It is to be expected that promoting a scientific publication through social media channels increases its online visibility. But such an initiative can also expand the publication’s reach within the scientific community and eventually lead to higher citation rates by peers. The European Society of Cardiology (ESC) performed a study to assess if there was a correlation between social media presence and article citation rates.
The ESC Journals Study randomized 695 papers published in the ESC Journal Family between March 2018 and May 2019 in intervention or control arms. The articles in the intervention arm were posted onto the @ESC_journals feed for promotion on Twitter, whereas the ones in the control arm received no active tweeting from the ESC channels. For the duration of the study, one or two articles were posted per day on Twitter. Tweets were written in English and contained at least one key image, if applicable, and a free link to the full-text version of the article. The primary endpoint of the study was the article’s citation rate. The secondary endpoint was the number of users that had tweeted the article and if there was an increase in the Altmetric Attention Score (AAS) – a metric that gauges the impact of a research article beyond typical metrics such as impact factor or citation counts.
Tweet or not to tweet
The final analysis involved 694 articles, as one had to be excluded due to retraction. The median number from publishing on Twitter to the day that the number of citations was counted was 994 days (2.7 years). The results showed that papers in the Twitter arm had a median number of citations of 15, an AAS of 24 and a median number of Twitter users of 41. In contrast, the articles from the control group had a median number of citations of 14, an AAS of 5 and a median number of tweeting users of 6.
In conclusion, Twitter promotion was linked with a 1.12 higher citation rate (Poisson regression, 95% confidence interval: 1.08-1.15). AAS and the number of users tweeting were also positive predictors of the number of citations (Poisson regression coefficients of 0.27, 95% CI: 0.26–0.28, P, 0.001 and 0.23, 95% CI: 0.22– 0.24, P, 0.001, respectively).
To read the full article, click here.