HIStalk Interviews Frank Clark

Frank Clark, PhD is vice president for information technology and chief information officer of Medical University of South Carolina, Charleston, SC.

How is the IT world at MUSC?

I think it’s going well. We have about 1,200 physicians that are our own employees — we have a closed staff model. We started deploying McKesson’s Practice Partner, an ambulatory product we acquired a couple of years ago. It’s been in here a long time. We started in 2004 and we finished it up in 2006. We have it in all of the departments and it’s being used pretty well. I think we still need to do some work with some of the sub-specialists and some of their templates, but that’s going well.

We started rolling out the e-prescribing module. It’s part of that package. We hope to have all of that done by the end of the calendar year. As far as catching clinical data in the outpatient setting, we’re doing a pretty good job.

On the inpatient side, we started in 2006 putting in some clinical documentation and meds administration and CPOE. We’ve got two of the four hospitals finished; two adult hospitals are done. We’re starting on the children’s hospital and we’ll do the psych hospital last. But we’re aggregating all of that data into the Oacis or Emergis Clinical Data Repository and Viewer, which our caregivers really like.

Our strategy is to create an enterprise-wide EMR and not separate the outpatient setting from the inpatient setting, to try to give the caregivers an environment in which they can operate regardless of the care setting, and all the patient information is in one place. They can do it, trend it, look at it, all in one.

I think that the nice thing about the Oacis toolkit is that it gives us the opportunity to make changes and cater things to the caregivers’ liking, unlike some other more fixed systems.

So that’s where we are. We have Telus working on meds reconciliation. We’ll do that out of the Repository/Viewer environment. We’ll do discharge summary and inpatient notes, and they’re working on those pieces as we speak.

So on the McKesson side, you’ve got Horizon Expert Orders, I assume, going to Meds Manager, and then you’ve got Horizon Expert Documentation.

Yes, that’s done. So we have that closed-loop medication process. If you look at our clinical IT environment, I guess the center of the universe is the Oacis Repository and Viewer. We’re using a number of the McKesson products. We use Cerner lab. We use IDX radiology, and of course the Practice Partner functionality, which is a McKesson product. We try to pick and choose fast, specific functionality to capture data in the various care settings, and we aggregate that into the repository.

So that’s all of our strategy. The Oacis toolkit gives us a good bit of flexibility to fill in the gaps with the discharge summary, meds reconciliation, and physician inpatient notes.

How does Oacis tie in with the McKesson parts?

We tried to identify certain pieces of functionality that are appropriate in certain care settings, certain areas like the nursing or clinical documentation. Anywhere nurses deliver care with that service, we want to capture that data electronically, and we’re using McKesson’s clinical documentation to do that. The same thing is true for meds administration — we’re using the McKesson piece — and also CPOE.

The key pieces of the closed-loop medication process come from McKesson, but as far as gathering that data and making it available for caregivers, we’re using the Oacis Repository and clinical results viewer. Given that it is a toolkit, that it is an open system technology, we have quite a bit of expertise, so we can go in and tailor those views.

We just did a really nice view for the ICU, what we call the Critical Care Viewer. It’s a view of data that the ICU docs need to look at. We’re pulling all of that data that’s captured with this task-specific functionality into the viewer. We looked at the Portal, but our caregivers said, “That is a step back.” What we have is much more advanced, much more flexible than McKesson Physician Portal.

I implemented the Portal in 2001 or 2002 when I was in a community-based hospital organization. And those physicians, independent contractors, thought the portal was great. But when you come into an academic medical setting in a closed-staff situation, our physicians said, “What we have is much more advanced.”

You had mentioned to me before that both Duke and Vanderbilt are using their own separate versions of a repository and viewer. Would you say that’s a good compromise between not trying to go off building your own clinical systems and yet having the presentation and data retrieval flexibility that you can get from having this third-party tool?

Absolutely. We don’t have the resources that Duke or Vanderbilt have. It’s kind of ironic. We’re going to spend all day with Bill Stead and his people, trying to fill in some gaps, because our strategy is very similar to that of Vanderbilt’s and also Duke’s. You are right — they use their own home-grown repositories, respectively, but they both are using a number of McKesson products in those task-specific areas. We talk to Vanderbilt probably at least two to three times a week trying to understand how they did some things with the clinical documentation.

You know, I think this would be true of any big vendor. It’s been difficult for McKesson to fully appreciate what we’re trying to do because they are used to the community-based setting where an organization just buys all of their products, like I did when I was in Covenant in Knoxville.

But academic medicine is, as you know, a little different. We were saying, “We don’t need all the functionality of what you might have in clinical documentation or meds administration. We want to pick and choose those pieces that we feel would fit nicely into our setting.” It’s been really difficult for them to understand that.

