SP2.2.3How to make the perfect tweet: exploring factors which affect retweetability of surgical research

Aims Twitter is a powerful platform for sharing information. Retweets allow users to reach a wider audience quickly. Our aim was to determine what tweet and twitter-account related factors affect the “retweetability” of surgical publications. Methods All papers published (in print) in Annals of Surgery, BJS and JAMASurgery in 2019 were selected. We identified each paper’s first appearance on Twitter and analysed characteristics of the posting twitter account and the tweet itself. The association between such characteristics and the number of retweets was investigated. Results 413 papers were selected. Median number of retweets was 3 (range, 0-121). The first tweet created about a paper was usually from a reader (42%) with only a minority created by the journal (17%) or the authors (20%). Use of visual abstracts (22/413, p<0.001), pictures (145/413, p<0.001), hashtags (149/413, p<0.001), mentions (183/413, p<0.001), number of followers (p < 0.001), age of twitter account (p = 0.003) and authors as creators of the first tweet (p < 0.001) were associated with higher retweets. When the journal created the first tweet, retweets were also increased (P < 0.001). Posting the title of the paper alone attracted fewer retweets (p < 0.001). Multivariate analysis confirmed visual abstracts (p < 0.001), hashtags (p=p.008), pictures (p < 0.001) and authors as creators of the first tweet (p < 0.001) were predictive of increased retweetability. Conclusions Authors and journals should create the first presence on social media to improve engagement with their papers. Such tweets should use visual abstracts or pictures, relevant hashtags and avoid solely posting the paper title to make their research more widely shared.