LAUSR

BACKGROUND Little is known about the key composition of a successful tweet in urology. OBJECTIVE To investigate for predictors of engagement with urology content on Twitter. DESIGN, SETTING, AND PARTICIPANTS This was a cross-sectional study based on 2-wk Twitter data surrounding a major international urology conference. INTERVENTION We examined the engagement for all original tweets containing the hashtags for the European Association of Urology conference ("#EAU19" and/or "#EAU2019"). OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS Study outcomes included engagement with tweets, as measured by the number of "likes" and "retweets." Tweet- and Twitter user-related parameters of each individual tweet were recorded. Multiple linear regression analyses were performed to investigate for predictors of likes and retweets. RESULTS AND LIMITATIONS From March 9 to 22, 2019, there were a total of 37 222 tweets. Among them, 3534 were "original tweets" that had 31 889 likes and 10 031 retweets. On multivariable analysis, the word count, number of mentions, and presence of a photo were predictors of likes and retweets. An increasing number of hashtags were associated with fewer likes. The number of "followings" and "followers" of the contributor, and their time since joining Twitter did not have any associations with the number of likes or retweets. The major limitation of the study is the lack of assessment about the quality of the tweet content. CONCLUSIONS Based on the Twitter data from a urology conference, we concluded that the word count, number of mentions, and presence of a photo within the tweet were associated with audience engagement. PATIENT SUMMARY We could engage the audience more successfully by increasing the number of words and mentions, and including a photo within a tweet. The results formulated the basic principles in creating successful tweets for sharing urological knowledge.