Invited Commentary: Estimation and Bounds Under Data Fusion

Am J Epidemiol. 2021 Jul 7:kwab194. doi: 10.1093/aje/kwab194. Online ahead of print.ABSTRACTOgburn et al. (Am J Epidemiol. 2021;000(0):000-000) raise a cautionary tale for epidemiological data fusion: bias may occur if a variable completely missing in the primary dataset is imputed according to a regression model estimated from an auxiliary dataset. However, in some specific settings, solution may exist. Focusing on a linear outcome regression model with a missing covariate, we show that the bias can be eliminated if the underlying imputation model for the missing covariate is nonlinear in the common variables measured in both datasets. Otherwise, we describe two alternative approaches existing in the data fusion literature that could partially resolve this issue: one estimates the outcome model by leveraging an additional validation dataset containing joint observations of the outcome and the missing covariate, and the other offers informative bounds for the outcome regression coefficients without using validation data. We justify these three methods on a linear outcome model and briefly discuss their extension to general settings.PMID:34240101 | DOI:10.1093/aje/kwab194

Source: Am J Epidemiol - July 9, 2021 Category: Epidemiology Authors: Wang Miao Wei Li Wenjie Hu Ruoyu Wang Zhi Geng Source Type: research