HIStalk Interviews Beau Raymond, MD, CMIO, East Jefferson General Hospital

Sidney “Beau” Raymond, MD is VP/CMIO at East Jefferson General Hospital of Metairie, LA.

Tell me about yourself and your job.

I’m internal medicine trained. I did LSU for my medical residency. I initially went to a multi-specialty clinic that went bankrupt because they didn’t know how to transition from capitated payments to fee-for-service. I went out on my own, was independent for a while, another physician joined me, and then I was recruited by the hospital here to become employed and become medical director for the employed positions. With that role, I became more involved with the IT aspect of things as well as other tasks and items and got more and more involved in the administration side, became CMIO, and now I am almost fully administration, but still doing some clinic work.

What are the major systems used at East Jefferson?

We use Cerner for inpatient. We were using the same platform in the ambulatory setting as well. Due to some difficulties in getting some information out of the system for quality reporting to outside payers, we felt the need to change to a different system in the ambulatory setting, so we just recently switched to GE Centricity. We’re in our second week of go-live. Being on the same platform had benefits as well as some problems, and some of our problems were getting really good information out of the system.

I haven’t heard of many health systems that had the same systems for both inpatient and outpatient and then replaced the ambulatory one. What was it about Centricity that was appealing?

One of the problems that we had with Cerner is when we initially had it – which we’ve had it for a while — is that you had the customization. You were given lots of choices as to where you wanted to go and how you wanted to do things. Because of that, some choices were made at that time which then affected things later on because it wasn’t standardized back then. They now have become more standardized with fewer options. We were part of that learning process.

Flaws that were made way back when – that’s why we had to make some adjustments. That was the drive. GE has a very standardized database, which allows us to get information out of it to outside payers and outside resources that we need to get that information to in an easily standardized fashion.

You mentioned that your previous practice had challenges trying to survive in a value-based payment model. Will having two systems present a challenge in that kind of world?

Let me clarify. The group that I was with before was in a highly capitated system. When I joined them, which was 15 years ago, it switched from that capitated system of the HMO world to fee-for-service. They were so entrenched on how to manage capitated care, which they were doing well with, they couldn’t make the switch to fee-for-service. That was a little bit different than what we’re doing.

It’s interesting that they failed in trying to move from value-based care to fee-for-service. 

Yes, that was 15 years ago, but now the pendulum has swung back around and we’re back into value-based care. Our thing is that we want to be sure that we get the information out there as to what we’re doing and make sure that that information gets there. Because some of the problems we’re having when we deal with the Humanas and Blue Crosses out there, they look at our information, they’re not able to get all the data that’s really there. For us to be scored properly and reimbursed properly, we need to make sure they’re getting the information correctly.

You probably answered my question by saying you just went live on ambulatory with Centricity, but what projects are taking up most of your time?

That’s in the past two weeks. Actually more recently than that, inside the hospital, we went live on Dynamic Documentation from Cerner. We switched from paper progress notes to electronic progress notes. A lot of the other stuff was already being done via dictation anyway. People could still write a consult if they wanted to, but most people were dictating them. We made that switch to Dynamic Doc, which went extremely well. We had postponed it for nearly a year because of some issues that we were finding regarding how Nuance interacted with the system and with Citrix and all sorts of other interactions that were going on.

Once we got all those things to the level that they needed to be, it was working very smoothly. The first weekend of go-live, I was here walking around campus over and over and over again and was very pleasantly surprised as how well it went. Did the same thing for the two weeks, just constantly walking around talking to physicians, and they were very happy with the system. It was a very smooth transition to a fully electronic documentation.

What are you doing with population health management, patient portals, and patient-facing technologies?

We are engaged with a company that was Medseek that is now Influence Health. We’re going to have a portal that has information from both the Centricity side and from the Cerner side, all in one area. The patient can access all that information at one spot. That hasn’t actually gone completely live yet. It’s finishing up development right now because they needed Centricity to be up and running more. That should be going live soon as a single portal for both sides, ambulatory and inpatient.

On that note, that is going to be a way for us to do outreach regarding patient engagement as well as scheduling that they’ll be able to do directly inside the portal. Also possibly doing population management using that tool as well – possibly. We’re evaluating that now. We also have a clinical integration network called Gulf South Quality Network, which also is engaged in population management, especially with Blue Cross at this time. We have a new tool we’re implementing with them as well regarding trying to get that information. But linking to numerous EMRs is difficult, so it’s taking a little bit longer than we’d like it to.

The direction is value-based care, yet most organizations still get most of their revenue from traditional fee-for-service. Are you finding it difficult to try to straddle those two worlds with your technology?

