The seamless sharing of patient data has long been a goal of many in healthcare. Just ask Peter Winkelstein, MD.
He has led the health informatics program at the University of Buffalo for 25 years. He is also chief medical information officer for Kaleida Health and UBMD Physicians' Group, both based in Buffalo, N.Y.
So he's certainly qualified to answer the question: Why aren't we there yet?
As part of a series on healthcare data-sharing, (parts one and two are here), Becker's recently discussed the topic with Dr. Winkelstein, executive director of the University at Buffalo's Institute for Healthcare Informatics.
Note: The interview has been condensed and edited for clarity.
Q: Why do you think interoperability hasn't happened yet? What might it take to get there?
A: Well, the short answer is: It's hard. It's easy to say let's move data freely among electronic health records and other sorts of platforms. But it actually is really hard to do, and that's why it's not been done yet.
However, I think over the past 10 years, or an even shorter period of time, we've made significant progress. And that's happened for several reasons. One is that there's industry interest in providing interoperability, and so there are now industry platforms for interoperability. And that's very powerful because the industry is relatively consolidated. And so they're in a position to be able to set interoperability standards pretty well.
The other is, of course, the government has been pressing for this.
The goal is pretty straightforward. A patient clearly wants their data to be able to travel with them, so that when they go from one provider or doctor to another, or one health organization to another, when they arrive the appropriate data is with them. And they want that done in a secure fashion.
This all strikes me as extremely reasonable and something we should be striving to achieve. And there's not much debate about that. I don't know anybody who says we should not do health interoperability. Everyone agrees that the data should be able to flow. But as I said, the fundamental problem is that it's hard. It's hard because the data is complicated. And it's hard because the data is scattered over so many different places, because patients typically go to many different offices, and many different healthcare facilities often each have their own EHR, and even if it's the same vendor, it may be their own implementation of the EHR. Also the insurance companies have their own systems.
One possible advancement in technology is the development of relatively easy-to-access API's, or application programming interfaces, so third-party apps, non-EHR vendor apps, can access your health data in a secure fashion with your permission and can begin to provide some ability to aggregate it and transport it from the patient standpoint.
So there are a number of examples of that. Probably the one that readers would know the most is Apple Health. It's a little app that comes with your phone. If you go into Apple Health, you'll discover there are a lot of different health organizations that have already opened up their EHRs so that you can connect to them with the Apple Health app. I can connect two or three different EHRs where I have personal medical data into the Apple Health app, and then that allows it to aggregate in one place.
Q: Can you quantify where the healthcare system is at now in terms of interoperability, and what will it take to get to 100 percent?
A: I don't quite know how to quantify it. I would say most, if not all EHRs, offer patient portals, so a patient can access their data by signing up for the patient portal and going and looking at the data in the EHR. Patients can at least see some of their data, and now with the information-blocking laws, the amount of data that patients will be able to see is going to be greater and greater and greater. And then, as part of the same regulations, they'll be able to transport that data into these third-party apps.
Patient access to their own data has improved dramatically, is close to 100 percent and will be at 100 percent or more with the current information-blocking regulations and the APIs. How many patients are taking advantage of that access? I don't really know. How much do they really want to use portals? How much do they really want to look at their data? I don't know.
Moving the data from one system to another — that's more complicated. That requires a higher level of technology. And there are a couple of different routes for that. One is using health information exchanges. So you send the data to a central place, and then you bring it down from there to your computer as you need it. And now there's direct messaging, which is essentially a secure form of email that's designed for health information. That's in the early days of penetration, but I think it will increase dramatically. And that is designed for provider-to-provider, EHR-to-EHR type communication. So when you think about interoperability and flow of data, sometimes you need to think about who it's flowing to. Flowing it to the patient is one thing. Flowing it between electronic health records is another thing.
Some of this has nothing to do with the technology. Things like patient portals — are you encouraging your patients to use the portal? So every time a patient comes in, you say, 'Hey, sign up for the portal area. If you're not, we recommend it.' That's not a technology question. That becomes a question of how does the office want to connect with their patients.
When I hear medical organizations talk about patient portals, their concern usually falls into one or two different categories. One category is they're worried, I think legitimately, about making sure that patients don't get bad news before they've had a chance to call the patient. They want to deliver bad news themselves. They don't want the patient to find out bad news by looking it up on the computer.
The other related concern is that sometimes it's difficult, depending on how the information goes to the portal, for patients to interpret the significance of abnormal results.
And then there's sometimes they're worried about, 'Boy, you know, we have trouble handling all the phone calls we get as it is. Now in addition to that, we add messaging from the patient portal, then we're worried about being able to handle the volume of calls plus portal requests.' Despite the fact that there's pretty good data that portal requests are much more efficient than phone calls. So actually, if these patients preferentially use the portal rather than the phone, you actually do better. But nonetheless, it's a concern that they don't want to miss a message.
