This is a transcript of my podcast interview with Dr. Henry DePhillips, chief medical officer of MEDecision.
David Williams: This is David Williams, co-founder of MedPharma Partners and author of the Health Business Blog. I spent yesterday at Caesar’s Palace in Las Vegas, where I was a speaker at the disease management leadership forum. I didn’t hit the jackpot, but did have the chance to interview Dr. Henry DePhillips. He’s Executive VP and Chief Medical Officer of MEDecision, a collaborative care management company, which enables payers to share clinical information with physicians to improve care and lower costs.
Dr. DePhillips, thanks for joining me.
Dr. Henry DePhillips: David, thank you. Thanks for asking us to speak. It’s a pleasure to be with you today.
David: What is MEDecision?
Henry: MEDecision is a 19 year-old collaborative care management company. Let me tell you what I mean by that. I’m the chief medical officer there, so I’m responsible for the clinical content that’s embedded into the software offerings. The software offerings are specifically targeted at entities that are responsible for the cost of administering health care. Our traditional audience has been payers, but we’re also marketing our software products to TPA’s, disability insurance companies, and integrated delivery systems who take on the responsibility for the cost of care.
We have a wide arrangement of software systems to enable them to do their job. We have an integrated UM, CM, DM platform to make care management more efficient. We have an analytics engine to help them analyze their populations and act on the results of those analytics. We have a provider connectivity electronic solution that allows the exchange of information with our target customers and their providers, their networks. We have a turnkey case and disease management offering that allows health plans to administer state of the art UM and DM programs. We have a utilization management tool.
We have tools that deal with enrollment, eligibility, benefits, essentially every software piece that a health plan will need to run their business, other than a claims operating system. We are not in the claims operating system business. We build interfaces to them, but don’t do those.
David: What sort of data do you work with? Is it just claims data, or do you use other data as well?
Henry: Great question. Our historical experience, for 19 years, has been with payer data sets. Payer data sets include mainly claims data: facility data, physician data, any type of medical claims data, behavioral health data. We also deal with pharmacy data sets, which tend to be the highest quality and most timely data, but we also deal with care management data now, lab results data. We’ve been asked for more and more of what the health plans have, and a number of other data sources within the payer.
We also are working to integrate non-payer data sets, and there are some very useful data sets out there. There’s EMR (Electronic Medical Record) physician office data sets, hospital information data sets, and very importantly and exciting now is personal health record data sets. So, there’s a lot of information that patients only know about themselves, that no one else knows: their medication allergies, diet program, exercise program, family history, over the counter medications. Those are data elements that are very, very important to look at, but are not really gotten any other way.
David: What about the electronic health records from the physician office? Are those ambulatory records? Are those hospital records? Are those both? How do you integrate those data sets?
Henry: We have not yet actually integrated an ambulatory electronic medical records system. However, we’re in discussions, and at the request of our customers, we certainly are willing to do that and will do that. We just haven’t had a solid, funded request come through yet.
The direct answer to your question again is, mostly we’re talking about physician ambulatory electronic medical records, although I think there’s an opportunity for hospital information system data to be integrated as well. That data tends to be very deep in the area where the patient was cared for at the facility, but the limitation of hospital information system data is the breadth, the information about the care of the patients outside of the hospital system. That’s where the other data sets come in.
David: One of the case studies that you point to is with an emergency department. Can you tell me why that is a high-value place to be deploying your system, how that worked, and how well?
Henry: You’re referencing a third-party study that was done on some of our software capability. The software we’re talking about is called the patient clinical summary, and it’s a situation where we take all of the data sets we talked about that are available from the payer, run it through our analytics engine and create a summarized, clinically validated, electronic historical record on each patient. It’s stored on a server. It’s done currently on a batch basis about once a month, although we’d like to move to real-time, and we’re building that capability as well.
This information is available to any provider at any point of care on behalf of the health plan. We deployed this over two years ago at Christiana Care Health System, which is the dominant level-one trauma center in the state of Delaware, in conjunction with Blue Cross Blue Shield of Delaware, which insures approximately 50-55% of all Delawareans. So, you have a dominant payer, dominant provider system, it was a perfect testbed to pilot this. We also learned a bit about what it takes to deploy a new technology like this.