Of late, we’ve been able to work a McKesson individual who works with both Duke and Vanderbilt, so he understands what it is we are trying to do. Finally, it took us a while to get around to getting McKesson to understand that, but I think we are on track now.

We want to get away from buying these complete systems. Vendors want to sell you a standalone ambulatory electronic medical record. Well, we don’t want it to stand alone — we’re trying to bring two care settings together, because many of our physicians see patients in their clinics, and then of course there’s those patients in the hospital. So we want them to have a longitudinal view of their patients’ data regardless of the care setting. 

That’s our strategy. So far, it seems to be working out OK. It’s difficult for the McKessons and I’m sure it would be for the Cerners and the Siemens, too, because they just want to sell you their stand-alone systems.

The same thing is true in the emergency room. They all have an emergency room product, but it turns out that a lot of that functionality you already bought when you bought the outpatient system or the inpatient system. They just want you to buy a lot of their functionality over and over again.

Meditech is on the low end and Epic on the top end, but people seem to like them for the same reason — they have everything. The territory in the middle is up for grabs.

You’re right. When I look at the big vendors to see who is probably most attractive to closed-staff model organizations like Cleveland Clinic or Mayo and most of the academic centers, it would be Epic, because they do have a repository strategy. A lot of the others don’t. They’re struggling because the relaxation of the Stark provision and the anti-kickback in this pending healthcare reform — I think independent physicians are going to bind themselves more to hospital organizations.

Hospital organizations can offer these independent physicians more systems, some kind of ASP subscriber model electronic medical record, and they can come together on the data-sharing agreements that hospitals will house the physician office clinical data in a single repository.

I think the people like Epic probably have a product that’s appealing if you don’t want to try to fit it together like we’re doing, piece it together like Duke and Vanderbilt, and also there may be some academic centers that’s pursuing the same strategy. But most of the big vendors — you’re correct, they’re kind of struggling because they don’t have that single repository strategy. They’ve got a separate electronic medical record for the outpatient, one for the inpatient, so it’s kind of bifurcated. But I think those two worlds are coming together.

Do you think you’re well positioned with the core clinicals from McKesson, plus Practice Partner, plus the Oacis Portal that would be the equivalent of MyChart, to do what the Epic folks are able to do?

Yes, I think so. I think it would compare very favorably. We’re really pushing McKesson on the Practice Partner side, because in order to do the kinds of things that I mentioned earlier, we’ve got to have their cooperation. I don’t know how much you know about Practice Partner, but we are really pushing them because of the size. We probably have 2,500 users, and when Andy here developed that product, it was designed more for smaller practices.

We’ve really pushed them to try to make it robust. We’ve gone to the Oracle database. We’ve moved on to Unix servers, both for database and application. But in order for us to do meds reconciliation within the Oacis environment, we have to have a bidirectional interface between Oacis and Practice Partner. So we were really challenging them to kind of open up and let us get down into the details of that system to make it work for us.

I would think that they’re open to understanding what you need, knowing that there are potential other sales just like you out there.

I think so. In my understanding, it was one of the biggest selling products in 2008. They have a competing product, it’s called Horizon Ambulatory Care, and maybe they’ve already made the decision, but they have to decide which of those two products they are going to fully integrate into their enterprise release strategy. I think they’ve made the decision, as best as we can discern, that Horizon Ambulatory Care will be the product they will integrate fully into their enterprise releases.

That was disappointing to us because that would have made it a lot easier for us, I think, as we try to do things, trying to closely knit the two care settings together. But in the absence of that, we’re really working with them to try to let us open up the architecture, because in order to do meds reconciliation through Oacis, we have to have that bidirectional interface.

There’s so many legacy products out there that the architecture of the framework doesn’t really lend itself to interoperability. That’s huge. 

Do you think the market’s going to pressure vendors to talk to each other’s systems so that you’re not stuck in your own vendor’s silo? 

I think so. If we’re going to achieve any modicum of success as far as HIEs and exchanging data, it’s got to. But it’s going to be a tough, long battle, I think. When you look at “meaningful use”, wherever that will end up, they can’t set the bar too high, because if they do, nobody is going to be qualified in October 2010 to get any of this in use.

Is HITECH something you’re looking forward to and planning around, or is it just “if it happens, it happens” but it’s not really going to be part of the strategy?

It’s part of our strategy. We’ve been thinking about it since the beginning of the year. We’ve been planning, trying to anticipate what will be the final requirements or the initial requirements for October 2010. I think we will be well positioned. I mentioned we’ll have meds reconciliation, discharge summary, and inpatient notes. We’ll have that done in months, all of those. We’re already doing CPOE, we’re doing outcomes reporting, we’re doing health registries, so I think we’ll qualify both for the physicians and the hospitals for that first round of funding starting in October 2010.

