Progression of chronic disease is often manifested by repeated occurrences of disease-related events over time. Delineating the heterogeneity in the risk of such recurrent events can provide valuable scientific insight for guiding customized disease management. We propose a new sensible measure of individual risk of recurrent events and present a dynamic modelling framework thereof, which accounts for both observed covariates and unobservable frailty. The proposed modelling requires no distributional specification of the unobservable frailty, while permitting exploration of the dynamic effects of the observed covariates. We develop estimation and inference procedures for the proposed model through a novel adaptation of the principle of conditional score. The asymptotic properties of the proposed estimator, including the uniform consistency and weak convergence, are established. Extensive simulation studies demonstrate satisfactory finite-sample performance of the proposed method. We illustrate the practical utility of the new method via an application to a diabetes clinical trial that explores the risk patterns of hypoglycemia in type 2 diabetes patients.
Publication:
Biometrika. Available online 19 Nov 2020
Authors:
Huijuan Ma
School of Statistics and Academy of Statistics and Interdisciplinary Sciences, East China Normal University, Shanghai 200062, China
Email: hjma@fem.ecnu.edu.cn
Limin Peng
Department of Biostatistics and Bioinformatics, Emory University, Atlanta, Georgia 30322, U.S.A.
Chiung-Yu Huang
Department of Epidemiology and Biostatistics, University of California at San Francisco, San Francisco, California 94158, U.S.A.
Haoda Fu
Eli Lilly and Company, Lilly Corporate Center, Indianapolis, Indiana 46285, U.S.A.