Yueyan Wang1, Ninez A Ponce1, Pan Wang1, Jean D Opsomer1, Hongjian Yu1. 1. Yueyan Wang, Ninez A. Ponce, Pan Wang, and Hongjian Yu are with the Center for Health Policy Research, University of California, Los Angeles (UCLA). Jean D. Opsomer is with Department of Statistics, Colorado State University, Fort Collins. Ninez A. Ponce is also with the Department of Health Policy and Management, Fielding School of Public Health, UCLA.
Abstract
OBJECTIVES: We propose a method to meet challenges in generating health estimates for granular geographic areas in which the survey sample size is extremely small. METHODS: Our generalized linear mixed model predicts health outcomes using both individual-level and neighborhood-level predictors. The model's feature of nonparametric smoothing function on neighborhood-level variables better captures the association between neighborhood environment and the outcome. Using 2011 to 2012 data from the California Health Interview Survey, we demonstrate an empirical application of this method to estimate the fraction of residents without health insurance for Zip Code Tabulation Areas (ZCTAs). RESULTS: Our method generated stable estimates of uninsurance for 1519 of 1765 ZCTAs (86%) in California. For some areas with great socioeconomic diversity across adjacent neighborhoods, such as Los Angeles County, the modeled uninsured estimates revealed much heterogeneity among geographically adjacent ZCTAs. CONCLUSIONS: The proposed method can increase the value of health surveys by providing modeled estimates for health data at a granular geographic level. It can account for variations in health outcomes at the neighborhood level as a result of both socioeconomic characteristics and geographic locations.
OBJECTIVES: We propose a method to meet challenges in generating health estimates for granular geographic areas in which the survey sample size is extremely small. METHODS: Our generalized linear mixed model predicts health outcomes using both individual-level and neighborhood-level predictors. The model's feature of nonparametric smoothing function on neighborhood-level variables better captures the association between neighborhood environment and the outcome. Using 2011 to 2012 data from the California Health Interview Survey, we demonstrate an empirical application of this method to estimate the fraction of residents without health insurance for Zip Code Tabulation Areas (ZCTAs). RESULTS: Our method generated stable estimates of uninsurance for 1519 of 1765 ZCTAs (86%) in California. For some areas with great socioeconomic diversity across adjacent neighborhoods, such as Los Angeles County, the modeled uninsured estimates revealed much heterogeneity among geographically adjacent ZCTAs. CONCLUSIONS: The proposed method can increase the value of health surveys by providing modeled estimates for health data at a granular geographic level. It can account for variations in health outcomes at the neighborhood level as a result of both socioeconomic characteristics and geographic locations.
Authors: Hongjian Yu; Ying-Ying Meng; Carolyn A Mendez-Luck; Mona Jhawar; Steven P Wallace Journal: Am J Public Health Date: 2007-02-28 Impact factor: 9.308
Authors: Xingyou Zhang; James B Holt; Hua Lu; Anne G Wheaton; Earl S Ford; Kurt J Greenlund; Janet B Croft Journal: Am J Epidemiol Date: 2014-03-04 Impact factor: 4.897
Authors: Wenjun Li; Jennifer L Kelsey; Zi Zhang; Stephenie C Lemon; Solomon Mezgebu; Cynthia Boddie-Willis; George W Reed Journal: Am J Public Health Date: 2009-01-15 Impact factor: 9.308