Characterization of Anonymous Physician Perspectives on COVID-19 Using Social Media Data

Area of Science:

• Public Health
• Social Media Analysis
• Healthcare Communication

Background:

• Physicians' perspectives on COVID-19 are crucial for pandemic response and identifying healthcare system gaps.
• Understanding healthcare workers' (HCW) sentiments and topics offers valuable insights into pandemic challenges.
• Social media platforms provide a unique avenue for anonymous expression of professional opinions.

Purpose of the Study:

• To analyze anonymous physician perspectives on the COVID-19 pandemic using Twitter data.
• To identify key topics and sentiments expressed by physicians regarding the pandemic response.
• To compare physician discourse with general COVID-19 related social media conversations.

Main Methods:

• Extracted and analyzed English-language tweets from an anonymous physician user group on Twitter.
• Employed bigram and trigram frequency analysis to identify prevalent topics.
• Utilized sentiment analysis methods and compared them against a larger, general COVID-19 Twitter dataset and manual analysis.

Main Results:

• Significant differences were observed in topical phrases between the anonymous physician dataset and general COVID-19 discourse.
• No statistically significant difference in sentiment was found between the two datasets; both showed a slightly positive trend.
• Current automated sentiment analysis methods require improvement to accurately capture sentiments within anonymous physician data.

Conclusions:

• Anonymous physician social media data offers unique and valuable insights into the COVID-19 pandemic.
• Topical analysis reveals distinct concerns and discussions among physicians compared to the general public.
• Enhancements in sentiment analysis are necessary for reliable interpretation of healthcare professional social media data.