Too Simple of a Story: Contextualizing Child Protection Responses to Prenatal Substance Exposure

Infants and children with prenatal substance exposure (PSE) experience high rates of child protection system (CPS) involvement.1 Reporting to CPS is often initiated around the time of birth due to concerns about future risk of abuse and neglect, but reporting mandates vary by jurisdiction, state, and/or country and interpretation of these mandates can further vary by hospital and provider.2–5 Few studies have been able to look at cumulative risk of CPS reports, investigations, and substantiated abuse and neglect following PSE throughout childhood because data systems involving administrative claims, CPS data, and substance use records are rarely linked or tracked longitudinally. In this issue of Pediatrics, Powell and colleagues were able to link population-level health, substance use, and CPS data spanning a 12-year period to compare childhood CPS involvement among children with PSE compared with those without in New South Wales (NSW), Australia.6 They found, unsurprisingly, that by age 12 years, when compared with children without documented exposure, children with documented PSE had significantly higher rates of CPS reports that were screened in for investigation (90% vs 18%), CPS cases that were substantiated (61% vs 5%) and out-of-home placement episodes (39% vs 1%).There are several ways to interpret these important data. It is tempting to look at the stark differences between children with PSE compared to unexposed children and focus only on measures of substance use prenatally as a causal factor. However, it may be more illuminating to examine associations among CPS involvement and the socioeconomic vulnerabilities and systemic discrimination disproportionately experienced by individuals with prenatal substance use, as thoughtfully highlighted by Powell and colleagues in their Table 1 and discussion. Without adjusting for these key confounding variables, there is a possibility that postnatal outcomes may be overattributed to PSE.For example, in their sample, children born to mothers who were younger, living in poverty, and with history of living in public housing were more likely to have children with PSE. Also, children identified as Aboriginal or Torres Strait Islander were overrepresented in the population of those with documented PSE (38%) compared with the overall child population (<7%). Although Powell and colleagues acknowledge the transgenerational impact of colonization and the intersections between trauma, racism, and substance use that may contribute to CPS surveillance and intervention, these factors are not accounted for in the unadjusted life tables they present. Unadjusted cumulative incidence measures of CPS involvement limit our ability to understand the complex interplay of prenatal, postnatal, familial, and contextual factors that shape reporting decisions and CPS responses, including removal of a child from their biological caregiver. Quantitative researchers can apply an intersectional theoretical framework by using model adjustment strategies and grounding their analyses in the social and political contexts from which the data originate.7 Contextualized, adjusted analyses are necessary to avoid reinforcing stigmatizing narratives and to better inform policy and practice.Importantly, in comparison with measurements of child developmental or medical outcomes, interaction with CPS is not a direct health outcome but rather a policy response that should be analyzed within the political and social context in which it is deployed.8 The extent to which prenatal substance use prompts CPS intervention varies based on how caregiver substance use is statutorily reflected in child welfare policy, as well as the availability of community supports and the training and sensibilities of CPS reporters.2 This is highlighted in the author’s Figure 2, which shows that by age 1, 60% of children with PSE born in 2018 had experienced a CPS investigation compared with 35% of children with PSE born in 2007. The close to doubling of CPS investigations by age 1 year following PSE over the study period may speak more to changes in social policy, particularly given the similar doubling (0.9% to 1.7%) among the much larger population of children without PSE over that same period.Analytic techniques that account for the role of social policy and systems of care in determining family outcomes can provide needed insights into effective and replicable preventive measures. For instance, a recent article measuring the effect of policy change on CPS reporting of PSE in the United States in Connecticut exemplifies how model design can examine system performance rightly as the intermediary actor between parental presentation and family outcomes.5Prenatal substance use and its detection are markers of other axes of social disadvantage and unmet service needs—elements that are essential to address if we are to create a service system capable of preventing child maltreatment through community-based supports.9–11 In 2021, following the time period studied by Powell and colleagues, NSW issued a number of guidance materials and instituted service systems that are laudable in this space.12 Specifically, the NSW Health Child Well-being Unit provides clinical consultation support to any mandated reporter with questions about whether a CPS report is needed. The Family Connect and Support Program is a statewide network of supportive services designed to sustain custody and support family well-being. Jurisdictions across the United States are inconsistently responding to PSE and could benefit from research and practice guidance that evaluate the performance of these innovative initiatives for families affected by substance use and other social, economic, and behavioral health challenges.13Finally, it is tempting to interpret the 90% cumulative incidence of CPS reports among children with PSE by age 12 years as a sign of overwhelming risk. Yet we think one of Powell’s most compelling findings is that by age 12 years, more than 60% of children with PSE had never experienced any out-of-home placements. This suggests that, despite high rates of CPS surveillance and the confounding effects of intersecting social disadvantages, the majority of children remained in the care of their biological parents. Further investigation into what supports and systems contributed to family cohesion would be beneficial. Additionally, although Powell’s cascade of CPS involvement ends at out-of-home placement, examining rates of reunification following child removal is also important, as both removal and reunification are opportunities for interventions including engaging parents in substance use treatment and providing other supports and services.Linked administrative data sets increasingly allow for population-level analyses of the intersection between perinatal substance use and CPS involvement.3 This line of research is valuable for informing the design of better preventive, public health-oriented systems of care that both support families in their recovery journeys and safeguard children from unsafe living environments. Powell and colleagues contribute to this conversation by describing the trajectories of CPS involvement among children with and without documented PSE in NSW. Although leveraging these sorts of data sets offers valuable opportunities to explore how families navigate public systems, we encourage future research to thoughtfully incorporate intersectionality and political temporal contexts. Doing so can help ensure that the resulting narratives are both accurate and actionable.