Sep 4, 2015


Announcer: These are the conversations happening inside health care that are going to transform health care. The Health Care Insider is on The Scope.

Dr. Chapman: Today we have with us Dr. Alistar Erskine and Dr. Stan Huff. Dr. Erskine is the Chief Clinical Informatics Officer at Geisinger Health System in central Pennsylvania and Stan Huff is the Chief Medical Information Officer at Intermountain Health Care here at Salt Lake City, Utah.

I'm going to tell a story that'll introduce the topic that we're discussing. At Intermountain Health Care, they have a program called The Antibiotic Assistant. When the patients in a hospital, if they develop a new infection, then the electronic medical record system collects information on that patient in the background. And if it looks like they're developing a new infection, it will alert the physician and say, "We think your patient is developing pneumonia and here's the type of antibiotic that we suggest you treat them with. And here's the evidence why."

This is the program that has been in place in Intermountain Health Care for a while and a woman in another state heard about this. She sent one of the physicians there, Dr. Scott Evans, a blank DVD. And she said, "I read your article of The Antibiotic Assistant and they don't have it in my hospital. Would you put that program onto this DVD so that my hospital can use this?"

This illustrates there are a lot of great innovations happening across the nation. A lot of different types of application are being developed, but they're not shareable across different health systems. And if we really want to get to a place where we have the best care in our nation and in the world, we have to be able to share these applications.

So Dr. Erskine and Dr. Huff are both working on technologies of how you can create applications, plug them into different electronic health records to be able to share those types of innovations. So, Dr. Erskine, would you talk about what you do in Geisinger?

Dr. Erskine: So, yes, I think at Geisinger, we have recognized that there are a series of processes we're re-engineering that are necessary in order to be able to maximize the value we get out of the care we deliver. And we've also gone through a process of focusing only on certain kinds of illnesses and procedures and re-engineering those processes to improve the unjustified variation in care and the outcome with the patients.

When we went through a process of trying to share that knowledge using our electronic health record with other hospitals and clinics that use the same electronic health record, we found the task practically impossible to be able to lift the work that we had done and share those workflows with another hospital. And if we try to do that with another electronic health record vendor, that process was even more difficult. So by following the work The Office of The National Coordinator had been promoting, the SMART platform, which was an effort led by Ken Mandl and Zach Klein and Josh Mandel at Boston Children's hospital . . . developed a platform they call SMART, which allows an application, an app to be able to interact with a number of different electronic health record systems using a nonproprietary, public means of doing so.

So the app can make sure that it's got the authority to access the patient information and then the app gets the context of what's going on with the patient and the provider and the encounter. And then, using a series of publicly available exchange mechanisms, which is referred to as FHIR, this smart, on-fire as it's called in the industry approach makes it easier for people to share their application that's built, for example, at Geisinger, for example in rheumatology, and have that plug into another electronic health record system in another hospital to share that knowledge and functionality across systems.

Dr. Chapman: So, Dr. Huff, at Intermountain Health Care, you're adopting a new electronic medical record system called Cerner. And Cerner has invested in developing this type of platform to share these types of apps. Can you describe why that's of interest to a health system?

Dr. Huff: One of the things that we're concerned about in purchasing the commercial system is that we would not have the flexibility to develop the kind of clinical decision support apps you already described like the Antibiotic Assistant. And we have probably 150 or so of those things that we've developed. And we want to continue to develop those because we've proven that those things will improve the quality and decrease the cost of health care.

So we were very interested in preserving that ability to create applications that would continue our legacy of really using medical information to improve the quality of care. Besides preserving that capability, another thing we realized is that if we're doing 150 or so of these kinds of apps, the real opportunity is to do 5000 because the apps that we choose to do are things that are the common processes where there's a lot of variability or it's a common illness like diabetes or heart failure or some other process. And we're really not doing anything for people with hyperthyroidism or myasthenia gravis or rheumatoid arthritis or all of those other things.

