Improving EHRs at Vanderbilt University

For all its promise, using data effectively in health care presents a special set of challenges. From his post at Vanderbilt University, Middleton, a leading informatics expert, weighs in on what the field needs to do to ensure data can drive improvements across the industry.

Sync: For as long as there has been talk about electronic health records (EHRs) and medical data management, why haven’t they become more integral parts of US health care?

Dr. Blackford Middleton: There’s plenty of evidence of the clinical advantages of EHRs, but there wasn’t widespread adoption because of misaligned incentives. Payers accrued most of the benefits, not providers or health-care systems. In other words, “He who pays for health-care IT is not he who gains.” That changed with the Office of the National Coordinator for Health Information Technology (ONC), in 2004, and the American Recovery and Reinvestment Act (ARRA), in 2009, both of which included health-care IT initiatives and incentives. From a practical point of view, we’re still in the early stages.

Plus, keep in mind that the US system is volume-based, not value-based. So even though access to critical data might help me avoid having to admit a patient, it’s the system that saves money. I might actually lose money because of that lost admission.

Sync: But with so many technological advances on the clinical side, why does implementation of medical informatics still seem to be lagging behind?

Middleton: Even now, implementation is being driven by outside stimuli like ARRA and meaningful-use programs. The bet was that the government could stimulate adoption and meaningful-use programs, which would push physicians and health systems to innovate. That’s happened to some extent. Adoption has gone up, but the focus has been on meeting the statutory and incentive requirements. Usually there would have been negotiations between technology users and providers about features, capabilities, and how they’d be implemented. But now meaningful use has taken over, instead of the marketplace and ease-of-use—and poor usability doesn’t lead to rapid adoption.

That being said, implementing IT is the first step to laying the foundation for true reform and transformation in US health care. Without EHRs in place we can’t gather data for providing support for clinical decisions, so what we’re doing now is creating the infrastructure for a new business model for health care.

Sync: If you had just five minutes to speak to a group of hospital CIOs, what would you tell them?

Middleton: It’s critical that we move away from relying on single applications or vendors for doing everything. We need an open architecture so that external suppliers can develop highly specialized solutions with very specific capabilities. It’s obvious how successful that’s been for smartphones and Facebook—it really stimulates innovation, because it invites advances that go far beyond the company hosting the app or service.

“The new tools provide access to more resources and information than anyone can keep in their head. No one can practice medicine anymore with an ‘unaided mind.’ We all need cognitive support from health-care IT.”

Sync: How should we educate patients about the pros and cons of data-driven care?

Middleton: Patients have adopted technology in lots of different areas of their lives. It’s common to have them come to appointments with pages of information they’ve downloaded from online sources, so many are already highly informed and tech-savvy. As a result, they also expect a certain level of convenience, just like they get in their other technology or service experiences. For health care, that means offering services like online or automated prescription refill requests and appointment scheduling. It’s interesting to note that those types of services may carry more weight with patients than concerns about technology and privacy, security, or knowing that their medical information can’t be used against them to deny insurance coverage.

Using data to help manage patient relationships results in more engaged, activated patients and better outcomes. That might mean providing a diabetes management tool as part of their EHR. They can go online to track their history and report medication use and glucose measurements before in-person visits. When we did this at Harvard Medical School, not only did 75 percent of patients use it, but the health care providers themselves became more engaged and activated, too.

Sync: What opportunities and challenges excite you about data-driven health care?

Middleton: The technology gives us the ability to better understand what’s happening with patients throughout their lives. We’ll have enough information to provide continuous care that doesn’t always require an office visit and that gives us much greater insight into their health outside the clinic environment. Patient-accessible apps, for things like lab tests or home monitoring for sleep, exercise, and diet, all make this possible.

When patients need to be seen in person, the technology enables flipped office visits. Because so much data can be collected beforehand, less face-to-face appointment time is spent collecting specimens and information for testing, which leaves more time for actual patient care and communication.

We also have unprecedented opportunities to personalize care. When I have a continuous data stream and can combine phenotypic [observable patient characteristics, test results, or symptoms] and genotypic [genetic] information, I can truly offer precision, customized medicine. That’s really the new frontier. The new tools provide access to more resources and information than anyone can keep in their head. No one can practice medicine anymore with an “unaided mind.” We all need cognitive support from health-care IT.

Sync: What is the “best of all possible worlds” application of medical informatics?

Middleton: Ideally, the presence of the computer itself would be diminished, so the technology would be in the background and we could focus on the patient and delivering care. I envision a hands-free EHR with speech recognition and a dynamic screen to help present graphics and other explanations to patients. The system would “listen” in order to provide appropriate support information. It would also record the entire interaction for clinical archives and for patient reference, too. What we have now is good, but it’s like WordPerfect in 1985. We need to make things dramatically easier to use.