Skip Navigation

Publication Detail

Title: A framework for widespread replication of a highly spatially resolved childhood lead exposure risk model.

Authors: Kim, Dohyeong; Galeano, M Alicia Overstreet; Hull, Andrew; Miranda, Marie Lynn

Published In Environ Health Perspect, (2008 Dec)

Abstract: Preventive approaches to childhood lead poisoning are critical for addressing this longstanding environmental health concern. Moreover, increasing evidence of cognitive effects of blood lead levels < 10 microg/dL highlights the need for improved exposure prevention interventions.Geographic information system-based childhood lead exposure risk models, especially if executed at highly resolved spatial scales, can help identify children most at risk of lead exposure, as well as prioritize and direct housing and health-protective intervention programs. However, developing highly resolved spatial data requires labor-and time-intensive geocoding and analytical processes. In this study we evaluated the benefit of increased effort spent geocoding in terms of improved performance of lead exposure risk models.We constructed three childhood lead exposure risk models based on established methods but using different levels of geocoded data from blood lead surveillance, county tax assessors, and the 2000 U.S. Census for 18 counties in North Carolina. We used the results to predict lead exposure risk levels mapped at the individual tax parcel unit.The models performed well enough to identify high-risk areas for targeted intervention, even with a relatively low level of effort on geocoding.This study demonstrates the feasibility of widespread replication of highly spatially resolved childhood lead exposure risk models. The models guide resource-constrained local health and housing departments and community-based organizations on how best to expend their efforts in preventing and mitigating lead exposure risk in their communities.

PubMed ID: 19079729 Exiting the NIEHS site

MeSH Terms: Child; Environmental Exposure*; Humans; Lead/toxicity*; Models, Statistical*; North Carolina; Research Design; Risk Assessment*

Back
to Top