Prior to the 2016 ASRA spring regional anesthesia annual meeting, a new social media education and engagement strategy was implemented and directed towards members and meeting attendees [11]. No specific strategy existed prior to this year, and implementation continued for the 2016 ASRA fall pain medicine meeting. In August 2016, ASRA posted an article on its website with specific tips on writing blogs, which would also apply to other social media posts (https://www.asra.com/news/117/tips-for-writing-blog-entries). All correspondence (e.g., emails, flyers) prior to the 2016 fall pain medicine meeting featured the hashtag #ASRAPain16. Although the 2015 fall pain medicine meeting had a hashtag #ASRAPain15, this was not heavily promoted. A ‘Social Media Team’ was identified in 2016 based on members and past meeting attendees who had been previously active on Twitter. These individuals were asked to promote the upcoming fall pain medicine meeting in advance as well as ‘live tweet’ [13] and actively participate in the #ASRAPain16 Twitter conversations during the conference.

Tweets for each of the conferences were collected as transcripts starting with registration opening on the first day of the meeting and ending on the last day of the meeting at the conclusion of the last session. Impressions (the number of tweets multiplied by the number of followers who can view the tweets) and number of Twitter participants (those who created at least one tweet) were collected from Symplur free analytics [11]. All tweets were manually counted and reviewed by two investigators independently. Among the complete list of tweets, replies (tweets in response to another Twitter user’s original tweet), and retweets were differentiated from original tweets. Tweets were assigned to one of four categories (scientific, logistical, social, or other) based on a classification system described previously [11]. Specific content of tweets was assessed for the inclusion of media (e.g., images, video), external links (‘for more information’), and mentions (including another Twitter user’s ‘handle’ or username in the tweet). Analyses were performed with NCSS-PASS statistical software (NCSS, LLC, Kaysville, UT). Comparisons of proportions were conducted using the Chi square test or Fisher’s Exact test if n < 5 in any field. For all analyses, P < 0.05 was considered statistically significant.

The number of tweets generated by the ASRA fall pain medicine meeting increased from 148 for #ASRAPain15 to 312 for #ASRAPain16. The number of impressions increased from 527,489 for #ASRAPain15 to 1,746,713 for #ASRAPain16, and the number of participants also increased from 128 for #ASRAPain15 to 196 for #ASRAPain16. Nearly all tweets were original and not replies or modified retweets: 136 (92%) for #ASRAPain15 and 284 (91%) for #ASRAPain16 (P = 0.758). Most tweets were scientific in nature: 57% for #ASRAPain15 and 161 (52%) for #ASRAPain16 (P = 0.242). Some form of media (e.g., image) was included in 92 (62%) of #ASRAPain15 tweets compared to 164 (53%) of #ASRAPain16 tweets (P = 0.053). Links were included in 15 (10%) of #ASRAPain15 tweets compared to 71 (23%) of #ASRAPain16 tweets (P = 0.001), and at least one other Twitter user was mentioned in 63 (42%) of #ASRAPain15 tweets compared to 168 (54%) of #ASRAPain16 tweets (P = 0.024). The number of tweets with at least one retweet was 65 (44%) for #ASRAPain15 versus 175 (56%) for #ASRAPain16 (P = 0.015).

This is the first description of a deliberate social media strategy to promote Twitter activity and engagement for an annual pain medicine conference. Despite being one of the oldest subspecialty societies in anesthesiology, ASRA has embraced technology and innovation and has made social media a point of emphasis in its communications. Following implementation of a focused strategy between the fall 2015 and 2016 annual pain medicine meetings of ASRA, the number of tweets, participants, and impressions generated as well as the content quality of tweets increased in the subsequent meeting.

Today, Twitter functions as a tool for the modern learner [2,14]. While the use of Twitter during scientific meetings is becoming commonplace [10,11], the selection of metrics in research studies analyzing Twitter data has not yet been standardized [2]. We propose four categories of Twitter metrics for the purposes of reporting Twitter use in medical education (Fig. 1). We further suggest the following best practices for live tweeting (Table 1) [15]. Our example of one professional pain medicine society’s integration of Twitter into its annual scientific conference demonstrates that a focused implementation strategy may be associated with improvements in the quantity and quality of live tweets and attendee engagement.