At the White House Summit on Precision Medicine, President Obama made good on his promise to "hijack the panel,” and it was worth it. He finally got to the real root of healthcare's data sharing problem. He stated, the "problem is that data is siloed." 

When my two little sons share their toys there are all kinds of "political" and "cultural" barriers they have to work through – whose toy was it first, which one is better, when do I get it back, what do you have to offer – but once they decide they want to share, it's easy. One hands the other one a toy and the playing begins.

Not so with healthcare data, and the core problem almost never gets discussed because we're busy talking about the cultural barriers to data sharing. It was great to hear the president finally acknowledge the problem is that the data is siloed. To simplify, think about it this way: Say you're in a large healthcare organization. You have claims data, financial data, clinical data, radiology slides, maybe even some cool IoT data. You decide for a very important initiative (let's say improving diabetic care or giving your customers or patients a better consumer experience) that you want to combine all that data into a 360-degree view of the patient, or what CHIME and HIMSS CIO of the Year Craig Richardville, of Carolinas Healthcare System, recently called a "panoramic view" of the patient. This is really the only way you're going to significantly improve the quality of care, not just by tacking on some new "app.”

If you want to improve the precision of patient care you need to be able to integrate and utilize all the data you have. Now remember, and this is key, we're not yet even talking about data anyone else has about your patient—that would really require "sharing”—just all the data you have. Your CEO has made the decision, the CIO and CTO have agreed to do it, but when you start the project you figure out it's going to be somewhere between really, really hard and impossible. Why? Because most data in healthcare is stuck in silos. It's stuck in 25-year-old technology that makes data integration, what we really mean by data sharing, very hard. In order for the data to be usable, it all has to be "transformed" into the same form. Getting the data “transformed" can take months, in some cases years, and millions of dollars. Nothing kills a project faster than years and millions. When we say "culture trumps strategy every time" we often forget that both culture and strategy can be trumped by reality. 

It's very encouraging to see big players in healthcare, including the government, stating publicly they are on board with data sharing, but in order for data sharing to turn into better outcomes – to have an actual effect on the quality and cost of healthcare - that data will need to be able to be utilized in an operational, transactional system, securely and at scale. A data lake off in the corner with an analytic package on top isn't going to do it. Healthcare organizations need to address the fact that their legacy systems are "hitting the wall" when it comes to data integration, and in order to deliver on the promises being made they need to address data integration at the enterprise level. When an organization can fluidly and securely integrate and utilize its own data to improve care and run its business more effectively and efficiently, the cultural barriers will start to fall away. This technology exists today and healthcare needs to understand the real problem is "the data is siloed," and fix it.   

Fox is vice president for Healthcare and Life Sciences at MarkLogic.