Unveiling Substantial Risk In Childhood Using Data Linkage

Adverse Childhood Experiences (ACEs) are foundational risk factors for poor health, behavioral, and legal outcomes across the life course. Conventional ACE measurement relies on retrospective self-report, which is vulnerable to memory failure, social desirability bias, and non-disclosure — particularly for the most severe and stigmatized adversities. These limitations systematically underestimate burden and obscure polyvictimization among children most in need of intervention. This study addresses that gap through a national, population-level linkage of administrative records, offering a more objective ACE surveillance framework with direct implications for child welfare practice and policy.

Methods

Using retrospective administrative records held by the Israel Central Bureau of Statistics (ICBS), we analyzed records for all children (N = 960,334) with documentation across child welfare, national population registry, criminal justice, health, and education systems spanning 1990–2019, with welfare-specific records available from 2000–2019. Deterministic linkage via permanent pseudonymous national identifiers connected each child's records to those of biological or adoptive parents and siblings. Fifteen ACE types were operationalized, guided by Finkelhor's expanded ACE framework, including direct maltreatment (physical abuse, sexual abuse, physical neglect), household dysfunction (parental substance abuse, mental illness, domestic violence, criminal involvement), and non-traditional adversities (war zone exposure, parental sex work, natural disasters). Individual ACE profiles were constructed by aggregating all documented indicators per child across the full observation window. Gender differences in ACE prevalence were examined using chi-square tests with Yates' continuity correction.

Results

Approximately 41% of children (n = 397,199) had at least one documented ACE. Among those, 12.6% (n = 49,996) experienced two or more distinct ACE types. Critically, 87,303 children had ACE indicators revealed exclusively through cross-system linkage, invisible within any single administrative source. The most prevalent ACEs were parental loss between ages 7–18 (n = 138,862) and mother treated violently (n = 110,789). Among polyvictimized children, co-exposure to mother treated violently and physical abuse was the most common profile. Gender differences were modest overall but significant for specific ACE types: females were overrepresented in sexual abuse records (56.2%), while males predominated in physical abuse, neglect, household substance abuse, and parental loss.

Conclusions and Implications

Administrative data linkage detected a substantially broader and more complex picture of childhood adversity than self-report or single-system approaches can capture. The 87,000 children identified solely through linkage represent a population whose needs would remain invisible and unaddressed under conventional surveillance. For social work practice and policy, these findings argue for integrating cross-system data infrastructure into national child welfare frameworks to enable more equitable identification of at-risk children, especially those unlikely to self-disclose. This methodology offers a scalable, reproducible model for ACE surveillance, supporting the design of prevention strategies and democratic accountability in child welfare systems globally.