Many things are happening these days at Duke University Health System in Durham, N.C. Under the leadership of Asif Ahmad, vice president, diagnostic services and CIO at Duke University Health System, and Kevin Schulman, M.D., professor of medicine and business administration at Duke University, the Duke organization has been moving forward strategically along numerous dimensions. Ahmad and Schulman spoke recently with HCI Editor-in-Chief Mark Hagland regarding their organization’s multi-pronged push forward.
Healthcare Informatics: At Duke, you’ve been working forward on a number of fronts with regard to information technology.
Asif Ahmad: We’re really approaching healthcare information technology more strategically now.
Kevin Schulman, M.D.: On the academic side, it’s been interesting to focus on how to train a generation of leaders with regard to the application of information technology. And organizationally, are we trying to make organizations more efficient? To create more value? To create innovative organizational forms? And how would IT apply to that? We’re looking to answer those questions.
Healthcare Informatics: What is your strategy around clinical IT at Duke?
Ahmad: We like to think for ourselves. And one of the things I’ve seen is that the EMR vendors tend to focus on functional issues such as how you get from point A to point B; they really don’t put intelligence into the system to determine whether you got from point A to point B optimally. So our strategy is either to work with the big EMR vendors to customize their products, or for us to go ahead and put systems around those EMR products. No commercial EMR really focuses on active engagement with the patients, so we’ve created a patient portal. And to build the best practices around clinical care models, you really have to engage the patients. No commercial EMRs right now really optimize patient portals. So we’ve built a very interactive patient portal, which gets data feeds from the EMR system, but really drives patient interaction around wellness.
So our clinical IT strategy is really about driving what Kevin has already said, which is creating clinical value at the bedside or at the side of the patient, even at home; so we’re trying to drive consumer adoption, positive patient outcomes, and positive financial outcomes for the system, by aligning clinical IT in the disparate packages available in the market, around our own systems. And it’s very much focused around healthcare analytics and business intelligence.
Kevin Schulman, M.D.
Schulman: I’ve been involved in a project looking at how we get patients access to their own information. There are two sides to this equation: What’s the right architecture to allow patients to get access to their own information and have a continuity record? And how can we bring the information to them in a portal, as we’re doing at Duke? So we’ve spent the past two years integrating the different clinical data sets, and we’re now integrating them into the clinical data repository, through the portal. And in terms of patient imaging, you come here to get cardiac catheterization and angioplasty, but you go back to your referral site, and then you can’t access the films. We’ve just brought that data repository live.
Ahmad: And again, this was driven out of the need to provide just-in-time communication to the patient and to the provider (whether the provider is at Duke or not), and two, to let the patient own their own information. It still amazes me in 2010 how few organizations believe in giving the patients their information, even though the federal government has a rule on that, in the context of HIPAA. From the IT side, from the EMR vendors to hospitals, they tend to make it virtually impossible to make that happen.
HCI: What is your strategy around handheld mobile devices, especially those being brought in by physicians and other clinicians on their own?
Ahmad: One of our physicians has already integrated remote monitoring onto his iPad. People get hung up on specific devices. But if you architect your technology to be scalable, adaptable, and flexible, and if you don’t make everything proprietary, but instead make it secure, you won’t have problems. We have a virtual patient information network. So if you make the architecture flexible and secure, you can hook up any device. And there are security limitations on iPhones and iPads. But our systems do not leave any data on any device. So the way we do it, you just use your device to access information, and then no data is left on the device. And I’ve been a CIO for over 10 years. But even back in 2000, I never had problems with physicians in terms of their not wanting to bring devices in—they always have, even when the laptops were very clunky. Indeed, physicians have been some of the first adopters of pagers, cell phones, and now smart phones. The question is whether the vendors produce technology that they ike and want to use.
Schulman: If we were improving the efficiency of care in a way that was meaningful to them to begin with, they would have been using these tools a long time ago. Physicians are very mobile. And if you’re logging in and out all day long, that’s not very efficient or convenient.
HCI: Where do you see clinical computing going generally?
