First Time Contact with Child Welfare and the Long-Term Risk of Recurrence

Background and Purpose

The current study investigates the long-term risk of recurrence for young children following their first contact with the child welfare system, which builds on previous research and makes unique contributions by: 1) creating the first entry cohort for children under age 8 and estimating their recurrence trajectory for nine years; 2) untangling case risk beyond binary classification of substantiated-unsubstantiated by introducing the structural decision making (SDM) risk score; 3) examining the role of the perpetrator’s relationship and specific maltreatment classifications. Two research questions are addressed: What is the risk of re-reporting & recurrence throughout childhood? And how does the likelihood of re-reporting & recurrence vary by child and case characteristics?

Methods

This study analyses administrative records for all child maltreatment allegations in the State of Michigan, 2008 - 2025 (n= 8475831). The analytic cohort comprises all children under 8 years old who had their first allegation in 2016. Children whose removal and placement in foster care happened within the initial 30 days after their first complaint date were excluded (n= 949). The final cohort was 28,031 children.

Re-reporting is defined as any subsequent allegation. Recurrence refers to children who return with a substantiated complaint. Cox regression was used to account for variations in the window of risk observation. The models included child characteristics, the SDM risk score, case-level maltreatment types, case-level perpetrator relationships, domestic violence, substance use.

Results

Interestingly, and perhaps surprisingly, 33% of children have no subsequent contact with child protection following their initial allegation and investigation. Among the full cohort, 67% of children were associated with at least one subsequent report, and 19.5% were associated with a subsequent substantiated report. Key findings from the Cox regression models included a higher rate of re-reporting among children whose SDM risk score is identified as level II (HR = 1.36, p <.0001), who experienced multiple types of neglect (HR = 1.23, p <.0001), whose relative & non-relatives were co-involved in the maltreatment allegation compared to those whose biological mother only, biological father only, parents co-involved (HR = 1.24, p <.0001); and when domestic violence risk is present (HR = 1.23, p <.0001). The risk factors for recurrence were similar, yet the SDM risk score was no longer significant.

Conclusion

Approximately 1/3 of children who experience a maltreatment investigation never experience a subsequent investigation. The field should investigate what has improved for these families. Yet equally as interesting and perhaps concerning is that nearly 70% of children return to child protection with a subsequent complaint. Children who experience more than one type of maltreatment and children who were maltreated by more than one perpetrator are more likely to experience re-reporting & recurrence. Cumulative risks in maltreatment type and perpetrator relationships may reflect the relatively unstable living environment and complex family system, which may explain the persistence of re-reporting & recurrence. Neglect cases were particularly high risk. These findings could help inform the frontend decision-making about the allocation of limited resources to identified high-risk children and prevent future maltreatment complaints.