The results of the study that was performed by an independent research company called Heath Core were that every time one of these patient clinical summaries was transmitted to the emergency room setting, so that a doc could have that historical electronic health record, medical cost savings was $545. That included all of the costs related to the emergency room visit itself, plus the costs associated with the first day of hospitalization for the patients that were subsequently admitted from the emergency room to the hospital.
The categories where statistically significant savings were achieved included medical and surgical supplies, cardiac catheterizations, laboratory testing (and that last one makes good sense because if you know a test has been done, you may not repeat it).
Another interesting byproduct of that study was a statistically significant increase in the reimbursement to the emergency room physicians for the cognitive care that they delivered to the patient. The example we use is if an emergency room doc is treating a patient as a fractured femur, they would bill for 9928… whatever the appropriate code is.
If, however, they have this electronic historical health record, they would then know this is a case with a fractured femur and a history of heart disease, kidney disease, or COPD, or asthma, and they would bill appropriately for the medical complexity of the patient. So, this is upcoding, not in the negative sense that’s in the industry, but upcoding in the appropriate sense: we’ve made the unknown known.
So we have good news for the docs. There’s a separate study that will be published showing the time through the emergency room for patients that have this historical record is also faster, which leads to more efficiency there.
David: I’m familiar with something that sounds related. In Massachusetts, a group called MA-Share did a program called MedsInfo where they were trying to get the emergency room physicians to look up what medications patients were taking when they came in, and there they found a very low use of the system. Only a few people –if somebody came in and they were passed-out, or the physician thought they were incoherent– did they bother to check. But it sounds like this was actually a greater use. Can you talk about how you achieved higher utilization if that, in fact, was the case?
Henry: Yes, you have hit on the single most important factor related to the deployment of an electronic health record in a real time setting, like an emergency room and that is adoption. Adoption is absolutely key so we learned a lot of lessons through the Christiana care, Blue Cross of Delaware deployment and I am happy to share those with your listeners.
Probably the single most important thing is to integrate the reception of the clinical data into the existing workflow in these settings, in this case the emergency room setting. So we literally sent a team of clinically trained people from our company and from Blue Cross to the Christiana Care emergency room and spent a few days watching the work flow. How do patients present? At what point is information gotten? At what point is it determined that they are with the particular health plan that has information available? Who is it that should be pulling the information?
I am happy to share with you that we initially thought that the doctors would be the ones pulling the information and the adoption on that was almost zip so we gave up on that. Then we went to the triage nurses, who were responsible for helping evaluate and stabilizing the patients at the very initial stages. The adoption there was zip. And then, pity the poor, I say this a little tongue in cheek, administrative support staff at the front desk who are responsible for gathering the insurance type information. We wound up working it into their workflow.
They have something called a COW, a Computer On Wheels. It goes to the patient’s bedside once the patient is determined to be stable, when insurance information is gathered. It was actually a link on that Computer On Wheels that went out to the website to pull the information where it was actually finally integrated into the workflow. These poor administrative staff were essentially told to do it. Now I think they realize the value of the information and how much it helps patient care so I think they are happy to do it. I don’t want to give the wrong impression.
What’s most important is once that workflow analysis was complete and once we found the single best place in the workflow to deploy the thing, virtually 100% of the Blue Cross patients that came into the emergency room over that study period had a patient clinical study query done. That speaks volumes to the adoption issue. It is the workflow analysis and integrating this as seamlessly as possible into the existing workflow that is the key to success.
David: You’ve talked about how your tools are used for the care of individual patients. Do you also have products or solutions that are used on a more aggregate basis so that people can look at populations or you could look at the level of a physician or hospital in terms of how they are performing, to help them improve?
Henry: Yes. We have much broader clinical capability. The patient clinical summary is one of the most exciting to work with. We also offer, for health plans, state of the art, evidence based, case and disease management programs. Clinical content for 30 different case or disease areas that can be embedded directly into the care management engine, which allows a health plan to administer state of the art CM or DM program themselves. That has had huge traction in the marketplace.
We have a full time clinical staff that does nothing other than continually update all of those algorithms so that they continue to comply with evidence based medicine best practices. That is a full time job for a number of people in our company. It is very popular. We do have an automated utilization management tool that we are revamping this year. Utilization management is as old as the hills in managed care. There is still a place in managed care for utilization management. There still are procedures, new technologies with a high propensity for over-utilization that are somewhat expensive and utilization management is still the tool of choice in those limited sort of scenarios.
