83.10 Characterizing #PhysicianBurnout through Social Media Posts, Responses, Traffic, and Influencers

C. Hwang1,2, T. Bellomo1,2, M. Byrnes2, M. A. Corriere1,2  1University Of Michigan,Vascular Surgery,Ann Arbor, MI, USA 2University Of Michigan,Center For Health Outcomes And Policy,Ann Arbor, MI, USA

Introduction:
Physician burnout is highly prevalent among surgeons. Negative impacts include depression, divorce, substance abuse, and attrition.  Survey and focus group methods have been used to study burnout, but potential non-response bias among affected doctors is a major potential limitation of these approaches.  Social media is a potentially rich source of information related to provider burnout because users can post while experiencing stress. We analyzed posts with hashtags related to burnout to characterize themes and influential users.

Methods:
A healthcare social media analytic platform was used to identify burnout-related hashtags. Posts specifically identifying physician burnout over a 90-day period (May26 -August 24, 2018) were analyzed. User influence was assessed using an impact factor algorithm accounting for healthcare relevance, conversation partners, and stakeholder roles identified through metadata.  Users, content volume, trends, and themes were evaluated using mixed methods. 

Results:
23 hashtags linked to burnout were identified. Hashtag names referenced burnout directly (#Physicianburnout), sarcastically (#joyofmedicine), through narratives (#ShareASToryInOneTweet), through impacts on patients and coworkers (#Thosewecarry), and countermeasures (#ProviderWellness).  #Physicianburnout was associated with over 5300 unique tweets, 3200 retweets, >20 million impressions, 3300 shared links, and 1800 visual file shares over 90 days.  Individual doctors accounted for the largest share of activity, followed by healthcare organizations, non-physician individuals in healthcare, and non-health individuals. Individual doctors represented 24/35 (68.5%) of top influencers, including the four highest ranking influencers. Trending terms included "Maslow's hierarchy," "broken work systems," and "Telling doctors to be more resilient…".  Prevalent themes included depression, stigma, depersonalization, negative impacts on patient care, activities that protect against burnout, and the need for collective action from physicians and medical schools. Trends indicated increases in social media volume ranging from 25%-39% from the preceding period.

Conclusion:
Social media provides rich and dynamic information about physician burnout. The majority of burnout-related content is posted by individual doctors who are potentially inaccessible through survey or focus group studies. Physician advocacy and support groups (including surgical societies) presently account for a limited portion of social media content related to burnout.  In addition to understanding burnout, social media represents a potential means of communicating initiatives and strategies to combat this problem.  Stakeholders invested in addressing physician burnout should leverage social media as a tool and consider partnering with influential posters.