Checking In About Innovation

Late last year, Carolyn M. Clancy, M.D., director of the federal Agency for Healthcare Research and Quality (AHRQ) spoke with HCI Editor-in-Chief Mark Hagland regarding her perspectives on current developments in healthcare, and her agency's involvement in pursuing innovation and change. Dr. Clancy will be the closing keynote speaker at the Healthcare Informatics Executive Summit, to be held May 11-13, in San Francisco. Below are excerpts from Dr. Clancy's interview with Mark Hagland:

Healthcare Informatics: What is your perspective on where healthcare is right now in terms of the ability to leverage information technology strategically to improve care delivery, quality, patient safety, and efficiency?

Carolyn Clancy, M.D.: There are so many opportunities right now, and we're making progress. I'm thinking in particular about all the opportunities from the work that really began in 2004, and obviously was built on work done previously at AHRQ and the National Library of Medicine-work using health IT in clinical care.

ONE AREA OF OPPORTUNITY INCLUDES LEARNING IN A SYSTEMATIC WAY FROM THE OFF-LABEL USE OF MEDICATIONS. ABOUT 20 PERCENT OF OFF-LABEL PRESCRIBING IS LEGAL AND OFTEN VERY APPROPRIATE, YET WE DON'T LEARN A THING FROM IT TODAY.

For example, one area of opportunity includes learning in a systematic way from the off-label use of medications. About 20 percent of off-label prescribing is legal and often very appropriate, yet we don't learn a thing from it today. And another area-and Peter Pronovost [intensive care specialist and healthcare innovation leader Peter J. Pronovost, M.D., whose work with intensive care checklist protocols has become widely known] refers to this as ‘feed-forward’-and that is, getting information way upstream in the care delivery process. So, imagine that a new breakthrough treatment is identified, and is applied to a particular patient and doesn't yield the results expected. In real life today, if that signal gets back to researchers doing the basic research work, it's only very weakly communicated; and really, you'd like to get that signal to the researchers much earlier. So that area of potential innovation includes using evidence in clinical care, and changing the metaphor to one from hunting to farming-growing information, and learning as a byproduct, and taking advantage of the data generated every day from taking care of patients.

HCI: Regarding the data that's essentially discarded now, what could it do for clinicians?

Clancy: It could help us learn from exceptions. Sometimes, for example, a clinician opts out from a clinical protocol or guideline, saying that their patient doesn't meet certain criteria that would make it appropriate to follow a protocol. In such a situation, what I'd like to see happen is for us to learn from that situation, and to use the data from that encounter in the future. In that sense, there is a real potential to change the use of care pathways towards being proactive rather than reactive. One of the very first things you learn in medical school is to take a patient's history and physical. In preparing physicians to do this, you can actually give them a printed-out sheet, based on your organization's performance metrics, that will show them exactly where they're behind in preventive services or in managing patients' diabetes; in other words, this is all about customizing data based on what we know already.

The other really cool potential is being able to identify what's going on with patients over a period of time. Most of the quality measures we use now are snapshots in time, and describe only what happened in an individual encounter. What we really need to know instead is, how is the new arthritis treatment helping a particular patient, Mrs. Smith, specifically, and over time? And so forth. And if a new treatment isn't benefiting individual patients, learning about and understanding those exceptions not only accelerates advances in delivering care, but also can help transform practice. We need to begin to help shorten the translation lag, the time between the introduction of new treatments, and our understanding of their effectiveness. And if I can get a customized sheet-either on paper or electronically-before I see each of my individual patients, that can help improve the care of those patients. Think of it this way: it's analogous to what Amazon does for us as customers of its products.

IF A NEW TREATMENT ISN'T BENEFITING INDIVIDUAL PATIENTS, LEARNING ABOUT AND UNDERSTANDING THOSE EXCEPTIONS NOT ONLY ACCELERATES ADVANCES IN DELIVERING CARE, BUT ALSO CAN HELP TRANSFORM PRACTICE.

For example, at a healthcare organization interested in enrolling people in clinical trials, if you could actually figure out how to let oncology patients know what clinical trials they were eligible for, and the distance from their house-this is not hard, but again, it's more proactive than reactive. And the last game-changer is ultimately reporting patient-reported information. A personal favorite of mine-a project at Denver Health-is trying to improve care for older Hispanics with diabetes. And how they're doing this is through text-messaging reminders to check your weight, get enough exercise, and send us your blood-glucose readings; and you see, they've got the IT infrastructure to deliver this directly into the electronic medical record, as opposed to these data bits ending up on scraps of paper in clinicians' lab coats. And one of the things that patients should be asked to do is to keep a little diary so that we can begin to track precipitating events for blood-glucose level drops, for example, and putting things into the record.

HCI: What's your perception of the pace of IT-facilitated advances taking place right now in care delivery and quality? How far along on the 1,000-mile journey of transforming care quality are we?

Clancy: We're seeing modest, steadily significant improvements in quality, 1 to 2 percent year over year. But when you go out into the field, you find some innovators. And we've created a space for them to share what they've learned, at our AHRQ Health Care Innovations Exchange, which can be found at www.innovations.ahrq.gov. It's kind of cool, because we're trying to create a space where people can share what they've learned. And in terms of development, are we toddlers or unruly adolescents or what? I think we don't really know. There's a leading edge, and I don't know whether it's 2 percent or 10 percent [of patient care organizations nationwide]. But as a healthcare system, we've failed to learn from those leaders, and we need to do better. Many hospitals are doing well in at least one key clinical area. But we don't have enough detailed information yet on precisely how organizations are moving from being transactionally driven to being information-driven.

HCI: Are we becoming more patient-centered, in your view?

Clancy: In very broad strokes, I think the very biggest challenge we have in healthcare is around chronic illness. And in some ways, that's a good place to be, right? Because many years ago, people were dying much more frequently from acute illnesses. But many patients have multiple chronic illnesses; and in many of those illnesses, the only way to learn as much as you need to know about a particular patient is to ask the patient. And that's where information technology can really help us. When I was thinking about accelerating advances in clinical practices, I was thinking about some advances in disease management. The data mining piece is not so hard; getting actionable data to the patient in a timely way is more of a challenge. I mean, some doctors are still using snail mail for really important communications. And one thing that is underemphasized is letting clinicians know regularly about the kind of care they're giving to patients. And when I look at the success of the Keystone project, where they made dramatic improvements in reducing infections in ICUs, it's worth noting that the people in Michigan are still excited, and have maintained those successes.

HCI: Do you think physicians are more engaged now in performance improvement in care quality than they were a few years ago?

Clancy: I think there are more examples to point to now. Physicians have had a substantial role in designing some innovative programs. And doctors are very data-driven. And if they have a role in designing programs, they can become very excited. For example, with regard to a currently operating national surgical quality improvement program, we helped the College of Surgeons expand a program that began in the VA [Department of Veterans Affairs healthcare system], into about 250 civilian hospitals. With recovery act funding, we are providing support to a statewide initiative in Washington state, the Surgical Care Outcomes Project, SCOPE. And they use this surgical checklist that Atul Gawande [healthcare policy and quality expert Atul Gawande, M.D.] wrote about. And I've been told that what they do is phenomenal, but the data collection element has been laborious. We're supporting them in terms of developing IT for this, for translating from manual, dedicated data collection, to a system that draws from and is populated by electronic medical record systems in participating hospitals. It's very, very exciting work.

HCI: What thoughts are uppermost in your mind right now, as you meet with CIOs, CMIOs, and other healthcare IT leaders?

Clancy: The goal is using innovations in health IT to put patients at the center of care. The innovation is in the usability and applicability of projects.