Last but not least is the capability that you asked me about, David and that is our analytics capability. Our analytics capability is clearly slanted towards clinical care. In other words, there is analytics capability that is offered to health plans that are for actuarial usage, for setting rates, making sure premiums are appropriate and following the finances of health care.
Ours doesn’t do that very well. What it does do extremely well is evaluate populations of patients, identify cases that are appropriate for intervention on a broad scale, measure those populations and allow the health plan to intervene in a very automated way. The results of our analytic capability is to fill, if you will, the care management workflow engine with work for the nurses to do the next time they come in. It’s not a paper output as an example.
You asked specifically about provider performance and pay for performance. We also haven’t really gone down that road. I have to be very honest with you; pay for performance is clearly on the radar of health plans but we have made a conscious decision similar to the GE model. If we can’t be first or second in a certain segment then we don’t go there. We have chosen not to go to pay for performance at the moment.
David: What is the relationship that this would have with other initiatives that people are doing? It sounds like you use data from electronic health records but I know people are talking about clinical data repositories. Some of the same customers that you are working with are Blue Cross plans where they have a huge initiative with the Blue Cross Blue Shield association. People are doing e-prescribing, they are doing tiered networks, they are doing a whole variety of things. Where does this fit in with all those? Of course you mentioned care management where you are at the disease management conference as well. How do you actually not get lost in all this? Where do you see yourself fitting?
Henry: The most important thing about the era of health care that we are in today is the era of clinical data exchange; sharing timely information with clinical decision makers at the point of care to allow for more informed decisions is the generation that we are in today. What is exciting from a technology standpoint –remember that I am a physician and relatively new to the technology side of this business– as a former private practice physician is that once you gather all this information and get it into a format that is usable for a clinician at the point of care and it is in electronic format, you can send it anywhere.
Obviously there are limitations under HIPAA and there are compliance issues that you need to pay a lot of attention to but consider this. You can send it, as we have, to the emergency room doctor so that they can have an historical medication and medical problem record on patients that would otherwise not be able to give a history at all. You can also send it to electronic medical records systems in doctors’ offices.
If you are a doctor you spend $50,000 – $75,000 per physician in your practice for an EMR system that’s great but the first day you have it, what good is it? The answer is that if you could pre-populate those EMR systems with all of the information that the payers have on their members, there’d be an incredible increase in value. Pre-populate it with medications and so on.
Similar to hospital information systems: a complex patient comes into a hospital; the hospital knows all about the care that was given there – but what about the out-of-system data? To have those extra data would be tremendously helpful to the hospitals.
And then the whole industry of personal health records. I think late 2007 and all of 2008 will be the year of the PHR. Imagine: you know, if you’ve done one so far, that, when you went to build your own personal health record, it’s essentially an electronic blank piece of paper: list your medications; list the problems that you have; list the doctors that you’ve seen. It’s a memory-recall sort of thing.
Imagine if all of the data that your payer knows about you, through claims and pharmacy processing and other things, pre-populated that record. The question isn’t then “List all your medications”; the question becomes “Here’s a list of the meds that we have a record of you taking; please, edit the list.” How much more useful would that be?
So it’s an exciting time. And the sharing of clinical information is clearly going to be a focus for the foreseeable future.
David: I know you’re not a salesman. But, if you think about the value proposition that you offer to a health plan, are there particular types of health plans that you approach? It sounded as if, in the Delaware instance, it made sense where there was a dominant plan; they were able to achieve something quickly. Are there particular pitches that are used for different plans? What would the overall rationale be for a health plan to sign on with your service?
Henry: David, also a very good question. Historically, MEDecision has specifically targeted large, complex, regionally dominant payers: Blues plans are a good example of that; they tend to be fairly dominant in the marketplace.
We did that specifically because our mission is to improve the relationship among payers, patients, and providers. We targeted regionally dominant plans because they’re the most influential payers in their respective marketplaces. So it was a conscious decision that MEDecision went out and worked with those customers first.
We now, I think, have a much broader appeal to health plans essentially of all types and sizes – probably above a certain membership level, on the basis of the complexity, and, quite frankly, the cost, of our offerings.
