Association between public demand and sentiment and the newly confirmed cases of COVID-19 from 2020 to 2023: A time series and lag-correlation analysis

Abstract

Introduction

Emerging infectious diseases continued to be a great threat to the whole world. Online public opinion, which reflected public demands and sentiments, was likely to contribute to the surveillance and prevention of the epidemics. However, the associations between the public demand and sentiment analyzed from online public opinion and epidemics remain unknown.

Material and Methods

We obtained text data from “People's Daily Online-Message Board for Leaders”, representing for online public opinion in this study. The period was divided into three stages, namely Containment stage (I), Ongoing Prevention and Control stage (II), and Steady Transition stage (III). We defined and explained the indicators of public demand and sentiment. Through natural language processing (NLP) technology, we conducted sentiment analysis on the message texts and described the trends of public demand and sentiment using interrupted time series analysis. Subsequently, we assessed the lag-correlations between the newly confirmed cases of COVID-19 and public demand/sentiment by stages.

Results

Public demand showed a L-shape trend in general, and public sentiment presented a U-shaped trend. Public demand and epidemic showed mainly positive lag-correlations (Stage I: rs=0.42, lag=+14, P<0.001; Stage II: rs=0.46, lag=+8, P<0.001). For public sentiment, it was negatively correlated with the epidemic in Stage I (rs=-0.2, lag=+10). Lag correlation between public sentiment and the epidemic was weak in Stage II, and varied in different lags in Stage III.

Discussion

The findings suggested robust associations between 8-days-posteriority public demand and the newly confirmed cases in Ongoing Prevention and Control stage, 10-days-priority public sentiment and the cases in Containment stage. This work introduced two potential monitor indicators and lag-correlation model on the real time warning, surveillance of emerging infectious diseases.