Evaluation of a commercial database to estimate residence histories in the Los Angeles Ultrafines Study.

ABSTRACT Background Commercial databases can be used to identify participant addresses over time, but their quality and impact on environmental exposure assessment is uncertain. Objective To evaluate the performance of a commercial database to find residences and estimate environmental exposures for study participants. Methods We searched LexisNexis® for participant addresses in the Los Angeles Ultrafines Study, a prospective cohort of men and women aged 50-71 years. At enrollment (1995-1996) and follow-up (2004-2005), we evaluated attainment (address found for the corresponding time period) and match rates to survey addresses by participant characteristics. We compared geographically-referenced predictors and estimates of ultrafine particulate matter (UFP) exposure from a land use regression model using LexisNexis and survey addresses at enrollment. Results LexisNexis identified an address for 69% of participants at enrollment (N=50,320) and 95% of participants at follow-up (N=24,432). Attainment rate at enrollment modestly differed (≥5%) by age, smoking status, education, and residential mobility between surveys. The match rate at both survey periods was high (82-86%) and similar across characteristics. When using LexisNexis versus survey addresses, correlations were high for continuous values of UFP exposure and its predictors (rho=0.86-0.92). Significance Time period and population characteristics influenced the attainment of addresses from a commercial database, but accuracy and subsequent estimation of specific air pollution exposures were high in our older study population.