Ahmad: Clinical computing needs to follow the world of Google and others; it needs to be very intuitive. Rather than having very rigid architecture, it needs to be supported by very engine-driven architecture. And as soon as you want to go outside a script for some things, it should quickly allow you to have mining capability and quickly get to your next step. We need to architecture systems to be more flexible and less rigid. And clinical computing should become very proprietary architecture-independent; and it’s slowly getting there. What needs to change is people constantly having to throw away old stuff, old legacy systems. And you now need to move towards aggregation, not just at your own patients. You need to look at your individual patient in comparison to other patients on your panel and beyond it. So clinical computing needs to have health analytics built into it, to provide you with better understanding of patient populations. You’re seeing some of those models being tested outside the U.S., but they will also come into the U.S. over time.
Schulman: And clinical computing will begin to substitute and begin to eliminate static systems. Kaiser [Permanente] has a system in place that allows for the PHR, patient support, and e-consults, virtual visits. Things like prevention: Why can’t I check my blood pressure at home and check a little visual graphic? Why should I have to drive and pay for parking for a very small interaction?
HCI: Where are the physicians right now in relation to clinical computing and working with IT and clinical informaticists?
Schulman: Clinicians are very diverse; we seem to have no problem adopting technology in different realms. And you think about society and age brackets, and you could make some generalizations about people now in their 50s. But I personally started playing with technology in high school. And my wife, the least technology-friendly person in our household, has a BlackBerry, and e-mails all day long. We’re surrounded by technology now in our lives. I would say there’s a business model transition taking place here. I’m on the faculty of the business school, so I keep focusing on business models. And here’s the Apple business model—spend $1 billion on research and development, and we’ll disseminate it at low cost. But for some reason for both hospitals and doctors, we had this model that every one of them had to develop their own technology. Obviously, that’s not a scalable model; but it also leads to a quality of product that’s way below what it should be. And I think vendors that can’t build truly scalable, fleible models, will disappear.
Ahmad: I agree, but I’ll go one step further. I just think the whole notion that there’s any generational device is ridiculous. I think you will see more grandparents online Skyping their grandkids than you see young kids doing it. And that’s because the good technology is so intuitive that no one will be scared of doing it. And I think some CIOs in healthcare have helped to create a myth about users. It’s like, ‘Oh, my physicians are so hard!’ But that’s not true. The problem we have in this industry is that most CIOs don’t connect with their physicians; that’s really one of the barriers. I mean, the iPhone is so intuitive—my three-year-old can figure out how to use it. So the problem in the CIO realm and in the IT vendors’ realm is that they’ve never taken the time to connect and figure out how to improve the clinical process; so they haven’t really linked in with the physicians. So no wonder the products are so horrendous and why the product evaluation processes and RFP processes are so flawed.
But that’s beginning to change. And that’s why in my deployments at Duke and previously at Ohio State, we’ve been so successful, because the products have been developed and led by physicians, and then deployed in a way that made sense. So we had one of the most successful CPOE deployments—we were live within Duke University Medical Center within nine months, and including the pilot, within 18 months. And it’s the same technology that McKesson [the Alpharetta, Ga.-based McKesson Corporation] has, but we were able to drive the content so well; and it has everything to do with how intuitive the interface was and how rich the content was, pertaining to each clinical service and department. And that’s something we did six years ago. So the point is that you don’t need breakthrough advances in technology alone; certainly a billion-dollar R&D investment into a new product will get you there. But you need also to reengineer the logistics of existing systems. So it’s a matter really of using existing technology to move us towards clinical functionality and better processes and better-driven outcomes around the patient. When you don’t adhere to such principles, you end up disconnected from the physicians and patients. What’s more, nobody understands what healthcare IT is. That’s why Kevin and I have launched a new master of management science in clinical informatics, out of our business school, Fuqua, and it’s a one-year degree focused on understanding the applied impact of medical informatics in healthcare.
Schulman: This speaks to the concept that Asif and I—he really argued that people had to do clinical informatics, applied, and not theoretical informatics. And again, they need to articulate processes, business concepts, or business strategies. So in a one-year program, we give people seven management and five informatics classes.
Ahmad: And they are very applied informatics classes. And the premise of this whole degree is, why should one do health IT? What’s the ROI one is trying to achieve? This is about preparing the next generation of leaders who can understand how powerful this technology is.