Title: Understanding the dynamic impact of COVID-19 through competing risk modelling with bivariate varying coefficients
Authors: Wenbo Wu - New York University Grossman School of Medicine (United States) [presenting]
Abstract: The COVID-19 pandemic has exerted a profound impact on patients with kidney failure. Motivated by request by the U.S. Centers for Medicare \& Medicaid Services, the analysis of their post-discharge hospital readmissions and deaths in 2020 revealed that the COVID-19 effect has varied significantly with post-discharge time and time since the pandemic onset. However, the complex dynamics of the impact trajectories cannot be characterized by existing varying coefficient models. To address this issue, a bivariate varying coefficient model is proposed for competing risks within a cause-specific hazard framework, where tensor-product B-splines are used to estimate the surface of the COVID-19 effect. An efficient proximal Newton algorithm is developed to facilitate fitting the new model to the massive Medicare data for dialysis patients. Difference-based anisotropic penalization is introduced to mitigate model overfitting and the wiggliness of the estimated trajectories; various cross-validation methods are considered in determining optimal tuning parameters. Hypothesis testing procedures are designed to examine whether the COVID-19 effect varies significantly with postdischarge time and the time since the pandemic onset, either jointly or separately. Simulation experiments and applications for Medicare dialysis patients demonstrate the proposed methods' estimation accuracy, controlled type I error rate, sufficient statistical power, and real-world performance.