Individual risk versus population incidence: a case example with congenital heart defects.

BACKGROUND AND OBJECTIVE: Many studies have shown that individuals in low socioeconomic status neighborhoods are likelier to have a severe congenital heart defect (CHD) compared to those in higher socioeconomic status neighborhoods. These studies conclude that this finding should be leveraged for population health policy. However, individual-level associations may not extend to populations. Thus, our objective was to estimate associations between county-level median home value and county-level incidence of live-born, severe CHDs. METHODS: Ecological analysis of statewide panel data at the county-year level. All neonates with a severe CHD born in Illinois, July 2013 to December 2021, were identified from the Adverse Pregnancy Outcomes Reporting System (Illinois Department of Public Health). The outcome was county-level incidence of liveborn, severe CHD. The exposure was county-level median home value. Associations were estimated with multivariable Poisson models, stratified by census region (west, south/central, and northeast regions of the state), a modifier that was identified during analysis. RESULTS: All 102 counties in Illinois were included. After regional stratification, low county-level median home value was associated with higher severe CHD incidence, but only in the western region of the state (IRR for lowest versus highest home value quartile: 1.96, 95%CI 1.14-3.38). There was no association between census tract-level median home value and individual probability of severe CHD in the western region. CONCLUSION: Differences exist between county-level and individual-level associations of CHD and socioeconomic status. Ecologic and individual-level analyses have distinct, but complementary, roles in informing evidence-based policies and public health initiatives.