Daniel Barchi is SVP/CIO of Carilion Clinic of Roanoke, VA.
You just finished your massive Epic project, with eight hospitals and 100 practices brought live over a couple of years. Tell me about that project.
When we decided that we wanted to integrate from 11 different medical records — 10 electronic and one paper — down to one integrated system, it was 2006. At the time, we were also merging from being a confederated health system into a single Carilion Clinic, so the two projects converged nicely.
We knew if we wanted to operate like an integrated system and have a continuum of care for our patients, we needed one tool to do so. It was an easy choice to select the one integrated EMR.
We ended up choosing Epic because they had a reputation at that time, and I think still do, for implementing well. We knew we had the focus of the entire organization, so we took advantage of it and used it as a tool to integrate all of our hospitals and our practices onto one platform as rapidly as we could, while still protecting the use of clinical data and the health and safety of our patients.
Carilion was a high-profile Siemens client and an early Soarian adopter. What led to the change from Soarian?
Soarian had a great reputation and was doing a lot of innovative things. But when we traveled to Malvern in the summer of 2006 and saw where they were in the development cycle, they were making progress, but they weren’t going to be ready to do everything that we wanted to do as rapidly as wanted to do it.
We knew that we wanted to implement our integrated EMR in every clinical area and work financials in at the same time — front office, back office, hospitals, physician practices — and so it really forced us to find another vendor, which we did relatively quickly.
Soarian has been on the drawing board in some phase of rollout for forever and it still seems like it’s always going to be a year or two more. Do you think it will be a tough job for John Glaser to make Siemens a little more competitive with the window of opportunity that’s out there?
You know, that’s the reputation Siemens has had, and yet it’s a great company. We’re still a big Siemens customer for its many imaging products. If there’s anybody who can make this happen, well, it’s John Glaser. I was alternately surprised and thrilled that he was the one going to take the helm, so I’m happy for him.
I do think, with the right leadership, they’ve got all the right tools and financial backing. There’s certainly a need for more integrated systems. Two or three big companies should not have a monopoly on everything that is out there. The more systems that meet all of the needs of our health systems, the better.
When you looked at systems, what were your overall thoughts?
At the time, in late 2006 to early 2007, we were looking for one that flowed seamlessly. At the same time, I didn’t want our technology team to make the decision. We were adamant that it had to be our clinicians and our financial teams that chose a product that was going to work well, both face-to-face with the patient and in the back office as well.
When we did the selection, we got literally 300-400 doctors, nurses, therapists, and financial folks in the room with some of our IT folks as well and scored them. We went from eight, through a quick cut down to four. We had four vendors on site in front of 200-300 people in a large auditorium. We narrowed that down through the choices of my colleagues in Carilion Health System down to two.
That’s when we had 300-400 people in a room looking at day-long shows from the final two, which were Cerner and Epic. We put a team of about 10 doctors, nurses, therapists, financial folks, and IT folks on the road to do a couple of site visits to see Cerner and Epic in action. Like I mentioned, we chose Epic in the end because they had seemed like a great partner and we were eager to select a vendor that was going to work with us for a rapid implementation.
When you look back at what you chose and where you are today, was Epic something that you were truly excited about compared to your experience with Soarian or was just the best of what your choices were?
We were truly excited. This was not that many years ago — three, three and a half years ago — and certainly, they’ve made a lot of progress even since then, but it was state-of-the-art at that time. In fact, we are still operating all eight systems and about 110 physician practices on the original version that we installed, the Epic 2007 version.
We have plans to upgrade for Meaningful Use purposes next year, but we are very, very happy with the way that the system is operating in all aspects of our health system. At the time, we were very excited about consolidating our many different systems onto one platform. It has certainly met our needs.
Tell me about the structure. You did it all internally — who led the teams? How did you actually set about doing this to get it done on time?
