Hospitals Are Mining Patients’ Credit Card Data to Predict Who Will Get Sick

Shannon Pettypiece and Jordan Robertson:

Imagine getting a call from your doctor if you let your gym membership lapse, make a habit of buying candy bars at the checkout counter, or begin shopping at plus-size clothing stores. For patients of Carolinas HealthCare System, which operates the largest group of medical centers in North and South Carolina, such a day could be sooner than they think. Carolinas HealthCare, which runs more than 900 care centers, including hospitals, nursing homes, doctors’ offices, and surgical centers, has begun plugging consumer data on 2 million people into algorithms designed to identify high-risk patients so that doctors can intervene before they get sick. The company purchases the data from brokers who cull public records, store loyalty program transactions, and credit card purchases.
 
 Information on consumer spending can provide a more complete picture than the glimpse doctors get during an office visit or through lab results, says Michael Dulin, chief clinical officer for analytics and outcomes research at Carolinas HealthCare. The Charlotte-based hospital chain is placing its data into predictive models that give risk scores to patients. Within two years, Dulin plans to regularly distribute those scores to doctors and nurses who can then reach out to high-risk patients and suggest changes before they fall ill. “What we are looking to find are people before they end up in trouble,” says Dulin, who is a practicing physician.
 
 For a patient with asthma, the hospital would be able to assess how likely he is to arrive at the emergency room by looking at whether he’s refilled his asthma medication at the pharmacy, has been buying cigarettes at the grocery store, and lives in an area with a high pollen count, Dulin says. The system may also look at the probability of someone having a heart attack by considering factors such as the type of foods she buys and if she has a gym membership. “The idea is to use Big Data and predictive models to think about population health and drill down to the individual levels,” he says.

Related: Making pizza delivery more efficient.