Social media has become an important platform for promoting research within the interventional cardiology community. However, the popularity of a social media account does not necessarily reflect the academic reputation of its owner, a paper published in JACC: Cardiovascular Interventions has concluded.
The research paper, authored by Paolo D’Arrigo, Antonio Greco (both University of Catania and G Rodolico-San Marco, Catania, Italy) and colleagues, looked at the Twitter metrics of faculty members and participants from the Transcatheter Cardiovascular Therapeutics (TCT) meeting. The primary outcome—popularity—was measured based upon an account’s higher followers’ number (HFN), defined as those belonging to the top quartile of followers’ numbers on Twitter—those with >736 followers.
Researchers examined three predictors of HFN, which included: 1) the Hirsch index (h-index), described as a marker of academic proficiency combining the numbers of published papers and related citations; 2) the activity index (a-index), the total number of tweets indexed by weeks from Twitter subscription; and 3) the Kardashian index (k-index), a measure of the potential discrepancy between social media reputation and publication record.
At one year, academic and social metrics were retracked and variations from baseline were analysed. On the basis of the increase in followers’ number from baseline to one year, fast-growing accounts were identified as those belonging to the top quartile—those with >344 additional followers.
D’Arrigo and Greco report that of the 1,423 TCT participants, they were able to locate active Twitter accounts for 382 individuals (26.8%, 89.3% male, 10.7% female). More frequently these accounts had a US affiliation. Academically, they had higher median values of h-index, publications, citations, and total amount of references used in their papers, and they were more likely to belong to a large cooperation network defined as a total of at least 150 co-authors over their academic career, the researchers found.
After univariate and multivariable analyses of the academic and social metrics of the Twitter users, the authors of the paper identified six independent predictors of HFN, which included the volume of posts and affiliation. The authors report that a- and k-indexes, but not the h-index, were independently associated with HFN. In particular, they report that accounts generating a stream of valuable content are more likely to be followed. “Some accounts may take advantage of celebrity to amplify their success, in a kind of cycle,” they suggest.
Follow-up at one year showed a faster increase in followers for accounts with a higher baseline number of tweets and higher a-index, whereas the baseline k-index was no longer associated with a significant increase in popularity, denoting non-lasting effects. “In aggregate, our results suggest that a prediction rule for durable popularity on Twitter is to be active and generate valuable contents rather than relying on individual academic or social reputation,” the authors note.
Citing some limitations of the research, they note that identifying individual accounts can be challenging due to the use of aliases and avatars online, and also that some accounts can be made private by their owners. The study also only applied to the TCT faculty, and may not be representative of the interventional cardiology community entirely.
In conclusion, D’Arrigo and Greco surmise that the popularity of social media accounts among interventional cardiologists is independently associated with their level of social media activity but did not seem to be associated with the academic reputation of their owners.