And we can imagine getting a little more budget and getting more efficient and maybe we can get from doing 150 things to doing 200 things. We've got to fundamentally change the paradigm if we're going to get to 5000 things and have the impact that we want, because at some point you have to get to the less common diseases with clinical decisions support to do that. And so this whole strategy that Alistar has described of exchanging knowledge as executable apps we see as that new paradigm that will allow us to really move from the small set of things that we're doing to sharing knowledge with a much broader community so that we can get to 5000 or 10,000 things that we could really do.

Dr. Chapman: Alistar, describe how the smart "on fire" paradigm that you've both been discussing my look in five years.

Dr. Erskine: The potential that a new market is established, not unlike the market we've seen emerge from the iPhone, and the Appstore and the iPhone where unimaginable types of products that serve not just the clinicians but also patients and other components of the health care continuum, I think that is a market that I think we would all benefit from. It will not come without a series of growing pains associated with how to curate the app information that has been produced, how to interact, even from a contractual standpoint with various different hospitals to support the app in a long-term kind of way.

But I think in five years, as we continue to democratize the functionality that is currently locked in only a few systems and a few EHR vendors, we're going to be poised to be in a position to be less dependent upon the system of record to extract knowledge and workflow and more dependent upon the collective, innovative approach from all the people at the front lines that maybe the subspecialist for in more rare disease that an EHR developer may not spend the time, effort and energy to develop. So we'll be able to level the playing field so that there's a low point of entry to even those who typically aren't the primary points of focus in terms of developing apps.

Dr. Chapman: Here at the University of Utah, we have the top-rated video game design undergraduate degree in the country. It's called the Entertainment Arts and Engineering. And our health system here at University of Utah Health Care has invested in a group of students there to help design some apps that apply to health care and help patients monitor their health. And this is an area where their apps that the individual can use on their own and they don't interact with the electronic health record.

But once you start needing to collect data that your doctor has about you and to give data back so your doctor can see how your wound care is coming or can look at pictures that you've posted about your new infection, once that starts interacting like that, then you need to have this type of a platform like the SMART FIHR to be able to connect the app to the EHR. And so this is a really a great area that I think we'll be heading down in the future. It's a great vision. What are the obstacles in getting to that vision?

Dr. Erskine: I think one of the obstacles is the fact that this is still fairly new and broad adoption of these nonproprietary advanced application interfaces have not been fully adopted by all the EHR vendors. It's not to say they're not planning on doing so, but that will occur over a period of time and it will not occur in a way that is complete at inception. So there will still be a need to engineer a process that still maps into the proprietary nature of an electronic health record system while we wait for there to be full adoption from the EHR vendors. So that's one obstacle.

The other obstacle is it's very much an unproven business case as to how the app developer will negotiate with the given hospital system to support that app. So I think that market has yet to declare itself. It will be useful to follow that over time to see how successful that is.

A third obstacle is we're dealing with the patient information and as a result we have to be cautious about the privacy issues around moving the patient information out of the secured electronic health record into a series of apps. So the various app developers will have to go through a careful curation process and certification process to know at any one time where that patient data is located. Is it in the cloud? If it is, is it secured in a way that is reliable and known? And the various different security parameters around the app to be able to make sure that untoward, nefarious things are being done to patient's data.

Dr. Huff: I would add one sort of nerdy, technical challenge to that. The hard work, if you will, is for these standards-based services to work, data has to be transformed from different physical representations and different internal codings into the standard structure. And so if you're going to share data, I think you can mathematically prove that that's the best way to do it, but it's work. It's still work. For people to be able to understand how data is represented physically in their own system and create the representation of the data that's presented by the service. And so people need to recognize the value of what we're doing and be willing to invest in actually installing and configuring the system so that you make that mapping from the physical representation in the given system to the logical representation that's now the standard for creating these interoperable applications.

