Tear Down the Walls and Liberate the Data

(cross-posted on Microsoft on the Issues) 


In 1987, President Ronald Reagan gave one of his most well remembered speeches.  Few of us could forget his words to Mikhail Gorbachev to “tear down this wall”—proving to be prophetic when the German Democratic Republic announced the re-opening of the border in 1989, and the subsequent destruction of the Berlin Wall.   What followed?  A new flow of people, ideas, commerce, and capital—creating the groundwork for unification and a better way of life in Germany and Eastern Europe,  benefiting all of us economically and politically in unanticipated ways.


A similar type of disruptive change needs to happen in the health ecosystem today.  Just as the free flow of ideas and capital were the foundation for dramatic improvements in society, so should the free flow of health data be the foundation for realizing a future of secure, personalized, data driven medicine in health.


Yet many outdated ideas and mechanisms stand in the way of change; the most important of which are the now impractical walls that were erected and still exist around patient-data.  And all of us—IT vendors, providers, payers, pharmacy benefit managers, policy makers and others in the health ecosystem have enabled the walls to exist in spite of the obvious benefits to patient safety and the overall health economy.


Real-time, accessible, meaningful and comprehensive data is fundamental to health care as a whole–to make a diagnosis, provide quality care, pay the right bill, discover new therapies, and so on.  What’s of paramount importance is liberating the data and making it available for re-use in different contexts.  This is critical for improving outcomes, paying for value, creating a learning healthcare system, enabling discoveries and fundamentally changing the dynamics of the ecosystem.  We should be treating health data as a vital asset—health enterprises and consumers—to drive an efficient, high-quality, value-based, evidence-focused future for medicine.


So why isn’t there data liquidity or the appropriate flow of data in the ecosystem?  There are two major walls preventing the liberation of data and each is starting to have some cracks.


First, there is the “it’s-my-data” wall put up by hospitals, insurance plans, pharmacy benefit managers, and others.  They believe there is some competitive advantage by keeping the data inside their walls.  And there are lots of excuses supporting their position–patients don’t want it, they can’t understand it, it might do them harm blah, blah, blah.  This wall is starting to crack.  Many institutions recognize that the consumer has a right to a copy of their data and are making the appropriate connections to personally controlled health data repositories like HealthVault or Google Health.  Many others are writing about it too—John Moore asserted in a recent post, “Personal health data belongs to the consumer and the consumer should decide how it is shared. This is a very radical concept that still has most providers, payers and other data holders shaking in their boots.”  


The second wall is the “waiting-for-the-right-standards-set-by-government” wall.  There are multiple excuses buttressing this wall; the core of which come down to technology, standards or policy excuses.  Without debating each point–one inexcusable barrier is the IT enterprise system vendors who make it difficult or expensive to get access to the data, to separate it from the application.  They believe that proprietary “lock-in” provides them with a strategic advantage.  We’re all in business and need to create strategic value for our products, but let’s do it in the application layer—rules, workflow, user experience, price, or services—not by trapping patient data in a proprietary database structure. 


Our customers and partners and their customers need to be able to re-use their health data, and in ways they haven’t always thought of or anticipated. They have to be able to build cross-vendor systems to improve care.  If information can be made liquid—flowing from where it is generated to where it is needed, and combining it with other bits of information to provide a comprehensive view—it can be tremendously powerful.


We, in the IT industry, can step up and be a driving force in enabling data to become liquid—specifically, doing this by separating data from applications.  This is one of the recommendations from a study by the National Research Council of the National Academies that takes a look at what types of computational technology and investments are best for improving health outcomes.  Let the excuse not be that the data is trapped in systems that we built, that we have to wait for standards.


We need to enable this, and we can start to do it today.  Just look at the Health Information Exchange in Wisconsin and CVS MinuteClinic.  In the former, value is being added immediately to users in the ED, without requiring all the participating EDs to change their systems or to be standards compliant (or CCHIT certified).  At MinuteClinics, summary after-visit health data are made available to customers online using the Continuity of Care Record standard.


There’s a proven model for extracting and transforming data in many ways—HL7 feeds, non-HL7 feeds, web services, database replication, XML and XSLT, and more—and along the way we can create value by interpreting the data and adding metadata.  At Microsoft, we’re doing it today both in the enterprise with Amalga and across enterprises to the consumer with HealthVault.  We hope other vendors follow this lead to drive better outcomes for patients, and we expect buyers of IT systems to demand vendors to meet this standard (excuse the pun).  Where standards are available, we should leverage them, and where standards do not yet exist, we should output the information in a consistent consumable format for the install base.


I understand that there are many, many complicated aspects of this problem, including the need to reform our payment system.  But we don’t have to work out everything first to begin building a better, more data enabled, data rich and accessible health delivery system now.  David Kibbe did a nice job simplifying and laying out some core recommendations in his post, Five Shovel-Ready Health Care Reforms. 


I have argued previously that HiTech should focus on investments which leverage existing digital data sources and drive better health outcomes.  With the dollars that are being allocated to EMRs, it’s critical that the data is liquid and that the consumer is connected to the data in a meaningful way


I understand that this kind of disruptive change can be uncomfortable because the forces unleashed can lead to unpredictable results for specific stakeholders.  But like the fall of the Berlin Wall, it is clear that the most important stakeholder—the citizen or patient—will be better off, and the other key stakeholders—providers, payers, policy makers, etc.—will participate in a healthier ecosystem.   There will be profound, new opportunities for everyone in this future.


The time for excuses is over.  Let’s tear down the walls and get the data flowing.  We can do it now.

Comments (4)
  1. Anonymous says:

    The Food and Drug Administration recently announced that the Office of the National Coordinator for Health Information Technology is launching the Sentinel Initiative with the ultimate goal of creating and implementing the Sentinel System – a national,

  2. Anonymous says:

    I’ve been following your blog for a few months now, and I have to say that I have found your posts very much on-the-money in regards to improving healthcare with technology / IT standards. I’m a resident physician at the University of Washington and I wanted to add another way that opening up health data can improve our system and lower costs. For generations, residents have wasted hours of their valuable training time on data-gathering. Even in this day and age of electronics and lab data that’s available at the touch of a button, we are STILL spending hours a day doing things like, calling various departemnts in the hospital figuring out if X test has been done / if not, when it will be / when the results will arrive / finding out that peice of particular data is provided in an esoteric form that is only accessible by XYZ manuvers or that the fax has been lost. And that’s just within your hospital. Good luck trying to track down records from an outside hospital, because even if they are faxed over appropriately, maybe it’s 2 days later and the patient has changed floors in the hospital and now those records have gotten trashed.  And then if your patient requires a healthcare power of attorney to make decisions, hopefully their contact information is readily available or you’ve signed yourself for many times, cross country phone tag figuring out who to talk to. Or even just reading an EMR. I’ve worked with several popular EMRs, and while they are sophisticated, the clinical notes are just a pile of electronic files equivalent to a paper record and there is perhaps a 1:10 ratio of physician notes to other notes which you must wade through.

    This is the reality of our physician training. And who pays for training residents?  Medicare. Which means taxpayers. The public is likely unaware that they are paying quite a few dollars for all physicians to first train as glorified secretaries.  

    For me personally, that’s what it means to liberate health data. Better patient care — the Veterans Administration provides some great care for veterans because all their patients stay within the same system, same health data accessible from any VA — and, better physician training.

    I was wondering if your group / Microsoft is working with the University of Washington at all in developing health IT software or standards?

  3. Anonymous says:

    After many long months of discussion and debate , the first draft of Meaningful Use has come out. I’m

Comments are closed.

Skip to main content