Abstract: In this talk we present a pilot study evaluating strategies for the dissemination of pro-health messages with regards to the United States opioid epidemic. A ChatterGrabber instance was deployed to track tweets relating to an outbreak of fentanyl and fentanyl derivative contaminated percocets on the US black market. 5,406 original tweets were collected from Virginia, Georgia, and adjacent regions from June 14th, 2017 to February 6th, 2018.
From these tweets, 847 were found to be retweeted a total of 3,574 times. Tweets were manually labelled for the presence of messages of general avoidance, health and legal consequences of abuse, public policy interventions, contamination, and use witnessed. Tweeters’ profiles were classified as individuals, social pages, law enforcement pages, organizational pages, or government agency pages.
The efficacy of message transmission was evaluated with regards to message type and source to seek optimal timing, sources, intervals, and keyword contents to promote health-positive message dissemination via retweets. Continuing research will expand the geography to include the entire United States and seek to develop predictive models of views as a function of tweet contents and user profile types.