Trend Analysis and Intervention Effect Starting Point Detection of COVID-19 Epidemics Using Recalibrated Time Series Models - Worldwide, 2020.

Shu Li,Chen Chen,Xiaolin Xu,Xifeng Wu,Qinchuan Wang,Hao Lei,Kejia Hu,Junlin Jia,Yuanqing Ye,Sicong Wang,Sisi Wang,Shuyin Cao,Changzheng Yuan

doi:10.46234/ccdcw2021.112

Abstract

Objective: This study aimed to identify a model for short-term coronavirus disease 2019 (COVID-19) trend prediction and intervention evaluation. Methods: We compared the autoregressive integrated moving average (ARIMA) model and Holt exponential smoothing (Holt) model on predicting the number of cumulative COVID-19 cases in China. Based on the mean absolute percentage error (MAPE) value, the optimal model was selected and further tested using data from the United States, Italy and Republic of Korea. The intervention effect starting time points and abnormal trend changes were detected by observing the pattern of differences between the predicted and real trends. Results: The recalibrated ARIMA model with a 5-day prediction time span has the best model performance with MAPEs ranged between 2% and 5%. The intervention effects started to show on February 7 in the mainland of China, March 5 in Republic of Korea and April 27 in Italy, but have not been detected in the US as of May 19. Temporary abnormal trends were detected in Korea and Italy, but the overall epidemic trends were stable since the effect starting points. Conclusion: The recalibrated ARIMA model can detect the intervention effects starting points and abnormal trend changes; thus to provide valuable information support for epidemic trend analysis and intervention evaluation.

Highlights

Coronavirus disease 2019 (COVID-19) is still a worldwide threat [1]
The results indicated that the recalibrated autoregressive integrated moving average (ARIMA) model was suitable for shortterm prediction of COVID-19 trends and could detect the intervention effect starting time points
Given the ongoing COVID-19 pandemic, this study could serve as a reference to initiate more adaptable and practice-based epidemic trend analysis tools that can benefit pandemic responders in different countries

Summary

Introduction

Previous studies have developed a variety of mathematical models to simulate and predict the disease transmission pattern recently [2,3], which mainly focus on macro-level and long-term prediction over the entire course of an pandemic. Those models may miss real-time trend changes and shortertime disturbances. The effects of interventions started to show on February 7, 2020 in the mainland of China, March 5, 2020 in Republic of Korea, April 27, 2020 in Italy, and May 19, 2020 in the United States This model can provide valuable information to support evaluating interventions, resource allocation, decision-making, and situation monitoring. Given the ongoing COVID-19 pandemic, this study could serve as a reference to initiate more adaptable and practice-based epidemic trend analysis tools that can benefit pandemic responders in different countries

Methods

Results

Conclusion