The first thing I’ll say is that you’ve got to have cooperation from all facets of the health system to make it successful. We had great executive leadership and good cooperation with physicians and nurses. While our IT team staffed and managed the project, I made sure that our governing structure was led by our chief medical officer and chief nursing officer.
We were led at an executive steering level by a small team — a CMO, CNO, chief financial officer, the head of our physician practice group, and me. That small executive team ended up making the hard decisions.
Below that, we had our IT team, led by a vice president of clinical information systems who was essentially our vice president of the Epic project team. Her name is Kay Hix. She did a fantastic job organizing the structure and allowing us to use our existing infrastructure to get it done.
Instead of going out and hiring a third party to do it, we decided that we would make our IT team work in tandem with a new Epic team that we set up, and largely within the confines of our existing organization, built a team to develop, train, and build and implement the system.
Underneath Kay, she had two directors, one primarily focused on ambulatory physician practices and one primarily focused on the hospitals. Beneath them was a large team of talented people, including about 35 trainers. We ran training from about 7 a.m. to 11 p.m., six days a week, to train more than 7,500 users.
People say, “Well, Epic just sends out a lot of inexperienced kids who follow the cookbook.” How would you describe how they got involved and contributed to your project?
I’d say the team that we worked with from Epic was top notch in every way. They are very focused on the area that they know well, and almost skill-typed. If you ask an expert in one area about another area, they are quick to get their colleague involved and won’t go out on a limb to guess at what they might not know.
In that way, it seems like Epic does a very nice job of training its people to be subject matter experts. They can have people very deep in knowledge without having to worry about being too wide.
It also works out well with the health system team because we’ve certainly had subject matter experts, whether it’s our OR team or our ED team, who was going from a legacy application that they knew well. They had been trained on Epic, but wanted to interact with an Epic person who was a subject matter expert in that area.
In that way, it seems like Epic has been able to replicate — and continues to replicate — its success with a slightly changing but relatively stable workforce.
Everybody wants to be Epic these days. Can other vendors copy what they’ve done?
I hope so. People ask me, “What’s Epic’s secret sauce?” I often say, if you look at the two biggest players — Cerner and Epic — Epic is Apple and Cerner is Microsoft. Both very talented companies, but each has a unique feel and flavor about them. Even when we were making our selection, Epic felt more university campus-like, while Cerner felt more business-like.
I think that other health information systems can do as good a job as Epic without trying to replicate all of its collegiality. But at the end of the day, I hope that other health systems continue to grow in the way that they are and that we have more systems out there that meet the needs of hospitals and physician practices.
Do you see that happening?
One of the challenges — and I’m amazed every time I go to HIMSS and walk the halls — is all of the small start-ups who think that there is an opportunity to break in at this point. I think if you’re a relatively established big player — one of the big ones already, an Allscripts, a GE, McKesson, Cerner, Epic — there is opportunity to grow and gain market share. I’m happy with John Glaser and his role with Siemens because I think he will make it happen there as well.
Growing from a smaller, unknown vendor at this point into one of the larger players? No, I don’t think so. I think this is a game of musical chairs. Within five years, every large and medium hospital and health system will be seated in the place where they’re going to be for the next 20 or 25 years without much opportunity for anybody smaller to work their way in.
You’re talking about for the major, core systems – correct?
CIOs could be fairly accurately characterized as risk-averse and finance people obviously are. How would you approach the market? People say they want innovation, but nobody seems to want to be the first to buy it in hospitals.
The funny thing is the tools that we have out there at our fingertips have been adopted in such a limited way. I think I saw the fact recently that fewer than 14% of US health systems or hospitals have achieved 10% or greater CPOE to this point. It’s almost funny to be out there demanding more innovative products when we’re not even using the products that are out there well.
I think there is a lot of runway for hospitals and health systems to use the systems that they have, tweak them, and make them more meaningful for their physicians and patients before we go out and try and demand something else. I think that the products that the big players have put out in front of us today should more than satisfy our needs for the next five or 10 years, even without a whole lot of smaller innovation.
