Your browser doesn't support javascript.
How Health Care Workers Wield Influence Through Twitter Hashtags: Retrospective Cross-sectional Study of the Gun Violence and COVID-19 Public Health Crises.
Ojo, Ayotomiwa; Guntuku, Sharath Chandra; Zheng, Margaret; Beidas, Rinad S; Ranney, Megan L.
  • Ojo A; Harvard Medical School, Boston, MA, United States.
  • Guntuku SC; Penn Medicine Center for Digital Health, Philadelphia, PA, United States.
  • Zheng M; Department of Computer and Information Science, University of Pennsylvania, Philadelphia, PA, United States.
  • Beidas RS; Leonard Davis Institute of Health Economics, University of Pennsylvania, Philadelphia, PA, United States.
  • Ranney ML; Leonard Davis Institute of Health Economics, University of Pennsylvania, Philadelphia, PA, United States.
JMIR Public Health Surveill ; 7(1): e24562, 2021 01 06.
Article in English | MEDLINE | ID: covidwho-1011352
ABSTRACT

BACKGROUND:

Twitter has emerged as a novel way for physicians to share ideas and advocate for policy change. #ThisIsOurLane (firearm injury) and #GetUsPPE (COVID-19) are examples of nationwide health care-led Twitter campaigns that went viral. Health care-initiated Twitter hashtags regarding major public health topics have gained national attention, but their content has not been systematically examined.

OBJECTIVE:

We hypothesized that Twitter discourse on two epidemics (firearm injury and COVID-19) would differ between tweets with health care-initiated hashtags (#ThisIsOurLane and #GetUsPPE) versus those with non-health care-initiated hashtags (#GunViolence and #COVID19).

METHODS:

Using natural language processing, we compared content, affect, and authorship of a random 1% of tweets using #ThisIsOurLane (Nov 2018-Oct 2019) and #GetUsPPE (March-May 2020), compared to #GunViolence and #COVID19 tweets, respectively. We extracted the relative frequency of single words and phrases and created two sets of features (1) an open-vocabulary feature set to create 50 data-driven-determined word clusters to evaluate the content of tweets; and (2) a closed-vocabulary feature for psycholinguistic categorization among case and comparator tweets. In accordance with conventional linguistic analysis, we used a P<.001, after adjusting for multiple comparisons using the Bonferroni correction, to identify potentially meaningful correlations between language features and outcomes.

RESULTS:

In total, 67% (n=4828) of #ThisIsOurLane tweets and 36.6% (n=7907) of #GetUsPPE tweets were authored by health care professionals, compared to 16% (n=1152) of #GunViolence and 9.8% (n=2117) of #COVID19 tweets. Tweets using #ThisIsOurLane and #GetUsPPE were more likely to contain health care-specific language; more language denoting positive emotions, affiliation, and group identity; and more action-oriented content compared to tweets with #GunViolence or #COVID19, respectively.

CONCLUSIONS:

Tweets with health care-led hashtags expressed more positivity and more action-oriented language than the comparison hashtags. As social media is increasingly used for news discourse, public education, and grassroots organizing, the public health community can take advantage of social media's broad reach to amplify truthful, actionable messages around public health issues.
Subject(s)
Keywords

Full text: Available Collection: International databases Database: MEDLINE Main subject: Health Personnel / Social Media / Gun Violence Type of study: Experimental Studies / Observational study / Qualitative research / Randomized controlled trials Topics: Long Covid Limits: Humans Language: English Journal: JMIR Public Health Surveill Year: 2021 Document Type: Article Affiliation country: 24562

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: International databases Database: MEDLINE Main subject: Health Personnel / Social Media / Gun Violence Type of study: Experimental Studies / Observational study / Qualitative research / Randomized controlled trials Topics: Long Covid Limits: Humans Language: English Journal: JMIR Public Health Surveill Year: 2021 Document Type: Article Affiliation country: 24562