Estimation of COVID-19 spread curves integrating global data and borrowing information

Abstract

Currently, novel coronavirus disease 2019 (COVID-19) is a big threat to global health. The rapid spread of the virus has created pandemic, and countries all over the world are struggling with a surge in COVID-19 infected cases. There are no drugs or other therapeutics approved by the US Food and Drug Administration to prevent or treat COVID-19: information on the disease is very limited and scattered even if it exists. This motivates the use of data integration, combining data from diverse sources and eliciting useful information with a unified view of them. In this paper, we propose a Bayesian hierarchical model that integrates global data for real-time prediction of infection trajectory for multiple countries. Because the proposed model takes advantage of borrowing information across multiple countries, it outperforms an existing individual country-based model. As fully Bayesian way has been adopted, the model provides a powerful predictive tool endowed with uncertainty quantification. Additionally, a joint variable selection technique has been integrated into the proposed modeling scheme, which aimed to identify possible country-level risk factors for severe disease due to COVID-19.

A simple model for the total number of SARS-CoV-2 infections on a national level

“…While predictions about the future course of an epidemic, especially one as novel as COVID-19 are difficult under the best of circumstances, the severity of the pandemic has resulted in an unprecedented amount of epidemiological data being produced with daily frequency making it clear that the spread of cases is not well-modelled by simple exponential growth equations [ 17 , 23 , 45 , 46 ]. This is not unprecedented as sub-exponential growth has been noted previously in a number of disease outbreaks [ 35 , 47 – 49 ].…”

Analysis of the SARS-CoV-2 epidemic in Italy: The role of local and interventional factors in the control of the epidemic

“…Generalized logistic curves (Richards model) are used to study the infection trajectories [ 10 ]. In this study, we adopted a re-parameterization of the Gompertz growth equation [ 11 ] as regression model to describe the cumulative number of infected subjects per day and to define epidemic parameters in Italy and in the Italian Regions.…”

NLSIG-COVID19Lab: A modern logistic-growth tool (nlogistic-sigmoid) for descriptively modelling the dynamics of the COVID-19 pandemic process

“…Epidemiological models such as the SEIRD variants (Lee et al, 2020;Okabe & Shudo, 2020) are just another form of representing sigmoidal growth (Xs et al, 2012). However, it has been noted (Christopoulos, 2020) that the SEIRD-variant models yield largely exaggerated forecasts.…”

“…The second is that during optimization, estimation of the logistic hyperparameters for the individual logistic-sigmoids that make the multiple logistic-sigmoid sum are computed separately, instead of as a unified function. The effect of this is that as the number of logistic-sigmoids considered in the sum increases, regression analysis becomes more cumbersome and complicated, as can be observed in a number of works (Batista, 2020;Chowell et al, 2019;Hsieh & Cheng, 2006;Lee et al, 2020;Taylor & Letham, 2018;Wu et al, 2020).…”