Big Data Holds Big Promise for Improving Health Care

Discussing Big Data and informatics can be technical, even ponderous. And anyone who has talked about how to make electronic health records meet the government's "meaningful use" goals knows it’s not exactly straightforward stuff.

Brent Lamm drew some laughs with a picture of a pumpkin and an analogy explaining health care data. He showed a picture of a pumpkin labeled simply "Pumpkin." Beside it was a pumpkin pie. The pie was labeled: "Meaningful use of a pumpkin. The reality of the matter is we need more pie," said Lamm, Director of Analytics at UNC Health Care. In other words, we need to take the data we have and make it useful. Doing so will lead to innovations and improvements in diagnosis and treatment.

Lamm was one of the speakers at the Thought Leadership Forum on Analytics & Informatics, an event Tuesday presented by the North Carolina Healthcare Information and Communications Alliance. More commonly referred to as NCHICA (pronounced "en-CHEE-ca"), the nonprofit group supports its members use of IT, analytics and informatics in health care. Joining Lamm were Eric Brinsfield, director health and life sciences at SAS Research and Development and Ed Hammond, director of the Duke Center for Health Informatics. RTI hosted the forum.

Hammond noted that the increase in processing power that is commonly referred to as Moore’s law holds true for health care technologies. We’re swimming in new technologies. And if you don’t think so, Hammond asks that you think about what passed for technology five years ago. Smart phones that are now ubiquitous are still a relatively new technology. But Hammond added that use of informatics with new technology has not kept pace. Greater use of data and informatics faces several hurdles. Brinsfield said that while SAS does a lot of work in analytics, that work is only as good as the data the company works with. If the data is of poor quality, analytics won’t yield good results.

Privacy of patient data is another concern. And clinicians must be on board, too. Sharing data is a must, Hammond said. But stakeholders must accept that cooperating and sharing data is a "win-win" situation.

How might clinicians use data? Brinsfield said someday a doctor will have all of the health care history and information about a patient upon entering the examination room. A doctor might base his treatment decision for one patient based on analysis of how 20,000 other patients who had the same condition have responded to a particular therapy, Lamm said. Technology advances are bringing us closer to the kinds of health care analysis depicted in "Star Trek," Hammond said. IBM envisions its supercomputer Watson being used to mine patient data for information that can help doctors make treatment decisions.

Easy as pie, right? Well, not quite. But Lamm, Hammond and Brinsfield agree that we’re on our way.