Healthcare conferences more and more often have active discussions taking place on various social media platforms, most notably on Twitter via the use of the event’s formal hashtag. These conversations include directly sharing what’s taking place at the event, quoting important points made by presenters, discussing the conference topics with other parties, and more. We studied how those conferences that have more patient participation in these Twitter discussions ultimately impacted the size and reach of the conversation on Twitter compared to those conferences with less patient participation.
At the Healthcare Hashtag Project, we have archived all tweets from 1,086 healthcare conferences that took place in 2013 from around the world. Of these conferences we selected those that had a minimum of 1,000 tweets using the conference hashtag during the days the conference was in session including two days prior and after the conference. A random sampling of 100 healthcare conferences was selected in order to scale down the data set. The tweet content from the final 100 conferences were analyzed to create a list of the top 100 influencers by mentions from each conference. These 10,000 participants were then categorized and identified as either a patient or non-patient. Our sole data source for categorizing was each participant’s Twitter profile description. Anyone that self-identified themselves as a patient was categorized as a patient. From this we separated the 100 conferences into two segments - those with more patients and those with less patients. Lastly, we analyzed the performance of each group on a variety of engagement metrics as a way of comparing the overall impact of the Twitter conversation between the two groups.
A total of 1,159,093 tweets were collected and analyzed for this study spanning 1,086 healthcare conferences. Of these, 382,468 tweets were considered to create the top 100 influencers by mentions for the final 100 randomly selected conferences.
We've identified 198 unique patients among the top influencers that participated in these conferences. Those 198 patients were scattered among 65 conferences. Since some patients attended multiple conferences the 198 patients created a total of 279 appearances out of 10,000 possible in the top 100 influencer rankings. 35 conferences did not have a single patient among its top 100 influencers by mentions, while the median was 1 patient per conference. Conferences with more patients outperformed conferences with less patients on social impact metrics. Conferences with more patients had an average of 4,983 tweets during the conference, compared to 2,761 tweets for conferences with less patients. Average number of participants were 869 for conferences with more patients, compared to 500 for conferences with less patients. A Welch two sample T-test was performed to test the two groups, and they were found to be statistically different at P-value 0.05. A multiple regression analysis was conducted in order to test the predictive power of our hypothesis that more patient participation results in greater social performance metrics for the conferences. The hypothesis was found to be true and statistically significant with an F-test P-value of 0.0004. And while the model's predictive power was positive, it was low with an Adjusted R-squared of 0.1494.
Healthcare conference participants use of social media, and Twitter in particular, have grown to become a significant part of the conference experience with 1,159,093 tweets for 2013. We have observed a recent strong trend to promote patient inclusive conferences. Our analysis concluded that 65 out of 100 conferences were found to have one or more patients in its top 100 influencers by mentions. While we believe this is a positive development, this study also highlights the lack of patient inclusion with 35% of conferences having no patients and the median number of patients for all conferences studied being only 1. Conferences wanting to increase their reach and impact should find every reason to encourage and facilitate patient participation. Conferences with patients were found to have a stronger signal with higher than average number of tweets, larger reach with higher a average number of participants and more dynamic conversations with higher average tweets per participant.