Though data analytics has the potential to improve health-care delivery, medical providers have yet to merge big data with health-care delivery en masse. Regardless, the possibility to one day mine sophisticated data sets to better understand patients’ living habits and health conditions excites scientists and physicians.
On Thursday, panelists at the Data Innovation Day event, sponsored by the Center for Data Innovation and held in Washington, D.C., discussed the emerging world of health-care analytics. They covered the current state of data in electronic health records, the openness and availability of datasets, and what health-care analytics could bring if fully realized: If doctors could access data that most people take for granted, it could revolutionize the medical field.
“Your refrigerator could know much more about your nutrition than I will ever elicit from you by asking you what you had for breakfast, lunch and dinner,” said Russ Cucina, the medical director of Information Technology at the University of San Francisco. “How do we get those data into electronic health records and other systems?”
But electronic health records themselves could use improvement as it is. The data is difficult to collate in large part because it’s in different formats from system to system.
“Despite the fact that electronic health record programs have been incredibly successful, it would be a mistake to imagine that the data in those systems is interoperable. It’s very much not the case,” Cucina said. “The lack of data interoperability across electronic health records represents a barrier.”
But the barriers aren’t always technical. The medical community, filled with competitive doctors, scientists and research institutions, is hesitant to share data regardless of a digital revolution.
Marcia Kean, chairman of strategic initiatives at Feinstein Kean Healthcare, mentioned the H. Lee Moffitt Cancer Center’s “treasure trove” of cancer data on 50,000 patients that the research institute would likely never share. Cancer centers are competitive with each other and like to keep their findings private.
Beyond competition, data disclosure can be limited due to protocol in general. According to Nina Preuss, program manager for the Neuroimaging Informatics Tools and Resources Clearinghouse, a lot of medical data doesn't become available until it’s cited in a research paper. People have to wait until a medical journal is published before they ever know anything was discovered, so data sits ignored unless someone’s prompted to do something with it.
The medical field’s digital upgrade, and the policy changes it may bring, can’t come fast enough.
“The health-care field, in going digital, is so backward that we’re almost leaping directly from the 19th century into the 21st,” Kean said.
But health care is a unique industry, she continued, with life-and-death consequences, so data-informed decisions carry extra weight.
“The implications of going from data to insight can be very serious, at both the individual level and the population level,” Kean said. “If you get insights on how effective a drug might be or not be and you make a decision about that drug as an individual, it may turn out to be the wrong decision. It could kill you or not kill you."
However, a key to improved health-care analysis could come from patients more than from doctors and policymakers… if the patients were mobilized and paying attention.
“Health care, as big a part of the economy as it is, is still something that people spend very little of their life doing or receiving. If you spend half an hour at your doctor every three months, that’s a lot,” Cucina said. “And even if you’re very ill, you’re only going to be hospitalized a couple times a year, but your health is something that is happening 24 hours a day, 7 days a week.”