Cox regression with linked data

Abstract

Record linkage is increasingly used, especially in medical studies, to combine data from different databases that refer to the same entities. The linked data can bring analysts novel and valuable knowledge that is impossible to obtain from a single database. However, linkage errors are usually unavoidable, regardless of record linkage methods, and ignoring these errors may lead to biased estimates. While different methods have been developed to deal with the linkage errors in the generalized linear model, there is not much interest on Cox regression model, although this is one of the most important statistical models in clinical and epidemiological research. In this work, we propose an adjusted estimating equation for secondary Cox regression analysis, where linked data have been prepared by a third‐party operator, and no information on matching variables is available to the analyst. Through a Monte Carlo simulation study, the proposed method is shown to lead to substantial bias reductions in the estimation of the parameters of the Cox model caused by false links. An asymptotically unbiased variance estimator for the adjusted estimators of Cox regression coefficients is also proposed. Finally, the proposed method is applied to a linked database from the Brest stroke registry in France.

Categorical linkage‐data analysis

“…To overcome these problems, Zhang and Tuoto 17 propose a pseudo OLS method for linkage-data linear regression, and Vo et al 18 develop similarly an adjusted estimating equation approach to Cox regression for survival analysis.…”

“…In this paper we shall address the gap by proposing methods for categorical linkage-data analysis from the secondary users' perspective, which include logistic regression, analysis of contingency table or log-linear modelling of such data. In the same spirit as Zhang and Tuoto 17 and Vo et al, 18 our methods can explicitly accommodate the complications of heterogeneous linkage errors, incomplete match space, and restrictions on the joint-data model induced by the linkage data structure.…”