This summer, Carolinas HealthCare System made the news rounds as a warning of the new levels of data mining available to healthcare companies. In Hospitals Are Mining Patients' Credit Card Data to Predict Who Will Get Sick, we get a very Big Brother type of picture of the invasiveness of such data mining with an illustrative picture showing a doctor saying, “Don’t lie to me, Susan, I know about the 2 a.m. Papa John’s deliveries.”
It makes for dramatic copy, but it’s still in the realm of fiction rather than fact, as I found our when I contacted Carolinas HealthCare and got a response from Jason Schneider, Director, Clinical PR. He explained that the article “focused on how providers could use data for in the future and didn't include details what data we are currently using and how we are using it.”
The data they are currently using does not follow an individual’s consumer trail but looks at things like socio-economic circles, neighborhood limitations, and cultural affiliation that could shape one’s access to healthcare. One example of that was identifying why patients in one particular area were not coming in for regular doctor’s visits. It turned out that it didn’t have reliable public transportation to a doctor's office. After identifying the geographic problem, Carolinas HealthCare set up a doctor in the neighborhood itself.
As the person quoted in each of the articles on Carolinas use of data is Dr. Michael Dulin, chief clinical officer for analytics and outcomes research at Carolinas, I contacted him and spoke with him on the phone. He explained that Carolinas has a decade of experience in using data to improve healthcare by identifying individuals within contexts that could pose obstacles to care.
Read more in