HIStalk Interviews Denis Baker, VP/CIO, Sarasota Memorial Hospital

One of Denis Baker's employees e-mailed and said I had to interview him, including a long list of reasons she enjoys working for him. I knew of Denis mostly because of Sarasota's work with Eclipsys and CPOE and was happy to visit with him by telephone.

Tell me a little bit about yourself and about your job.

I started in healthcare during Vietnam as a Navy corpsman, thinking that I would end up on a ship or a Navy base someplace. Then, out of total ignorance, I realized that I was probably going to end up in the Marine Corps, which I did for short while, but fortunately stayed out of Vietnam.

I got into laboratory medicine and then became a med tech. I worked in a hospital in Portland, Oregon for ten years, in a chemistry department. They were implementing their self-developed lab system. So I got involved interfacing all of the instrumentation to the computer system. This would have been early eighties. Then they thought the computer system was going to support itself, so they were going send me back to the bench. So I jumped to another organization and supported an HBOC Star lab system for a couple of years there.

I ended up being the manager of the clinical systems. Then a new CIO came into the organization and created a new position of Office Automation and End-User Computing Manager, which is the worst title in the world that I can think of. Was it meant, basically, was  supporting PCs. It was a four-hospital system back then. They had no centralized support. So I pulled together a good support group for training around PCs.

I left the organization in 1991 and then ended up working for an outsourcing company that has since disappeared, moved to Cincinnati, and worked as a director of IT at one of the suburban hospitals there. Then, ultimately became the CIO for the four suburban hospitals who were part of the system. The whole consulting company crashed. At about that time, the CIO that I worked with in Portland, Jim Turnbull, had since moved down here to Florida and there was a Director of IT position. So I moved down here in 1995. I was the Director of IS for five years. Jim left in 2000 to go to Denver Children’s. I was promoted to CIO.

Your background is as a clinical department end user. Do you think that's a good background for a CIO to have? 

I think it so, because I looked around early on at who the early CIOs were. It seemed to be most of them were promoted directly out of IT and really didn’t have any exposure to the clinical world. I think that has really helped me as the whole shift in the industry is gone to clinical information systems. I can talk the lingo with not only lab folks, but also nurses and other clinical folks. Physicians as well. I think that’s been a leg up.

I think the future for healthcare CIOs in particular is to have a clinical background, whether that’s nursing or one of the ancillaries. I think you really need to understand what happens in a hospital, not just producing bills.

Should the ultimate goal be to have a physician running IT or does it really require that?

I think that physicians bring a certain aspect to the job, but I don’t think they necessarily know how a hospital works. I think they know how their practice works and how they interact with the hospital, but I don’t think they absolutely know what nursing does, or any of the ancillary departments, and what they do.

What do you like most and least about being a CIO?

Most is seeing technology applied to operational improvement in the organization and moving healthcare out of the dark ages. We’ve been on paper for a hundred years and many people have said there are industries that are far beyond us in adoption of IT. With good systems and good implementations, it’s remarkable what you can change within healthcare with IT.

Do you think clinical systems are realizing their potential, or are those systems still a generation away?

My chief medical officer asked a similar question a couple of days ago. He compared it to the automobile industry, where the tires might last for five miles and then you’d have to replace them. He thought that maybe we’d moved now into maybe the thirties or forties as far as automobile technology in comparison.

But I think we’re really in the fifties. I think the systems work, but they don’t have all of the bells and whistles that the current car today would have. It’s going to be an evolving process. We’ve been at this, with the clinical information system, for twelve years and it seems like the work is never done. There’s always something changing and something new. New functionality or, typically, some new regulation or reporting that has to be generated out of your clinical system, so the work never stops.

You never stop implementing a clinical system once you start. That, I guess, one of the downsides. You asked me the pluses and the minuses. I think the downside is, it’d be nice to wrap up a project and move onto another one, but it just never goes away.

Nobody can afford to replace those systems every few years. How important is it for the CIO to establish a relationship with a vendor and stick with them?

Let me start from day one, with negotiations with the vendor. Obviously the vendor’s interested in sales, market value, and stock price, but I think you need to reach a common ground on what you’re trying to achieve. I’m not a big one for really tough, upfront negotiations. I don’t try to nickel and dime them, but I do want performance guarantees. I do want access to senior management. I do want them listening to us as a customer base as to where their systems need to go, and hopefully they’re listening to us. Because you’re right, this system we’ve had for twelve years -- I would not want to be here to be the one to replace it. I would not want to go through that agony again.

A lot of places just trade Vendor A for Vendor B while the hospital down the street is trading Vendor B for Vendor A.