What’s held everybody back? Why aren’t they using what they paid for?
It is very difficult. I’m surprised at even the way that some large health systems have achieved what they have. You know, and all of our colleagues do, that these are tough and challenging projects made challenging not only by the fact that they’re very complex, but they involve human lives, so there’s a premium on risk. They involve physicians who are well trained and want to be very efficient and good in what they do and see these tools occasionally as a threat to the way that they operate, and a threat to the way that they care for their patients.
Balancing all those factors is very difficult. Even a well-run project, which on paper has good governance and structure, if it doesn’t balance those needs, and especially if it doesn’t meet the needs of the physicians and the nurses using the system, it’s a recipe for failure.
What are some things that you learned that most people would not pick up on or that you wouldn’t have expected that really made a difference in how your project was completed?
I’d say the factor that made our project a success more than anything else was a buy-in to the schedule. We knew that this was something we needed to accomplish across all of our hospitals and practices. We knew that if we went very slowly it could take many, many years, and that if we were going to achieve the benefits for our doctors, nurses, and patients in a reasonable timeframe, that meant we had to implement in a reasonable timeframe.
That meant we had to make our hard decisions upfront and then stick to them and operate in a system fashion. When we focused on our order sets, for instance, we went from more than 3,500 order sets down to 500 common order sets. It was not 120 order sets for Hospital A and 50 different order sets for Hospital B. It was 500 order sets which you could use, and were the same at every one of the Carilion Clinic hospitals. We did all of the hard work upfront, set a schedule that we said that unless we were going to put patient safety at risk, we were not going to deviate from.
Then, once we had that focus that we knew what we needed to achieve and when we needed to achieve it, we looked for any outliers that would get in the way. As long as everybody was on board — our executive leadership, our clinical leadership, and our project team — we didn’t deviate from the schedule. It was more like riding a train than it was stopping and approaching every new hospital and physician practice and making decisions about it.
For a lot of hospitals, their problem is that their milestones are all wrong. They have to pay more money when they get the code loaded and then pay more money again when they get implemented and go live. It’s almost like an anti-sense of urgency.
That’s a great point. One thing we did was we bought an enterprise license upfront so that we had laid out the capital dollars initially. It was just a question of when we were going to use it, not if we were going to pay more when we got around to using it.
We also front-loaded the project. We did our largest hospital first. We have our largest, 880-bed hospital going to down a smaller 120-bed and even smaller hospitals than that. We decided to start with the 880-bed medical center first because we knew that we would run into the biggest, hardest issues there. Once we solved them there, we would just replicate the same process at our other hospitals.
By getting all of the right people on board, knowing that they had to make the right decisions upfront, and that there was not room for error at a smaller hospital that we could go back and fix later on, we really did have focus and cooperation in a way that I think we would have not had had we started at the other end and worked our way up.
You mentioned Meaningful Use. Are you comfortable with what you think Meaningful Use will be and where you are?
We are comfortable. I’m very interested to see what rules come out. It’s been a fascinating process to watch it all along. I do hope that the standards are held relatively high, but I agree with many of my colleagues who worry that they’re high, almost to the point of a lot of people not being able to participate. It almost seems like it begs for a common ground — that that bar is set high enough that it causes us to achieve more, but not enough that it decapitates anyone.
At this point, I expect that it’s going to be challenging for many hospitals across the United States to achieve it. We’re comfortable with almost all of the elements as they’re laid out today. There are four or five that don’t come easily that we have planned reporting for. We’re prepared, but we’re counting on achieving both certification and Meaningful Use in fiscal 2011.
You mentioned Carilion’s move to the practice-type model, the Mayo model, in 2006. What kind of IT changes did you have to implement to support that?
