the proportion of all families who ultimately fell under these disposition categories were of far too small a magnitude to account for the differences in outcome. Other limitations of this evaluation are that data linkages were only available within the State of Missouri; a majority of SP families were known to continue to reside in Missouri because they continued to receive clinical supports and services from the SP following adjudication; and we compared our statistics to rereferral proportions within Missouri which would have similarly underestimated recidivism for families moving out-of-state. Any disproportion between our cohort and that comprising the comparison data in such moves would have been expected to result in the timing of rereferral to have occurred sooner on average for SP families, however we observed the opposite. Since the available State-specific contrast data were not segregated by age, the magnitude of the difference between SYNCHRONY and care as usual outcomes in Missouri that are reported in Table 3 should be viewed as conservative, since a younger sample would be expected to have higher rereferral proportions than older samples, as evident from the national data. Our cohort happened to be under-representative of families of Hispanic heritage from a national population percentage, which is typically the case in Missouri because of relatively low prevalence of this ethnic group. Finally, the analysis was restricted to aggregate statistics following individual-level data linkage and anonymization of results by a government entity; the birth parents were not individually-consented to research because it would have been unethical to do so in the immediate aftermath of loss of custody of their children, and any decision of the Department of Social Services to enroll a child in protective custody into research of this nature would have represented a potential conflict of interest.

Despite these limitations, major strengths of this program evaluation are its prospective longitudinal nature, follow-up of the outcomes of children over many years of time, and complete ascertainment of rereferral information from all courts in the State of Missouri for the entire longitudinal follow-up period through CaseNET. These data represent an example of the importance and utility of State and Court administrative data in informing best practices in the legal and social service sectors; such data can have profound implications for the wellbeing of children and cannot straightforwardly be acquired by human studies requiring individual informed consent. We urge insurers, health systems, providers of pediatric health care, and family courts to use and organize data that is available for children in protective custody in the course of care as usual and to consider the potential benefits of systematic supplementation of care along the lines of clinical support encompassed by the SYNCHRONY Project for families of young children in first episodes of protective custody.

In this cohort, evidence-based parenting education, two-generation (2GEN) psychiatric care and consultation, and active information-sharing between clinicians, the Court, and caseworkers of the Missouri Department of Social Services was associated with a very substantial reduction in re-entry into protective custody for recurrent child abuse and neglect. Given that official-report child abuse and neglect is known to exhibit causal¹⁸  dose-response effects on serious adverse mental health outcomes,⁴  such interventions are a matter of both urgency and medical necessity. Future research should consider whether they should be restricted to the aftermath of a catastrophic lapse in the care or supervision of a young child, or rather offered proactively to young families at high risk¹⁹ , as a “package” of insurance-covered, evidence-informed preventive intervention.

We thank Victoria Wright and Megan Panther for their assistance with assembly of the clinical data for this program evaluation, and to acknowledge the dedication and effort of the SYNCHRONY Project intervention team who delivered the clinical care associated with the outcomes described in this report: Ann McAndrew, LPC; Denise Wilson, LCSW; Michelle Horwitz LCSW; Molly McGrath LCSW; Laura Pons LMSW; Angela Klocke, RN, BSN, MPH; Neha Navsaria, PhD, Sandy Polanc, BS; Margaret Hobbs, MSEd; and Lexa Martinez de la Cueva.

CM

child maltreatment