Once again, I think it comes down to personalities, and if you can develop a relationship with your vendor at a personal level and not slam your fist on the table every time there’s a hiccup, but thank them for when they help you solve problems ... I think that’s the key to the partnership. I think some people get caught up in egos. They’ve negotiated some super-duper contract and the vendor, for whatever reason, is unable to deliver, and potentially the CIO or whoever negotiated is being held accountable for making it happen. So the knee-jerk reaction is, "Let’s go find Vendor B. That’ll work out much better than Vendor A did."

How much are hospital executives involved in IT decisions?

Well, here, they’re pretty involved. The Chief Operating Officer, my new CEO that’s been here for about two and a half years – both of them have been involved in some pretty major IT implementations where they came from. I think they have a good understanding of technology and what the limits of technology are. Their caution to me and the organization is, "Just because you’ve got a problem doesn’t mean IT necessarily needs to solve it. We need to focus on our workflow and the operational improvement." And then if there’s an opportunity for IT to get involved and ease that along, that’s probably the best course to take. I think all of my peers within the VP ranks have that same understanding.

How often is IT part of the strategic solution?

It's hard for me to gauge at this point. We’ve had a new CEO for two and half years. Our Chief Operating Officer, he’s relatively new as well. I think the last thing they look for is an IT solution, but we’re still going through our version of Six Sigma or Lean Management or Process Control Management, whatever you want to call it. And then you take a look at, "OK,is there an opportunity for IT to get involved and help solve that problem?" So as we have a fairly new executive team, I think we’re still working our way through that.

Every CIO wants to run a world class IT operation, but hospitals don’t usually have large IT budgets. How do you choose your battles and stretch your dollars?

Number one, I’ve got a great staff. I have about a hundred people on staff. Being in Sarasota, Florida, it's fairly easy to recruit good talent to this part of the world. We pay well. So I think that’s the first key, I guess, to making it a success.

I can only think of one project in twelve years that we’ve backed out of. Some of the projects may have taken us a little bit longer or cost us a little bit more than what we thought, but we’ve only had one complete failure in ten or twelve years, out of I don’t know how many projects we’ve been involved in. I think we’ve developed a reputation, as a division, for getting things done on time and relatively on budget. That lends some credibility, not only among my peers at the executive level, but also with staff; and also, even more importantly, with the physicians. If you get them involved in something and it turns out to be a success, you get less and less resistance as you move into other things to implement.,

If you look back two or three years, what projects gave you a lot of bang for the buck or made you glad you did them?

Probably the first one would have been what’s now the Eclipsys Sunrise Clinical Manger. It was created by a company called HealthVision, then called CareVision, the product. We were the first customer. We started to roll it out in 1998 to deal with nursing documentation and physician order entry.

At some point, the voluntary CPOE hovered at about 25%, and finally, nursing got tired of having to deal with the paper and electronic world. Our elected board then told our physicians, "OK, a year from now, it's mandatory that you put your orders in." Almost immediately, we saw the percentage starting to rise. We run probably about 80% entered by physicians, 10% verbal, and another 10% written or faxed in.

I think the whole CPOE and at least the beginning of the medication order process of transcription illegibility and so on - that went completely away. It created other problems, but at least it solved the illegibility and who actually ordered something.

Another project that took us a few years, but I think was ultimately a good decision ... we needed an ERP system. We looked at Lawson and PeopleSoft and ultimately decided on PeopleSoft. That product has been rock-solid ever since we implemented it, even after the Oracle acquisition. In fact, we’re going through an upgrade to the HR side of this system right now. We had to engage some fairly expensive consultants to help us get it implemented and augment our staff. But I don’t lie awake at night worrying about PeopleSoft at all. It really helped with supply chain management, on the one side, and then we also had some issues with HR and payroll on the other. I think Peoplesoft solved both of those.

Anything on the infrastructure side that turned out to be a good investment of time and money?

Early on, as we implemented the electronic medical record, we were looking for a fairly robust network infrastructure. At the time, about the only thing was available was a technology called ATM. Implementation was good. It provided campus-wide network backbone capability up to gigabit speed. That served us well for a few years, until Cisco and the rest of the world got Ethernet up to speed.

Since then, we’ve gone with Cisco and that’s been rock-solid for us. Built in an awful lot of redundancy to make sure that the network never goes down and, knock on wood, it never goes down. Early on, the intent with the electronic medical record was to maintain all of the records on everybody forever. So we made an early investment with EMC and their technology. This would have been back in 1996 or 1997. We’ve been with EMC ever since. So from a storage perspective, expandability, once again, that’s worked out very well for us.

