A Shot in the Arm - Jason Burke on Health and Life Sciences

Dave makes the case that the real value in applying analytics to social media is not simply in detecting previously undetected issues -- it is in the fact that social media is one of a very few ways to see health trends in anything close to real time.  And he is absolutely right. 

While our industry struggles with data integration, data standards, and overall healthcare interoperability, 80 million US adults are online chatting, Twittering, and Facebooking about their health conditions and treatments.  Through text mining and other analytical techniques, it is easy to envision a digital radar of health activities capable of giving us insight into rapidly emerging health topics.  I continue to be surprised that this idea is not pursued more aggressively -- both the technology and the information is available.  Hospitals have been reasonably successful in adopting social media strategies, and consumer-oriented sites have risen dramatically.  But the intelligent use of advanced analytical methodologies, including data and text mining, is very hard to find.

A recent survey by John Mack over on the Pharma Marketing Blog collected responses from visitors over how social media is used in pharma.  The techniques being used are far from sophisticated -- most of the answers were related to simple keyword searches.  For me, the respondent's comments were telling -- all of the ones John posted (a subset) focused on the obligations companies and physicians have related to collecting, managing, and reporting safety information, but none on the positive opportunity social media presents.  This trepidation may be reflective of the survey coinciding with the upcoming FDA public hearing on the Promotion of Food and Drug Administration-Regulated Medical Products Using the Internet and Social Media Tools. 

A subtle distinction in Dave's article should not be overlooked: where else can you find real-time or even near-real-time health information?  Anyone trying to find reasonably current publically-available health data to analyze will be disappointed.  Most large-scale public health data collection is somehow related to the government: Centers for Disease Control, Department of Health and Human Services, NCI, etc.  I went looking through their data catalogs, and data sets from 2007 are the most current information I could find. Have you found good sources of near-realtime health data?

There is no question this is an evolving space.  Over time we can expect today's technology initiatives -- EMRs, personal health records, OMOP and FDA's Sentinel -- to pay dividends in data currency (get it??).  In the meantime, keep those Twitter accounts twittering.