Following the crowd: patterns of crowdsourcing on Twitter among urologists

To examine patterns of crowdsourcing on the social media platform Twitter among urologists.

Urologists' public Twitter accounts were reviewed for original posts seeking clinical advice or feedback, and associated reply posts, before and after the 140-character-limit expansion in 2017. Predictors of responses to crowdsourcing requests were determined using multivariable regression. When patient data were posted, we noted whether consent was documented.

A total of 276 posts in 23 crowdsourcing requests prior to character-limit expansion were analyzed. Reasons for crowdsourcing included requesting solutions to a clinical dilemma (82 posts, 30%); advice seeking about a surgical plan (77 posts, 28%); surveying colleagues' experiences with a new product (64 posts, 23%); and soliciting feedback about a proposed course of action (53 posts, 19%). Recent completion of training (as a proxy for inexperience) did not appear to disproportionately motivate crowdsourcing; authors' median time in practice was 7 years, and authors practicing for ≤ 7 years initiated 57% of requests. 22 (96%) crowdsourcing requests received ≥ 1 reply. Of 15 requests about a specific patient, eight included imaging, but only one cited patient consent. A second analysis of 184 posts in 17 crowdsourcing requests initiated after character-limit expansion demonstrated significantly more authors replying per request (P = 0.01), but no change in the frequency of patient-specific crowdsourcing or citation of consent.

Urologists are leveraging Twitter for crowdsourcing clinical guidance and experiential knowledge. Nearly all requests were answered, suggesting low barriers to entry for novice users. Even after character-limit expansion, dissemination of potentially identifiable patient data remains a concern.