You made CPOE mandatory in 2003. What advice would you have for hospitals considering doing the same thing?

I‘ve looked at some of the organizations that have tried the voluntary route, but I think you can only do that for a certain period of time before you have to make it mandatory. Like I said, we coasted along for a few years with a 25% compliance and that was driving our nurses nuts. Having to check not only the paper chart, but the electronic chart for recent orders and so on. That’s going to cause quality and safety issues. So at some point, if you’re not making it with the voluntary, I think you better go mandatory.

You’re not going win them over with technology. They’re always going to complain about the time it takes to log on and how much longer it takes to place an order. But after a period of time, in our case ten years, they can look at patient information back to 1998. There’s nothing archived. Everything’s available and I really think they see that as a value in exchange for the whole CPOE piece, but it takes a while for you to build up that database for them to appreciate that.

Are you seeing any impact of the Stark relaxation and are you doing anything with physician office computing?

In a very minor way. We’ve had Siemens' PACS system since 1996. I think three to four years ago, we implemented Siemens Magic Web, which is the online retrieval of images. They were able to do that within the organization. Obviously, they came to us and said , "OK, we want to see those images in our offices, and by the way, we don’t want to buy any equipment." 

So we were able to seed a few workstations out into some of our specialty physician offices; orthopedic surgeons and so on. That’s all they can do with them, look at our images. They can’t load it up with games and other stuff.

Physicians have the expectation that the hospital should provide them with an office EMR. We’re trying to figure out if our direction should be in that area. Obviously we’d have to charge some nominal fee. On the other hand, as my CEO reminds me, there’s a whole host of other companies out there like eClinicalWorks and so on that are offering ASP models that have relatively reasonable prices. They offer not only EMR, but also practice management.

So why, as an organization, should we get involved in that? The only challenge I have to that is that it would be nice to be able to have longitudinal medical history on our patients, whether they’re seen in an office or in the hospitals. I’m not sure how well some of those ASP offerings could be integrated into what we’ve got. So we’re kind of exploring that right now.

Are you seeing any impact of interoperability?

We’re the only not-for-profit hospital in four counties, surrounded by HCA, Universal, and HMA organizations,and they really have no desire to exchange data. I don’t think it makes sense, from a corporate perspective, for them to get into that. So we’ve really not been too successful in creating a real environment here locally.

I tried to get some money out of the State of Florida. Jed Bush budgeted $10 million to get RHIOs off the ground. I made an application and one of the requirements was that it had to be with a competitor. I tried to make the argument that, in some cases, our physicians are competitors, but the state wasn’t going for that, so I didn’t get any of that money.

How would you say Sunrise is working compared to a year or two ago?

We did the 4.0 upgrade probably close to two years ago. That was probably some of the worst software I’ve ever seen. It took us probably eight months and I don’t know how many hundred patches and service packs to get all of that fixed. But finally, everything settled down and the performance came back.

Two months ago, we did the 4.5 upgrade. That’s was probably the easiest upgrade that we’ve ever experienced. That was real quality software. I think you could see the impact of John Gomez and his development team on the quality of the software they’ve produced.

With Andy Eckhert involved, do you think the direction of the company or its likelihood to success has changed?

Yes. Andy made a few visits here since we were one of the early adopters and I’ve liked the changes he’s made in the company. I’m not sure how successful offshore development is. I’ve never dealt with a vendor who has really relied on that quite a bit. I know they’re expanding their office in India to four or five hundred developers. So hopefully we’ll see, once again, a continued emphasis on quality software when that’s released.

Some of the other changes he’s made is decreasing sales staff and so on, and focusing more on support and development folks. The consultants that we’ve had involved in the 4.5 upgrade ... the quality of the individuals, I think, has risen dramatically as well. As I understand it, they have to go through a three-month boot camp to learn the system before they’re ever turned loose on the customer base. I can remember years ago when a new hire would get hired on Friday and be assigned to us to fly in on Monday, knowing little to nothing about the system. They were just here as a body filler. But, like I said, the quality of individuals we’re dealing with now is much better.

Their future success in a having a broad clinical offering like the market wants is based on making Sunrise Pharmacy work. What are your thoughts on that?

I always thought that pharmacy really needed to be, not an interfaced system, but integral to the whole order entry process. Because they didn’t have that product five years ago, when we needed a pharmacy system, we went with McKesson’s Horizon Meds Manager. We had some transition issues with McKesson. So we implemented their system; we interfaced it with a bi-directional interface. That has its own uniqueness and causes its own problems. Now that Eclipsys has a pharmacy component, we’re going through an evaluation of, "OK, where’s McKesson right now? How would their new Meds Manager and Admin-RX compare to an integrated pharmacy module with Eclipsys?" So we’re going through that process this week, comparing and contrasting that.