When I talk to a lot of colleagues across the country, both in big hospitals and small hospitals, not that many that are doing physician order entry for the closed-loop medication things.

Have you calculated what’s on the table for you if you qualify for all the HITECH requirements?

The finance people have been doing that. The hospital side doesn’t seem to be that much money in the scheme of things. On the physician side, it looks pretty good. For a physician, I guess they’ll have to choose between the Medicaid and the Medicare; they can’t do both. Hospitals can.

So they’re doing the numbers, and we’ve already made the investments, so it’s not that we’ve got to come up with a bolus of money because we’ve already invested heavily starting in 2004.

Other than those three pieces I’ve mentioned a while ago — meds reconciliation, discharge summary and patient notes — we have all the functionality that we’ll need. Going forward, we’ll continue to design it. But I think that’s the nice thing about Oacis, that it gives us the flexibility to fine tune and do some things that otherwise you’d have to ask the vendor to do. It just takes a long time for them to do things as opposed to having the ability to do some things on your own.

Going back to your question a moment ago, we don’t have the resources that Duke and Vanderbilt have, so we’ve had to do it on the cheap, so to speak, or do it in a less expensive way. We’ve had to buy more pieces and parts, whereas Vanderbilt could probably write their own meds reconciliation functionality. And they’ve done their inpatient notes piece, whereas we would have to contract with Telus Health and Emergis to do that.

If you look down at where your time and energy and concern is going to be placed in the next three to five years, what do you think is going to be important?

We need to finish the inpatient functionality, the children’s hospital and the psych. We’ve got closed-loop medication in all of our inpatient facilities. We’ve got to make a determination if the functionality inherent in the Practice Partner piece is going to be flexible enough long term to fit into our enterprise-wide clinical IT strategy, because as you know, 95% of the care is delivered not in the hospitals, but in the clinics.

We want to be as efficient and as effective in the outpatient setting. For instance, like charge capture — right now, that’s manual. We’re doing charge slips; caregivers are writing out the charge slips. We need to be capturing that electronically. So that’s something we will do over this period that you’ve just alluded to.

We’ve got to be as effective and as efficient in delivering care in the outpatient setting to be competitive. Again, we’ll have to make that determination if the functionality that we’re using out of Practice Partner is flexible enough and robust enough to serve us well long term. That’s going to be a primary focus going forward. Does that make sense?

It makes perfect sense, yes. You never know why vendors acquire ambulatory systems — do they really plan to integrate them, or do they just want to latch on to the trend of the moment? Vendors have to figure out how badly they really want to get in the business of tying in and I guess it’s up to the customers to push them.

I’m not sure if they’re any different from some of the other vendors, with the exception of Epic. I think Epic seems to be well positioned as care settings come together and organizations look at acquiring a clinical strategy or solution that scales across both care settings, and Epic is really attractive.

Most hospital organizations are not going to do what we’re doing, and that is trying to knit it together to some degree, or they don’t have any development capability; it’s all commercial off-the-shelf. I know Pam Pure and her leadership team, right after they acquired Practice Partner, came down to spend a full day looking at how we were using it. And I think at the time they were trying to decide which of those two horses to ride. Should they continue with the development of Horizon Ambulatory Care or should they look at trying to integrate the Practice Partner product into their enterprise strategy? I don’t know, maybe that’s why Pam’s not there any more. I’m not privy to all the details, but they’re not unlike some of the other big spenders; they’re trying, through a lot of acquired systems, to coalesce and integrate them into a common framework, a common platform. This is a slow process.

Anything else you want to share or mention?

These are exciting times in healthcare. I can’t tell you how many calls I get from vendors trying to sell stuff and there’s just so much money out there around healthcare. It’s like flies around honey. There’s just so much money, so many opportunities out there for vendors, particularly in the health information exchange market.

We’ve got a project here that we’re trying to link eight emergency rooms in the Charleston area across four organizations. We call it the ER Alert System. When a patient presents in one of the EDs, a caregiver can query the other hospital organizations to see if there’s any clinical data about that patient. You know, some people are shocked, they just go to the ER to get drugs. We hope that it will cut down in procedures, if somebody’s already done a CT scan or MRI at one of the other facilities, they can access that, or if they have any labs or meds or anything like that.

So we’re looking at technologies for that sort of health information exchange, trying to decide on which technology to use. We’re looking at Oacis technology, but also this Vanderbilt-developed product that Informatics Corporation of America spun off. It seems a lot of businesses, a lot of companies say they have a product that will do that. I think we’ll see a lot more of those health exchanges. I think that the reform of ONC will push to try to make that happen.