Blog: Health Information Exchange Revisited

Next week I’ll be traveling to D.C. to speak at the Fourth National Accountable Care Organization Summit sponsored by the Engelberg Center for Health Care Reform at Brookings and The Dartmouth Institute for Health Policy & Clinical Practice. The topic is Health Information Exchange (HIE). This is a curious topic for such a prestigious conference. Why talk about a topic that has represented two decades of disappointment and missed opportunity? Maybe there is something interesting going on in this health care backwater. I will use the next several installments of this blog to explore what that something might be. In this installment, I will establish some historical perspective around information sharing in health care, and then future blogs will focus on its potential and examples of progress being made.

The aspiration for a clinically and financially integrated health care delivery system has been with us for several decades. The reason is that from a consumer perspective, our health care delivery system today is intimidating, confusing, fragmented and downright frightening. Since the practice of medicine produces variable outcomes in humans and comes with no guarantees, trust is an essential component. Trusted information from a trusted advisor provides the needed reassurance. The dream of a patient/consumer-driven health care system remains aspirational so long as the essential information required to empower consumers to help make better decisions remains elusive. Even then, the role of the physician will remain central because of the expertise required to interpret the information, complete diagnoses and access the right services. (That is why after years of careful consideration and research, we, at Lumeris, had the audacity to modify the simple, elegant Triple Aim and its ability to summarize the goals of our health care system by adding Physician Satisfaction to create the Triple Aim Plus One).

Hope has been fostered that one day all the information needed to provide the best possible clinical and financial outcomes for a patient will be readily available. We could optimize the right balance between patient expectations and perceptions (a simple definition of customer satisfaction) by having this information instantly accessible and actionable at any point in the delivery of health care services.

In the 1990’s, health care took note when the supply chains of both the financial services and manufacturing industries were successfully integrated on a scale that had previously been unimaginable. The consequences of this deep information integration were complete globalization of both industries and an economic restructuring of the countries they serve. Efficiency and quality standards soared while traditional, and frequently protective, market barriers collapsed, giving consumers an unlimited choice of products and services at much lower prices. The effect was bringing to consumer products many of the benefits Moore’s Law has had on technological innovation, accurately predicting a doubling of capability every two years with continuously declining prices. What a wonderful aspiration for our health care system.

In the new world of integrated financial services and manufacturing, consumers could just as easily purchase a stock on the Malaysian Stock Exchange as on the New York Stock Exchange. And, since every hour of labor available for manufacturing hard goods was now competing against every other hour of labor on the planet, manufacturing costs fell dramatically and logistics improved to the point that manufacturers no longer had to maintain much inventory. Just-in-time inventory management and information-based logistics led to consumer product orders showing up instantly and simultaneously at a components manufacturer in China, an air shipment company in Tokyo, an assembly plant in Iowa, a trucking company in Chicago and a local distributor in Pennsylvania ─ all contributing to the consumer having their purchase in two days delivered to their home.  Think of that UPS tracking tool that has eliminated the question, “When do I get my package?” Offering you the ability to track the location of your package in real time resulted from enormous investment in supply chain information integration; providing consumers with exactly the same information at the same time that historically was only available to the manufacturer, retailer and shipper. The root of all this innovation can be traced to powerful information exchange that obsoleted processes and business models that could not adapt or were built around constraining the free flow of information.

So why has information exchange been so difficult in health care?  Process engineers would suggest the health care industry’s structure has largely been built on practices more akin to a cottage industry that made the scope and timing of processes highly unpredictable, making automation more difficult. Except for the large integrated health care delivery systems that could impose standardization on all elements of the delivery system, the vast majority of the industry relied on highly customized processes driven by the local preferences of physicians and patients, making standardization difficult.

While there is truth in this argument, a closer look reveals a much more systemic challenge to information sharing incentives, or more appropriately, disincentives. As I have mentioned in prior blogs, I am a strong believer that behavioral economics is a factor in most human decisions, and health care is no exception. Health care over decades built powerful business models providing perverse incentives for not sharing information. Much of this behavior has been justified under the umbrella of patient privacy. Actually, health care subscribed to the axiom of “knowledge is power” more so than any other industry.

With powerful incentives not to share information or collaborate around the benefits that ensue from information transparency health care built barriers ─ some unintentional ─to facilitate broad-based information sharing. For example, one rationale given for the difficulty of information sharing in health care is that the industry’s IT systems aren’t architected for it. This is basically true, but the underlying reason health care’s systems infrastructure is so Balkanized and dated is that the business and care delivery models it supports did not require open architected modern systems technology. It is certainly not because of a lack of investment in IT.

Health care has an extremely high preponderance of installed “closed” architected systems. Why?  Health care tolerated the constraints of these systems because they were perfectly aligned with their business strategies. Collaborative, open architected systems were unnecessary and even possibly viewed as a threat. Health care IT vendors were likewise comfortable with this dichotomy because it reduced their need to innovate and increased the shelf life of their software well beyond what vendors serving other industries experience. This created an unusual and somewhat unintentional symbiotic relationship between vendors and their customers. Customers had a high tolerance for old software.

This perversity of disincentives to share information manifests itself in lots of unlikely ways. For example, the original expressed intention of HIPAA legislation was to encourage information sharing by having the regulators define standards for data protection. However, the actual and implied penalties for breaches in data security are so severe, the effect has been just the opposite, indirectly reinforcing the arguments of those against data transparency. A more subtle example is the unresolved dilemma of a physician receiving information about one of his patients stored in another physician’s EMR. If that information goes into the receiving physician’s EMR and she acts on it and it proves to be wrong, who is liable? Who provides the credentialing of health care information? This simple example can become a major deterrent to information sharing by otherwise well-intentioned collaborators, as well as an excuse to those committed to maintaining the status quo.

All that said, am I optimistic about the future of information sharing in health care? You bet I am. Because of breakthroughs in technology? Partly. But the real breakthrough is in incentives. For the first time in my lifetime, incentives to share information across the continuum of care are soon going to be pervasive. And it comes with a certain and defined date. It is all very exciting, and I aim to expand on this topic next week at the conference and in future blog posts.