Twitter’s Potential for Scalable Outreach

February 4, 2014

You can’t deny the fact that Twitter has become a dominant force in the micro-blogging realm. A relevant technology blog, socialmediatoday.com, recently reported that Twitter was the third most popular social media platform (behind Facebook and MySpace), with nearly 232 million active users worldwide, more than 5,000 tweets composed each second, and nearly 300 billion cumulative tweets (access the infographic here). A 2013 study commissioned by The Pew Charitable Trust suggested that 18 percent of online adults use the popular platform. The Centers for Disease Control and Prevention (CDC) use Twitter as their sole micro-blogging platform and actively encourage the use of Twitter as a cost-effective solution to reach stakeholders with relevant health information. The big question here is: Does Twitter have the potential to bridge the gap between scientific innovation and effective dissemination? Possibly. 

Here at the Center for Injury Research and Prevention (CIRP@CHOP), we have a research-to-action-to-impact outreach strategy that relies on a large network of child and adolescent professionals interested in injury prevention and treatment. In an effort to expand the reach of our network, we leverage the opportunity of micro-blogging to disseminate pediatric-related injury health information to researchers, physicians, advocates, and industry and governmental stakeholders, as well as to the general community. @safetymd, the official CIRP@CHOP Twitter handle, is led by the Center’s Scientific Director Flaura Winston, MD, PhD and is taking advantage of the viral nature of Twitter to disseminate injury-related health information across our network.

Evaluation is crucial, and obviously there’s no point in utilizing these resources unless they have truly been vetted as cost-effective solutions to disseminate our science across existing and new networks. Interestingly enough, Twitter has a well-documented and widely accessible application programming interface (API) that allows researchers, hackers, and the general public to mine its data and extract analytical datasets. In the past year and a half through a grant from the Pennsylvania Department of Health, CIRP’s Digital Health Initiative has partnered with Drexel University’s College of Computing & Informatics to conduct a rigorous and systematic evaluation of our social media dissemination strategies.

In a previous blog post, we reported on the results of using a Twitter chat to effectively disseminate health information across existing and new networks. CIRP is currently focusing on how messages travel through Twitter networks and how to apply quantitative techniques to classify users in networks based on the content they post. At the 2013 ISRII meeting, we demonstrated how you can actually observe and measure “message dilution,” a process in which original tweets exponentially become de-emphasized or “diluted” in successive layers of re-tweets. For the average Joe-Sally-Venk practitioner or researcher, this could be a potential new tool in your kit to accurately measure the reach of your messages across Twitter networks.

I recently presented some of this work at the Medicine 2.0 meeting in London and was able to show how roughly 50 Twitter users-worth of tweet data boiled down to about 5 differentiated clusters of users with similar tweeting content. We are currently in the process of applying this technique to nearly 200 user handles, but our preliminary results truly demonstrate the potential of using Twitter for targeted messaging campaigns. It was mind-boggling to see how many researchers, practitioners, and entrepreneurs were interested in applying these techniques and working in this space. The general consensus was a basic fear of breaching privacy, a lack of novel methodologies to conduct evaluation of these platforms, and the overwhelming anxiety of getting waist deep in big data!

There is no question that Twitter and other large micro-blogging platforms have the potential to help bridge the gap between scientific innovation and dissemination. More work, including developing and testing new methodologies for assessing effective dissemination, is needed in order to determine whether or not these platforms offer cost-effective and pragmatic solutions.