The Twitter factor: How does Twitter impact #Stroke journals and citation rates?. Issue 5 (June 2023)

• Record Type: Journal Article
• Authors: Sousa, João André; Alves, Inês Alexandra; Donato, Helena; Sargento-Freitas, João
• Is Part Of: International journal of stroke. Volume 18:Issue 5(2023)
• Page Start: 586
• Page End: 589
• Publication Date: 2023-06
• Subjects: Social media -- stroke -- Twitter -- altmetrics -- citations -- impact factor; 616.8005
• DOI: 10.1177/17474930221136704
• Languages: English
• ISSNs: 1747-4930

Abstract

Background: Twitter is a social media platform popularly used by health practitioners, a trend that has been followed by medical journals. The impact of Twitter in bibliometrics of stroke-related literature is yet to be determined. Aims: We aimed to qualitatively assess the usage of Twitter by stroke journals and study the relationship between Twitter activity and citation rates of stroke articles. Methods: We used Journal Citation Reports to identify stroke journals. We collected the 2021 Impact Factor (IF) and the top 50 articles contributing to each journal IF. Relevant metrics were collected through Twitonomy, Altmetric, and Web of Science. The association between Twitter activity and citation rates was tested by a negative binomial regression model adjusted to journal's IF. A bivariate correlation and a log-linear regression model adjusted to journal's IF tested the relationship between number of tweets, tweeters, and the number of citations. Results: We collected 450 articles across nine stroke-dedicated journals, five of which had a Twitter account. Only 95 (21%) articles had no Twitter mentions. The median number of citations in articles with versus without Twitter activity was 19 (10–39) versus 11(7–17) ( P < 0.001). Twitter activity was associated with higher citation rates controlling for the IF (odds ratio (OR): 2.7, 95% confidence interval (CI) 2.12–3.38, P < 0.001). We found number of tweets to be predicted by the number of citations controlling for the IF (B = 0.33, 95% CI 0.29–0.40, β = 0.54, P < 0.001). Conclusions: Tweeted stroke articles tend to have higher citation rates which can be predicted by the number of tweets.