University of Pittsburgh Medical Center committed to "meaningful use" long before it was considered meaningful. We saw it as a means to improve patient safety, service quality, and the patient experience. It was a lonely journey, fraught with a lack of standards, closed systems, vendors that refused to connect to one another, and immature technologies. But now that meaningful use is an accepted part of the national agenda, we realize that the much tougher job of creating truly "connected medicine" lies ahead and will fundamentally transform healthcare.
Information as a commodity. It used to be that information was power: The paper chart connected the patient to a specific hospital or doctor. Today, it's the ability to act on information — to screen, translate, and interpret it — that empowers clinicians. In a highly digitized world, we deal with images, structured data, and unstructured data, and the amount is overwhelming.
Healthcare will see real change when all this data converges. Context, in particular, will be important. The ability to compare me as a patient with people just like me across the nation requires that all valuable data must live outside of the system in which it was created, decoupled from the source and rich with context. The concept of context in medical data isn't new but to date has existed only within the silo of individual systems.
Vocabulary is key. If we hope to derive meaning from all the data we've stored and to interact with it using advanced analytics, we must standardize the medical vocabulary. Basic "rear-view" reporting and analytics involving metrics such as average length of stay and readmission rates can be built around code sets that are specific to one department or organization. But advanced analytics will require common vocabularies that span electronic health record vendors, episodes of care, legacy codes, and multiple standards.
Vocabulary harmonization must happen at the source, where the intelligence behind the definition is present. This is no small challenge. In my organization, our catalog of lab codes for 20 hospitals equates to a compendium of more than 10,000 orderable items. Convincing this one department of the importance of participating in a larger connected world is a major cultural change. And it's just one of 29 service areas at UPMC.
The principle of trust. As medicine moves from single providers to teams of providers interacting across multiple care sites, the ability of clinicians to trust the data they're exchanging is crucial. In the absence of a national patient identifier, our ability to assure exact patient identity is limited to the data in a record. Questions of trust will extend to patients as well. Our ability to harness the value of data — to effectively fire decision-support systems based on a scoring algorithm or to accurately predict and manage risk in accountable care organizations — won't be limited by technology but by the quality of the data in our systems.
Medicine's future will be defined by the ability to connect information in different ways around the patient all the time. Exchanges of medical data must become like transactions at bank ATMs, where I expect my money and data to be available to me all the time. This should be the goal of the connected health record. The judgment calls my doctor can make with this data, the specificity of the care I receive, and my ability to take a more active role then become as easy as making a cash withdrawal.
Together, we must provide a circle of information around the patient with far greater value than any of us can achieve alone. Connected medicine allows for better use of data and better decision making, and it yields better healthcare — and that's what's meaningful to our patients.
Lisa Khorey has 15 years of experience integrating electronic systems in healthcare. She leads UPMC's interoperability program and its efforts to create a unified and connected patient record. Write to us at email@example.com.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?