Toward Measurement for a Whole Child Health Policy: Validity and National and State Prevalence of the Integrated Child Risk Index.

Abstract

To develop, validate and estimate national and across state prevalence on a multidimensional index that assesses the complex medical, social, and relational health risks experienced by United States children. Data from the National Survey of Children's Health were used to construct the Integrated Child Risk Index (ICRI) which includes medical health risk (MHR), social health risk (SHR) and relational health risk (RHR) domains. Confirmatory factor analysis and logistic regression analyses were employed to assess construct and predictive validity. Validity outcomes were child flourishing, school engagement/readiness, emergency room utilization and forgone care. Confirmatory factor analysis confirmed the ICRI 3-domain structure and greater correlation between MHR and RHR than MHR and SHR. Logistic regressions confirmed strong predictive validity of the ICRI for all study outcomes and ICRI scoring approaches. Nearly two-thirds of children (64.3%) with MHR also experienced SHR and/or RHR. Nearly one-third of United States children experienced risks on 2 or more ICRI domains and 15% of publicly insured children had risks on all domains (16.2%; 9.0%-25.7% across states). Significant variations were observed across states and by age, race/ethnicity, health insurance and household income. The ICRI is a valid national and state level index associated with children's flourishing and educational preparedness and emergency and forgone care. National child health policies and Medicaid risk stratification and payment models should consider children's RHR in addition to SHR and MHR. Results call for integrated systems of care with the capacity to address medical, social and relational health risks and promote well-being. Substate and clinical applications require research.