Internet Searches for Affect-Related Terms: An Indicator of Subjective Well-Being and Predictor of Health Outcomes across US States and Metro Areas

The mediating effect of media usage on the relationship between anxiety/fear and physician-patient trust during the COVID-19 pandemic

OBJECTIVE

Our study explored whether and how media usage can mediate the path from anxiety and fear to physician-patient trust.

DESIGN

Study 1 was a population-based, longitudinal study using nationally representative data from 29 provinces in mainland China. The baseline sample (N = 3233) was obtained from February 1 to 9, 2020. Follow-up (N = 1380) took place during March 17 to 24, 2020. Study 2 was a machine learning-based sentiment analysis in which data were captured from Sina Weibo, a Chinese microblogging website, among the most popular official, unofficial, and health-related media accounts. The screened blogs from November to December 2019 and February to March 2020 were scored by Google APIs for positivity and magnitude.

MAIN OUTCOME MEASURES

Physician-patient trust.

RESULTS

Study 1 showed fear and anxiety affected changes in physician-patient trust through media usage, the indirect effect of which was 0.14 (0.03) and the 95% CI was [0.08, 0.19]. Study 2 indicated a more positive image of physicians after the outbreak compared to before [F (2, 3537) = 3.646, p = 0.026, partial η²=0.002].

CONCLUSION

The negative impact of anxiety and fear on physician-patient trust was mediated by media use, which can be explained by the more positive media image during the pandemic.

COVID-19 Increases Online Searches for Emotional and Health-Related Terms

Background

The COVID‐19 pandemic has powerfully shaped people’s lives. The current work investigated the emotional and behavioral reactions people experience in response to COVID‐19 through their internet searches. We hypothesised that when the prevalence rates of COVID‐19 increase, people would experience more fear, which in turn would predict more searches for protective behaviors, health‐related knowledge, and panic buying.

Methods

Prevalence rates of COVID‐19 in the United States, the United Kingdom, Canada, and Australia were used as predictors. Fear‐related emotions, protective behaviors, seeking health‐related knowledge, and panic buying were measured using internet search volumes in Google Trends.

Results

We found that increased prevalence rates of COVID‐19 were associated with more searches for protective behaviors, health knowledge, and panic buying. This pattern was consistent across four countries, the United States, the United Kingdom, Canada, and Australia. Fear‐related emotions explained the associations between COVID‐19 and the content of their internet searches.

Conclusions

Findings suggest that exposure to COVID‐19 prevalence and fear‐related emotions may motivate people to search for relevant health‐related information so as to protect themselves from the pandemic.

Twitter-based measures of neighborhood sentiment as predictors of residential population health

Several studies have recently applied sentiment-based lexicons to Twitter to gauge local sentiment to understand health behaviors and outcomes for local areas. While this research has demonstrated the vast potential of this approach, lingering questions remain regarding the validity of Twitter mining and surveillance in local health research. First, how well does this approach predict health outcomes at very local scales, such as neighborhoods? Second, how robust are the findings garnered from sentiment signals when accounting for spatial effects? To evaluate these questions, we link 2,076,025 tweets from 66,219 distinct users in the city of San Diego over the period of 2014-12-06 to 2017-05-24 to the 500 Cities Project data and 2010-2014 American Community Survey data. We determine how well sentiment predicts self-rated mental health, sleep quality, and heart disease at a census tract level, controlling for neighborhood characteristics and spatial autocorrelation. We find that sentiment is related to some outcomes on its own, but these relationships are not present when controlling for other neighborhood factors. Evaluating our encoding strategy more closely, we discuss the limitations of existing measures of neighborhood sentiment, calling for more attention to how race/ethnicity and socio-economic status play into inferences drawn from such measures.