One of the things I remembered about the hospital is you were one of the first, if not the first, to offer a turnaround time guaranteed time in the ED. Were there technology implications to that strategy?

Actually, no. That was all workflow. A new CEO came in two and a half years ago from Detroit. At least a couple of hospitals had implemented the thirty-minute guarantee. She walked in the door and said, "OK, we’re going to do that here" and turned to the ED folks and said, "Make it happen. Figure it out." And it really had nothing to do with technology. It was all workflow and handoffs.

Now, somewhat after the thirty-minute guarantee was in place, we purchased the Eclipsys ED module, displaced boards and all that other stuff in there, and I think that helped. Now we’re on ED doing nursing documentation. And then finally, ten years later, asking the ED physicians to do order entry. Back in 1998, they screamed bloody murder, so we started someplace else. So it’s taken us ten years to get back to them. I’m anxious to get that piece wrapped up. But no, the thirty-minute guarantee had nothing to do with technology.

Tell me more about your department.

There’s actually three departments that report to me. I had more at one point, but right now I’ve got Information Systems, which is the pure technology stuff: the servers, the network, PCs, and all of that. There’s about sixty people there. The original project team that implemented SCM has been maintained as a separate department. They used to be all clinicians, with nurses, pharmacists, radiology techs, whatever. Perhaps less so today, but I wanted to maintain a real emphasis that there was a support department called clinical systems. It was responsible for, not only SCM, but now they’ve taken on the rest of the world: radiology, pharmacy, laboratory, all the ancillary systems as well. Their focus is more on the application side, with the IT department really worrying about the infrastructure piece.

We’re about 3.1 or 3.2% of the operating budget. Our routine capital is about $30 million a year and typically we get $5 to $7 million of that, This year, we’ve got $7 million, which is about a quarter of it. In fact, that was one of the attractions when I came down here. I came from an organization whose IT capital budget for four hospitals might be $1 million. When I came down here, my predecessor Jim Turnbull had gone through a planning process and gotten a commitment from the board to spend $50 to $60 million over seven or eight years. So that was a big attraction -- being able to do things without scrimping on the basics. And I’ve been able to maintain that capital commitment board and administration. This is my third CEO. I’ve been able to continue the capital investment in IT for the last seven years since I’ve been CIO. So I feel pretty good about that.

With a large amount of money being invested, how do you decide where to spend it and how to justify the ROI that results?

I think I’ve been fortunate. We’ve really never been an ROI organization, which I appreciated as well when I first walked in here. It's been focused more on what are the problems that the organization needs to solve. What’s the solution to it? How much does it cost? And then it goes into the budget.

I don’t have an IT steering committee. My IT steering committee is my CEO and she can be very direct at times. We had a JCAHO survey a couple of years ago. We ran into a couple of situations that IT could solve and she said, "Go make it happen." And the real focus over the last eighteen months, if not more, has been on quality and safety. Now that we’re doing CPOE and eliminated the upfront transcription errors, how do we solve the problem of wrong meds, doses, and all that on the back end. That’s why we’re really focused on the barcode administration piece right now.

Are you worried that vendors seem to be moving toward hiring inexperienced employees right out of college?

I can’t say that I’ve seen that within Eclipsys. Most of the people that I’ve interacted with, all the way from implementation consultants to project managers, these people have got a number of years of experience behind them. I’ve seen the comments about Epic and the implementations and so on, but I cannot say I’ve ever seen that with Eclipsys. There always seems to be a requirement that either they have a clinical background and know something about how the department operates. And then they get educated in IT. And as I reflect on our original project team for SCM, that’s the approach we took. We attracted the best and brightest clinicians in the organizations and then took them through the IT training piece. That worked out very well for us. I think it would be very difficult to take some computer science graduate that just got out of school and teach them how a hospital works without a whole lot of supervision and good mentorship and/or project management.

Then we get into my concern about a company that is publicly traded is having to pay attention to what’s going on in Wall Street, and try to come up with, may be not the best model, but the most economical model, and hope that it actually works. I wish there were more healthcare IT companies that were privately owned. I see Wall Street as a huge distraction. A good example -- I don’t know if you remember Transition Systems Inc.?

Yes. Eclipsys bought their decision support.

They missed the mark on one quarter. Their stock price dropped and then they got scooped up by Eclipsys. At the core, I think TSI was a good company. I think they had a Cadillac of decision support systems at that time. Through acquisition, good talent left.

