Early Epidemic Detection of Ebola through Tweet Analysis: Nigeria as a case study

Background: Surveillance through electronic mediums like the Internet, provide tremendous opportunities for public health practice.  The current Ebola outbreak in Nigeria has provided a unique opportunity to examine the use of Twitter, a powerful Social Network Site (SNS) tool during a public health crisis.  We provide a snapshot of Ebola-related tweets to monitor trends of information spread and examine early epidemic detection.

 Methods:  During Ebola outbreak from July 24^th to August 1st, 2014, tweets were collected regarding Ebola.  Keywords of related data included # #EbolaVirus, #EbolaFacts.  Data elements collected for tweets included contents, time stamps, latitude and longitude codes.  Twitter communication was examined through natural language processing and time series analysis.

 Results: Tweets steadily increased (six-fold posted and twenty-fold disseminated) from July 24^th until July 31^st.  The first probable case announced by the Nigerian Ministry of Health was July 27^th .  Tweets increased three to seven days prior to the official CDC announcement on July 31^st.  Twitter users discussed the first case in tweets such as“#EbolaVirus 1^st case discovered Lagos, pls spread the word”.  Ebola news was first tweeted on July 24th and reached 1,196,793 with 120,574,549 reached on July 30th, a hundred-fold increase.

 Conclusion:  Our results demonstrates the effective use of social media in public health outbreak surveillance.  We have also demonstrated that although Twitter adoption and use in resource-limited settings is lagging, an observed increase in the frequency of Ebola-related tweets shows its usefulness.  SNS tools such as Twitter can support surveillance efforts in resource limited settings.