Yes, well, I will tell you that we have a significant amount of fee-for-service. Actually here at East Jefferson, we have a significant Medicare population around here, but we have a Blue Cross contract which has a shared savings component to it. We think we’re doing well according to the reports that we have. That final analysis, I believe, is in August, so we’ll see where those numbers are. But what we’ve found so far is that we’re doing pretty well with that. We’re heading in the right direction.

It is very difficult because of the fact that it is only one real contract that is value based. We have some others regarding Humana and some others that have some aspect to it, but not nearly to the extent as the Blue Cross one. Technology-wise, as well as contracting-wise with physicians and compensation for physicians for those that are employed, it’s just a difficult transition because you can’t go too far too fast. When you’re stuck in the middle, it tends to be problematic, which is where I feel like we’re a little bit there now. I would love for the rest of the payers to switch to value-based all at once. It would make my life a whole lot easier.

What are you doing with MedCPU?

MedCPU is a rather neat program as to how it works. It looks at all the data within Cerner historically on a patient, has all that information, gets it processed on the back end. It takes that information, sends it over to their processors, and depending on what it finds, it sends us information back if it falls within a certain algorithm. We’ve already gone live with them on VTE as well as stroke.

Our VTE numbers, when we looked at them, were not nearly where we wanted them to be. That’s why we got engaged with MedCPU. We thought it would be an innovative way to deal with that. It has helped us tremendously. We went from some of the measures being in the 60 percent, 70 percent, up into the upper 90s now.  We’re not quite to 100 percent, but that is our goal. We’re heading in the right direction regarding that. It has been rather dramatic as to how well it’s worked for us.

For stroke, our numbers have been very good. We’re stroke certified, so that’s been working well for us. We are piloting with them a product that Merck developed with MedCPU which looks at sepsis, UTI, and pneumonia. Basically it gives you an alert when somebody hits SIRS criteria – systemic inflammatory response syndrome — and lets you know that there are certain tests you may want to order. So far in this first few weeks, it’s averaging about six alerts a day. Most of them are pretty valid and are acted upon appropriately. Some have had to be tweaked a little bit.

It’s going to be interesting to see what tweaks need to be done going forward, because if you think about it, SIRS criteria was probably created with someone looking back at a chart over a time frame rather than having real-time information, which is what we have now. We may be acting a little bit earlier than we really need to sometimes.

There have been episodes where it’s a post-surgical patient and they had a mild fever, which triggered them to get into the right criteria. They had all their factors, of course, to actually hit SIRS criteria, but with that, the physician said, let’s give it a moment and see what the next temperature is. The next temperature came back fine, so the patient was no longer SIRS criteria. It’s one of the things where I think we may have to tweak it somewhat. Right now we’re following SIRS criteria pretty strictly, but we may have to tweak it somewhat and say that they have to have it for maybe four hours or something along those lines, but we’ll see where that goes. It’s a pretty effective tool so far as what we’ve seen, but it’s in the beginning stages.

Is MedCPU’s technology good enough to trust to accurately create discrete data from free text using your rules?

Yes. It’s looking at numbers that are discrete data from lab results, from vitals that are entered, as well as from looking at information that is entered by the physicians and nursing regarding the documentation. For pneumonia, for example, if the radiology reports mentions consolidation, it’s going to trigger and say, "Does this patient have pneumonia?" It’s going to ask you. It’s pretty active and pretty accurate. For VTE, for example, it knows whether you documented that the patient is ambulating and therefore doesn’t meet criteria for VTE. You say that in your note and it doesn’t fire. It’s pretty remarkable as to how it works now.

Will you use it more broadly going forward?

We’re piloting the ones involving SIRS, bacteremia, UTI, and sepsis. We’re piloting that now and that seems to be going well. We’ve had conversations with them about some other products that they already have. One is regarding radiology — appropriateness of ordering the right test. We’re looking at that as a possibility. They’ve already done that elsewhere, so that would be implementing something that they already have. There’s also a pretty good OB product that they have, so we may take components of that and implement it over time as well.

The beauty of it — and to be honest, the thing that’s been most beneficial to us — is that the tweaks that they’re making to the program are happening on the back end. It’s not happening on our servers. It’s not interacting with what we do. It doesn’t affect speed or anything else regarding how you document or view your information. It’s just pulling that information and sending it back to their processor and sending information back across. It’s not running on your own servers, slowing their progress as well. It’s nice because you just add something and it really is kind of seamless. They monitor to make sure it’s not affecting the Citrix servers, and if there’s an issue, then they stop it and re-calibrate whatever they need to do and turn it back on and go from there.

What will the most important IT-related priorities be at the hospital?

It’s going to be population health. That’s everybody’s answer, I’m sure. It’s just trying to get that information, trying to figure out a way to manage it to continue to do value-based care, and do so with getting compensated for keeping people out of the hospital, which is the trick that we have now.