Improving Foster Care Outcomes via Cross-Sector Data and Interoperability

Abstract

Social workers, educators, legal advocates, and physicians and other health care professionals depend on data to diagnose concerns, develop treatment plans, and evaluate outcomes for foster children, who account for 425 000 children in the US annually. 1 However, data concerning child welfare issues are often substantially decentralized, protected, and even suppressed in the US, limiting the effectiveness of interventions and subsequent social and health outcomes. Research inquiry and implementation of effective polices are limited by the lack of a population health infrastructure to establish the evidence for a more effective US child welfare system. While health information exchanges are used by third-party payers in managing risk of their subscribers, these data, with rare exception, have not been used with other sector data systems to affect relevant and meaningful outcomes.In this issue of JAMA Pediatrics, Sariaslan et al 2 determine associations between being placed in out-of-home care and subsequent risks of multiple clinical and social outcomes at a population level, using the benefits of integrated data systems in Finland. Using a retrospective cohort study design, the authors convened a population sample (N = 885 000) of all children born in Finland between 1986 and 2000, with longitudinal follow-up from age 15 years to an end point (migration, death, or the end of the study), with follow-up for nearly a decade. An out-of-home cohort and cosibling subcohort were identified for comparisons with the general population. While the study 2 is important on its own merit for the results gathered and reported, it has further relevance for US researchers attempting to grapple with interoperability of data to derive meaningful conclusions about health and other outcomes.Hazard ratios or risk ratios were generated using a series of Cox regression models to quantify associations between outof-home care placement and long-term outcomes. To examine specific outcomes, the authors 2 combined national population registers, patient prescription drug registers, causes of death, and crime registers. The authors relied on full-sibling comparisons in which 1 sibling was exposed to out-of-home care and the other remained in the home. Outcomes were categorized into 4 domains: psychiatric disorders (any inpatient episode or outpatient visit associated with a diagnosis for "severe mental illness"); low socioeconomic status, defined as receiving welfare benefits, long-term unemployment support, and/or a disability pension and/or having a low education level (ie, not achieving secondary school qualifications); experiencing unintentional or intentional injury (any inpatient episode, outpatient visit, or death associated with any diagnosis Related articleOpinion EDITORIAL jamapediatrics.com (Reprinted) JAMA