Healthcare's Missing a Big (Data) Opportunity

Cleveland Clinic's top 10 medical innovations for 2012, released at the annual Medical Innovation Summit last week, included a mix of cool medical devices, new treatment protocols and procedures, and other healthcare technologies that, according to the organization, have significant potential for short-term clinical impact and a high probability of success.

The list includes wearable robotic devices, genetically modified mosquitoes, and medical apps for mobile devices—and one item that's not quite like the others: Harnessing big data to improve healthcare.

"Healthcare data requires advanced technologies to efficiently process it in reasonable time, so organizations can create, collect, search, and share data, while still ensuring privacy," the organization said in a release. "In this way, analytics can be applied to better hospital operations and tracking outcomes for clinical and surgical procedures. It can also be used to benchmark effectiveness-to-cost models."

"Big data"—impossibly large and unwieldy data sets that contain, hidden deep within, a treasure trove of potential for healthcare research and discovery, could have a dramatic impact on efficiency, cost, and quality of healthcare.

A report by the McKinsey Global Institute estimates that better use of big data in healthcare could generate an additional $300 billion in long-term value, with approximately two-thirds of that coming from a direct reduction in national healthcare expenditures.

The federal government, which is the biggest source of big data, is looking for ways to help the industry use data to improve healthcare. Agencies and offices from the White House to the National Institutes of Health to the National Science Foundation to the Department of Health and Human Services to the Office of the National Coordinator for Health IT are partnering with researchers and private IT companies to develop tools to harness big data sets. 

NSF has funded several projects focusing on cloud computing to help researchers store, index, search, visualize, and analyze data, "allowing them to discover new patterns and connections," Tom Kalil, Deputy Director for Policy at the Office of Science and Technology Policy, wrote in a recent White House blog post on big data. HHS has spearheaded many projects, including efforts to ease data-sharing among rural healthcare providers.

And ONC reaffirmed its pledge in its recently released Federal Health IT Strategic Plan to create what it calls a "learning health system" that uses information to continuously improve health and healthcare, and has placed renewed emphasis on patient access to data. 

The healthcare industry is still struggling to get its arms around big data. Analyzing large data sets is not easy. But health leaders can emulate and implement some best practices, the McKinsey report authors write.

The report points to a few healthcare organizations that are doing a good job with big data, including the Department of Veterans Affairs' health information technology and remote patient monitoring programs. "The VA health system generally outperforms the private sector in following recommended processes for patient care, adhering to clinical guidelines, and achieving greater rates of evidence-based drug therapy," McKinsey says. These achievements are largely possible because of the VA’s performance-based accountability framework and disease-management practices enabled by electronic medical records and health IT.

But most hospitals and healthcare systems aren’t ready for big data—not yet, anyway. What's stopping them? Part of the problem is that healthcare data—provider clinical records, payer claims, and pharma research and development, for example—is fragmented. It lies in multiple systems in varied electronic and non-electronic formats and has many owners that often don't share well.

Farzad Mostashari, MD, ONC's national coordinator for health IT, has said repeatedly that the push to get providers to become meaningful users of electronic health records will not only improve quality, patient safety, and care coordination but will also set the groundwork for "massive liberation of patient data."

Patients "have the legal right to access [their] own health information. And that has been the case since HIPAA was written,” he said at a recent ONC town hall event. "The problem is … that’s not always so easy.” Doctors and other providers are "not too thrilled" to share data, in part because they think it will mean more work and in part because providers feel a proprietary ownership of patient data, he said.

It’s “not just a technical problem but a mindset problem,” he said.

Without that massive liberation of data—and a change in attitude about who owns it—big data for healthcare will remain a big missed opportunity.