This long view of the Lancet Infectious Diseases (TLID) Twitter account looks back over 2 years to explore the change in tweeting and retweeting, and the reception of these posts, before and during the COVID-19 pandemic. It uses methods explained in detail in a recent blog. Typically access to social media data via the Twitter API are only available for 7-10 days after posting. There are two workarounds that provide access to data over a longer period. First @TheLancetInfDis tweets and retweets were extracted in a NodeXL user search (up to 3,200 tweets and retweets without time limit), reconstructing the data in a NodeXL tweet ID extract (again no time limit). Retweets may accrue in the days after a post, so retweet data were refreshed using TAGS (using tweet ID) a week later. There were 3,196 tweets extracted successfully from 372 accounts. In total these 3,196 tweets received 436,977 retweets, 13,202 of which were received by TLID, the rest received by posts that TLID retweeted. For the most popular ("viral") tweets it is not possible to tell to what extent TLID contributed to this snowballing dissemination.  The top 100 posts that the Journal retweeted are listed in another Wakelet summary. Replies (5% of all posts) were excluded because these are displayed differently to original tweets on Twitter.
The top 100 tweets posted by TLID are listed chronologically below. In total they received 6,145 retweets. The top tweeters and hashtags are also shown. The hashtag word clouds, based on @TheLancetInfDis tweets are interactive so you can click the hashtags to view the individual tweets posted by the Journal.  Over a period of 630 days prior to the pandemic 263 hashtags were used. Over the 169 days of the pandemic period studied 51 hashtags were used, 10 of which were directly COVID-19 related. Looking at each hashtag used during the pandemic, two thirds (262/393) were COVID-19 related. There was therefore a considerable narrowing of topics covered during the pandemic.
Data plotted over time demonstrates a step change in the level of retweeting during the pandemic, peaking in March. The Journal's first COVID-19 related retweet was 2 Jan 2020, with their first original tweet on 9 January. Between 11 April 2018 and this initial pandemic coverage the Journal received an average of 8 retweets per tweet posted. From the start of the pandemic coverage the journal received an average of 16 retweets per tweet posted. The tweets that the Journal tweeted had a more marked increase: receiving an average of 39 retweets per post before the pandemic and 594 retweets per post during the pandemic.

Items