It was all about integration. We had 11 different systems, more than 512 different interfaces, and we had the challenges of trying to get down to a common way of operating because we really wanted the physician in Roanoke to be able to refer to his or her colleague in Blacksburg, and for both of them to see the same information on a discharged patient from our medical center. We thought we were only going to be able to do that, not by brushing up on our 500-plus interfaces, but by having it all operate on the same system.
One thing that we had done was that we had a whole lot of experience with the GE Centricity product. In the previous eight years, we had rolled out GE Centricity to every one of our 140 physician practices. When we implemented Epic, we knew that we had a responsibility — not only to our patients, but to the doctors that had put all the work of entering and maintaining that information — of having it available at their fingertips the first time they logged in to Epic.
One of the biggest early challenges we had in this project was converting literally eight years’ and about 800,000 patients’ worth of data from GE into Epic. We had a small team led by two of our physicians and about five of our IT people who did nothing for about four months than plan the migration of the data and test it over and over again.
Then, when we actually did push the button and convert the data, our data center literally chugged for about 11 days converting all of that information from the GE system into the Epic system, so when our first practice went live, it was all there. That was a commitment we needed to make so that our physicians had, in their new tool, all of the data that they had in their old tool.
There’s a new Virginia Tech Carilion School of Medicine. How are you involved with that?
We do have a new school of medicine that we’re very proud of. It’s been a challenge. In the past four years, we have implemented Epic. We’ve build 200,000-square-foot new medical center. We’ve acquired two major practices. We’ve acquired one hospital. We’ve built a research institute. We just started the Virginia Tech Carilion School of Medicine. All of which made for a very busy past four or five years.
My team and I are responsible for the technology for the school of medicine. It’s been a fascinating experience helping them stand up their practices, implement IT for them, and put the systems at their fingertips so that they’re ready to go when their first students show up 23 days from now.
One of the most fascinating things was participating in the selection of medical students. The Virginia Tech Carilion School of Medicine used an innovative interview approach where they had many people, lay people and clinicians alike, participate in the student interview process. Getting to do that was a highlight in addition to being the CIO for that school of medicine itself.
In terms of the curriculum for the medical students, are you involved in any IT or informatics training components?
Yes. In fact, we’ve made use of the medical record one of the components of the school of medicine. It’s not something that the students will have Year One, but by the time they get to their clinical rounds, we will have them trained on the Epic electronic medical record and built templates for them to use on their own, in ways that they can step slowly from viewing patient data initially to full CPOE over a period of about six weeks.
The other hat you wear is that you’re involved with the Virginia Information Technologies Agency. What kind of work is being done at the state level?
I’m proud of the way that throughout the Commonwealth of Virginia, we’re cooperating to make sure that the data that we have in each one of our health systems is available to others. Part of the HITECH Act was $2 billion set aside for HIEs. Five of the CIOs of the other large health systems in the state of Virginia and I serve on a Governor’s Commission to help define the standards for the HIE.
We’re also on the advisory board for implementing it. We’ve been meeting in our state capitol at Richmond for the past 12 months focused on how we’re developing HIE, who we will have implement it for us, and how we will begin to exchange data and interchange with the NHIN as well.
What are your IT priorities for the next several years?
One of the things that has been nice about getting to the point where we’re close to wrapping up this implementation is that we know that for the next five years, we’re going to be all about optimization. We don’t want to go into the next big project. We don’t want to go buy the next new piece of cool technology. We want to take what we’ve built and implement it and make it work as well as we can for our physicians.
We have an optimization team, which is getting larger all the time, and that we hope through our upgrade to the 2010 version of Epic next spring that we will bring even more useful technology to our clinicians. Our priority, instead of being very forward-thinking and cutting edge, is all about using this tool that we’ve built to its maximum advantage.
Any concluding thoughts?
It is a heady time to be part of healthcare IT. For the first time in my career, my mother understands what I’m doing because it’s front page in The New York Times. I’ve really enjoyed being part of it. Doing it in a large health system and trying to make it integrated is very rewarding.