B1736
Title: Semiparametric Bayesian two-stage meta-analysis for association between ambient temperature and new cases of COVID-19
Authors: Dongu Han - Korea University (Korea, South) [presenting]
Kiljae Lee - Ohio state University (United States)
Yeonseung Chung - Korea Advanced Institute of Science and Technology (Korea, South)
Taeryon Choi - Korea University (Korea, South)
Abstract: In environmental epidemiological studies, two-stage meta-analysis has been a popular tool to investigate a short-term association between environmental exposure and a health response by analyzing daily time-series data collected from multiple locations. We propose a novel Bayesian approach for a two-stage meta-analysis by innovating each of the existing first and second-stage models in Bayesian frameworks. Specifically, for the first stage model, we propose a new Bayesian distributed lag nonlinear model which accommodates three kinds of nonlinearities. For the second stage model, we propose a matrix-variate Dirichlet process mixture multivariate meta-regression that is robust when the assumptions of existing linear models are violated. The proposed second stage model also allows for identifying subgroups of locations through model-based clustering. We validate the methodologies through simulation studies and apply them to study a short-term association between ambient temperature and new cases of COVID-19 in South Korea and the United States.