Building off what Alistar said, a key part of this is understanding the business motivators that will make a new marketplace available. And there are, I think, at least two important aspects of that. One is why the big vendors like Cerner and Epic and Allscripts and Athenahealth and those people would be interested. And then why would other software developers be interested in this. I would go back. If you look at sort of a different aspect to this problem today, you go to the HIMSS Conference and you have sort of the prime real estate where all of the big vendors live.

And so there's, depending on how you want to count, there are six or eight or 15 or whatever of the really major vendors of electronic health record systems. But there are over 1300 people who display at HIMSS. And the majority of those 1300 are small companies who have really creative, innovative ideas and they have no market for what they're doing. They're creating things and they're dependent upon data that's in the electronic health record and they have no entry into that marketplace.

So they either have to create a relationship with the company and essentially partner with that company and now they have to make versions of the software that work with each individual vendor and then it becomes a derivative work of that vendor. So they don't have free ability to market or share or open source it because of those sort of contractual and business obligations.

The thing that this does, one of the really important things that creating these standards based applications is that the standards are open and everyone can use them and so that's not a barrier for them. And because they create the app on that basis it's not a derivative work from any one of the existing and so they have the opportunity now to be what we would think of as a much more free-wheeling, free enterprise entrepreneurial system in bringing their applications forward.

Now, traditionally, looking at the other side, the obvious motivation for most of those 1300 is that you're creating a marketplace where they can be creative and they have a place to sell their app. For the other systems, the people that already have a large market share, you have to ask, "What's the motivation?" And traditionally, they haven't seen the motivation. In fact, for the last 30 years, mostly what you have seen is, "We don't want to interact with those other things. You need to buy everything from a single vendor. It needs to be a single vertical market." And that's been the strategy.

I think recently there are a couple of things that those companies have realized. One is in spite of them becoming big and having thousands of programmers, there's always a gap between that what they're producing and things that their customers want. And that's because you have a process where if their customers want things, they submit those, they prioritize them, they figure out how many they can do and then they do them. And then you make progress. I mean, it's not like we're not making progress.

On the other hand, what happens is if I'm a particular customer, I come in with a list of 15 things and maybe five of my things make the list that are going to get done and I go home with 10 things that are undone. And by the next year, I've got a list of 20 things. And so there's always this ever-present gap between things that I want and need and what things can be accomplished through that kind of central planned economy, if you will, that's going on within the vendors.

And I think, ultimately, I think they actually want their customers to have better applications and they realize that they can't produce all of them themselves. But the other part I think Alistar also alluded to, and that is if you think about somebody who is very different from Geisinger or Intermountain Health Care, that's a 150 or 200-bed community hospital, they don't have any informaticists. They're not going to produce applications. They don't have anybody to install the applications or to monitor the applications. And so it's an entirely new market opportunity for the existing vendors because they are the ones that have the business relationship with those organizations.

And so the large companies can make a business of assuring that the applications run against their systems and that they're safe, that they don't cause performance degradation, that they don't cause any kind of downtime associated with the app. They can worry about updating the app when the new versions come app, they can help educate if education is needed. I mean, there is a whole set of services now that they can offer around these new apps that are value-added. It's not like they're trying to get something for nothing. They will be absolutely offering new value to their customers that their customers should be able to take advantage of should be willing to pay to have that new capability in their system.

And so I think we're still working through what the motivation is, but I think it's there and I think as we move into it, there will be greater and greater understanding and the marketplace will open up. But you have to actually have both things. You have to have buy-in from the major vendors and people with market share to support the standard based services against their platform. And then you have to have this entrepreneurial, creative, innovative set of app developers who create the apps that are going to populate the app store and that five years from now will blow our socks off with the kind of things that they'll create.

Announcer: is University of Utah Health Sciences radio. If you like what you've heard, make sure to get out latest content by following us on Facebook. Just click on the Facebook icon at

For Patients