But, nevertheless, we can benefit almost any health plan at this point. What are those benefits? And you’re right: I’m not in sales, so I’m going to have some trouble articulating this.
David: It’s more genuine, though…
Henry: I appreciate that. The benefit is that, if we can automate systems to allow each of the folks inside the health plan who administer the benefit plan to be able to do the work of two or three people, then we’ve brought tremendous value to the marketplace.
The administrative cost of health care is – you know, it may be only nine to eleven cents on the dollar, but that’s a very, very important nine to eleven cents and, if it can be nine instead of eleven, that’s a competitive advantage for any insurance company.
Secondly, our offerings have grown and have a much broader appeal. So, traditionally, maybe commercially insured – but we have very broad appeal on a large number of Medicaid customers; customers who do Medicare risk; TPAs; administrators; integrated delivery systems, who take on the responsibility for the cost of care. So the appeal has gone much broader.
And it’s really nice to see that… We have an annual user-group meeting, which actually started at MEDecision in 2004, the year I started with the company. So it’s a great way to get to know the company. But more than half of our customer base comes in for that user-group meeting and the sharing of ideas.
And some of our customers are doing things with our software that we didn’t even think about, in ways that leverage them even further. And for those customers to share with our other customers some of those best practices has been a lot of fun to watch.
So what’s really cool is that you have a situation in which the people who are responsible for delivering quality care in a cost-effective way – all of us are pulling in the same direction. I’m a doc; my goal is healthier patients. It happens to be in alignment with the insurance industry’s goal of lowering costs, because healthier patients cost less money.
So there’s a fantastic alignment here of incentives in the industry. And I think that’s one of the things that makes it exciting.
David: You talked about it being an exciting time. If you look five years into the future, what do you hope to be able to then that you can’t do now, either from a clinical standpoint or what kind of effect you could expect in that time?
Henry: Wow. Well, I think that technology clearly will continue to advance. And the fun thing about being in technology is that technology does not advance linearly; it advances exponentially: so, five years from today, we probably will have ten times the technology capability that we have today.
My goal is for MEDecision to be a part of that, of driving that. And I truly believe that we will be: our CEO is a very visionary guy; he’s a lot of fun to talk to about what health care looks like five years from today.
But I think there’s a couple of very, very important things. First of all, we can’t forget about the patients, and we can’t forget about the majority of patients. There’s a range of education, ability, technology orientation, even Internet connectivity; and we have to keep that in mind. So we have to bring all of the target population along as we go down this technology journey.
Number two: we have to bring things to the market that add value, because there’s a lot of technology out there that’s very expensive and very fancy but, when all is said and done, it doesn’t produce an increase in value in the industry. And the industry’s pretty good about picking that apart; but, nevertheless, we just have to be cautious about that.
Number three: there has to be a sustainable business model. There’s a lot of grant money and demonstration projects. And I’m not getting on the government or anybody else; but, while, I’m fine with seed money to get something started, there needs, at the end of the day, to be a sustainable business model, in which value is created and the stakeholders who receive the value are willing to pony up a little bit of extra money to attain that value.
Now, they need to get more value than the cost; so it has to be a positive risk benefit. But I think that the whole RHIO industry – which we could spend a lot of time talking about, although I know that’s not the discussion for the day – you see that breaking down. And I think the reason is that there’s been a lot of grant money, but not a sustainable business model, to come out of that.
If you look at – and I’ll highlight our experience in Delaware, with this patient clinical summary: all we did was take information from the payer, give it to the docs; the docs were happy, for they got it for free. The health plan was happy, for medical costs were reduced; and the patients were happy, because more-informed decisions were made about their care. A very simple business model: but it was a win-win-win situation; and there was, quite frankly, a positive revenue model there for MEDecision. So we now have more seed money to go create more technology.
David: I’ve been speaking today with Dr. Henry DePhillips, the chief medical officer and executive vice president at MEDecision, here in Las Vegas, at the Disease Management Leadership forum. Dr. DePhillips, thanks for speaking with me today.
October 8, 2007
Henry: David, it’s been a pleasure. I appreciate your listening; and I appreciate your listening to our ideas. And it’s really a pleasure to contribute to the information that you get out into the marketplace.