BACKGROUND

Medical societies such as the Society for Vascular Surgery (SVS) and the Society of Interventional Radiology (SIR) have been encouraging the use of social media at annual meetings by establishing unique meeting hashtags (eg. #VAM19, #SIR19ATX). These two specialties have similar number of active physicians and share procedural interests. We set forth to understand differences in Twitter activity and engagement by analyzing Twitter outputs from the vascular annual meeting (VAM) and the interventional radiology annual meeting (IRAM) in 2019.

METHODS

Tweets with "#VAM19" and "#SIR19ATX" from 30 days before and 30 days after respective meetings were collected. Proportion of distinct Twitter users relative to total number of meeting attendees, number of posts per user, number of hashtags per post, and number of engagement metrics (likes, replies, and retweets) were collected for comparison. As for the top 100 most liked tweets, specialty-related hashtags were categorized into identity, diversity, education, procedure, and medical care, and the authors of these tweets were also identified for comparison. Student's t-test* and Fisher's Exact were utilized for analysis.

RESULTS

A total of 362 and 1944 tweets were collected from the VAM and IRAM in 2019. There was no difference in proportion of active Twitter users relative to total number of meeting attendees between the VAM (7.5%) and IRAM (7.7%). Average number of posts per user from the VAM and IRAM showed no significant difference. However, tweets related to IRAM had significantly higher number of hashtags per post (2.67±1.96) than ones related to VAM (1.78±1.26) (P < 0.0001*). Additionally, these tweets on average received significantly higher number of likes (P < 0.0001*), retweets (P < 0.0001*), and replies (P < 0.0001*) than ones related to the VAM. Specialty-related hashtags from the 100 most liked tweets showed significantly greater proportion of hashtags associated with specialty identity (4% vs 28.6%, P < 0.0001**), diversity (2.9% vs 8.0%, p=0.0268**), and education (1.1% vs 9.1%, p=0.0004**) in tweets related to the IRAM whereas the proportion of hashtags associated with procedure and medical care was similar between the two meetings. Lastly, the 100 most liked tweets were authored by trainees (p=0.005*) and official societies (p=0.003*) in significantly greater proportion in IRAM whereas academic institutions/training hospitals authored in significantly greater proportion (p=0.004*) from the VAM. Contributions from attending physician users to the 100 most liked tweets were similar between the two meetings.

CONCLUSION

Analysis of Twitter activity centered around #VAM19 and #SIR19ATX indicates that there was no significant difference in proportion of Twitter users relative to meeting attendees and average number of posts per user. However, tweets with #SIR19ATX had significantly higher number of hashtags per post and had greater level of engagement than ones with #VAM19. The top 100 most liked tweets from the two meetings differed in proportion of hashtags related to specialty identity, diversity, and education, as well as proportion of contributing authors identified as trainees, official societies, and academic institutions/training hospitals. These data should help the SVS and its members to establish a more directed social media effort to facilitate its use during national gatherings.