That’s what I get tired of -- the mergers and acquisitions. When we were looking for radiology systems, Siemens had a partnership with IDX at the time. They didn’t have their own good solution. So we went with IDX for radiology. Then that faded after about two to three years. IDX went to GE, and I can’t say I’ve seen a GE rep in the last two years since the acquisition. So this whole vendor churn and having vendors figure out how the new products that they’ve just acquired are going to integrate with what they’ve got seems like a huge distraction, not only on the front end of the acquisition, but on the back end on how are you going to make this stuff work.

What technologies do you see on the horizon?

I’m not sure I’ve got any original thoughts. I know there’s a lot of negative bias against it, but it occurred to me after Katrina, the paper records in New Orleans were gone and the only organization that seemingly did well at recovery was the VA. They took their backup tapes from their data center in New Orleans to Houston and, within a week, everybody in the nation had access to those records. I was trying to think, since I’m in the potential path of a hurricane as well, what would we do?

The whole idea with smart cards appealed to me. Downloading the CCR from our inpatient systems; providing read-write devices to our physician offices so they can populate it as well. The card isn’t so much the issue. The opportunity is having a redundant data center in Dallas or someplace where all the data is stored. But from a smart card perspective, not only has the core clinical data on that card to be read any place, its also available on some website somewhere. It provides a marketing opportunity for us with our logo all over the face of it. And then from an efficiency point of view, them walking in with their card, we swipe them, they’re registered, and they are done. Then they can go on to their appointment.

One of the issues that I’m not sure is unique to us is the length of time to identify the right patient, get them registered, and double check the insurance information. I believe that smart cards would solve that. Some of the discussions I’ve heard is, "Well, we should be downloading that to people’s cell phones." Somebody’s always looking for the next technology and we’re really focused on trying to do smart cards this year, but we’ll see how well I do. It’s kind of a data concept, but I think it’s potentially could solve three problems for us.

One of your employees e-mailed me to suggest that I interview you and said, "As long as Denis is the big guy, I will work at SMH." How do you command that kind of loyalty?

I’m honest with them, sometimes to the point of probably saying things that maybe I shouldn’t. Like most larger organizations, there’s rumor mills all over the place and I want to make sure that my folks hear from me what I think is going on and what the organization is actually doing. So I think, honesty and also being upfront and fair. We’ve had certain situations with employees that have not been popular decisions. So when I go back to explain, to the degree that I can, what the situation was and why that individual no longer works here, they appreciate the fact that I made the right decision. They understand it.

I give them quite a bit of latitude into the decision-making, particularly to my management group. An idea will be thrown out on the table, we’ll talk about it, and sometimes I’ve overridden the consensus decision from the management group, and I’ve tried to explain why I made that decision. I’ve had very little disgruntlement because of that.

Who do you admire in the industry?

I would say John Glaser at Partners. He was way ahead of his time when they started writing their own MUMPS software in, I think, 1988. They’ve always been ahead of the curve as far as development of their clinical systems and the fact that they self-develop them. They’ve got a staff of six hundred or something like that, but to take something massive like that on and be that successful at that large an organization is remarkable.

The same employee that e-mailed me that said that you’re a faithful HIStalk reader. Why is that?

I appreciate the insight. You’re one of my twice-weekly reads and the Brev-It e-mails as well. It gives me an insight into stuff that typically wouldn’t be available to me regarding acquisitions of vendors. Sometimes the rumors are interesting as well. I appreciate the fact that you wait for secondary validation that its true. It's well written. I think you cover the industry pretty well. Obviously I think you have the trust of your readership. It’s a good read. I guess the other piece that I appreciate is that but you’ve always got the link. The article allows me to go out and find out more about it, so I don’t have to go someplace else.

Is there anything that you wanted to talk about?

I just received the invitation for the Most Wired survey again. I wish somebody would kill that. I’ve seen your comments. I share your sentiments about it.

I’ve talked to some of my peers that have been on the Most Wired list and asked them if they’re really doing some of that, and they said, "Of course not." So I think somebody needs to audit some of this and put this to rest. Fortunately my CEO doesn’t have a whole lot of belief in it either, so she’s not holding me accountable to what some of the other organizations are doing. Not that there aren’t some good, innovative things going on out there, but having an unaudited survey of what you’re doing ... the polling results are in from New Hampshire. Everybody thought they had the pulse on what they thought was going happen and then it changed overnight. So in that case, the pollsters were throwing out the numbers, but the voters really showed up and indicated what reality was. So I wish somebody would do that with the whole Most Wired survey as well.