No one says the patient shouldn't be able to see their own information. I think the concerns are that we want to make sure that, again, patients have access to good information and that we're able to respond to them in a timely fashion, and we don't want to get overwhelmed.
Q: Is consolidation in the hospital and EHR industry a good thing for interoperability?
A: It's simpler to figure out how to exchange information if you're talking about exchanging among three or four different systems than if you're talking about 40 or 50 different systems.
In terms of making it easier to agree on standards for information exchange, the smaller number of players makes that easier, and especially in our world where you really have two big players in the EHR world. If they basically agree on how we're going to exchange information, which basically, they've done, then that de facto is the standard. I mean, they're just going to call the shots.
So in that sense, having a few big players makes it easier and moves us closer toward interoperability, and I think that that has happened through CommonWell and other agreements, where the big players just said, 'Yep, we're going to agree to exchange in this format,' and then they just do it. I think that's helpful.
What's going on with competition in the EHR world? We basically have two big players and then some medium-sized players and some small players. And the two big players are enormously big. That's obviously Cerner and Epic, and they'll be setting the standard for the big pieces of this.
I don't see it likely that a third vendor can achieve the scale of a Cerner or an Epic at this point. I think that's unlikely for a couple of reasons, not the least of which is that most places have already implemented EHRs, and changing EHRs is an expensive, time-consuming practice. It happens, but you don't do it lightly. You really have to think that through.
So one big question is: What does the future look like in terms of smaller niche products that can be tightly integrated into the big platform? So the analogy is something like the Android or the iPhone. What really is important on your iPhone are the apps. And the apps can't run without a very sophisticated platform. But mostly you're using your apps, and many of them are not the eight apps that come with the phone.
So the question is, can you imagine a similar sort of ecosystem, for lack of a better term, with EHR vendors, so a Cerner or an Epic or one of the other big players is the iPhone- or Android-type of program? They're going to provide the fundamental platform on which you build.
And already this is happening. Epic has the App Orchard, and Cerner has an equivalent app store. And so they're encouraging third-party apps to build on their platform and opening up APIs. Competition in that area is very likely because it's simpler to get into. And it's relatively new. We don't really know where this is going to go.
But the analogy with an iPhone is simplistic. An iPhone is pretty much a blank slate, but that's not true of an EHR. An EHR is a complex piece of software with a particular look and feel, and we don't really know how apps are going to be able to integrate with that. But that's probably where the next wave of competition and innovation is, rather than somebody coming up with the next version of a big systemwide EHR platform. The next level is smaller pieces, maybe via apps.
Epic and Cerner are encouraging it as near as I can tell. We can call up and say, 'Hey, we're interested in this kind of function or that kind of function.' They're saying, 'Yep, our partner is doing that via an app. Here's where to look for that.'
I think the companies are also struggling with how much does a Cerner do, how much does an Epic do internally with their own developers, and how much do they say, 'We're going to work with external developers through an API or some kind of app infrastructure and let them do it, have it tightly integrated with our product.' I think that's still very fluid right now.
Q: Are there any downsides to the dominance of Epic and Cerner?
A: In general, oligopolies are where we tend to wind up — a relatively small number of airlines or a small number of oil companies. And I think we could spend a lot of time with our academic economic hats about whether oligopolies are a good thing or a bad thing, and I think the answer is that it's some of both.
Certainly, competition is really good for the soul, and oligopolies tend to minimize competition. On the other hand, you've got to be realistic. You can't create an airline from scratch easily. There are high barriers to entry. The Epics, the Cerners and the other companies that have been around for a long time — there's a lot of institutional knowledge there about how health systems work that is very difficult to obtain. I just don't think you're ever going to have a lot of big players in the field. Obviously, competition might help lower costs, might help spur innovation, make products better.
One of the problems in the health world is that the pace of innovation is slowed by the complex regulatory environment. Healthcare is not a field where you can go fast and break things. You don't really want to be going fast and breaking things when you're talking about taking care of patients. It limits you on innovation.
Perhaps the most obvious example is billing. In order to be able to submit a bill, to be reimbursed for medical stuff, there's a whole set of what you have to document. And that forces systems to do a whole set of things that are required by regulation, and that healthcare providers may feel aren't particularly valuable in terms of providing healthcare. But there's nothing you can do about it, because you've got to do it, because that's what the billing regulations are.
And the EHRs allow us to try to do that in as easy a way as possible. And so the EHR vendors have to spend a whole lot of time meeting all these regulatory requirements, which takes away from the opportunity to innovate. They can't because they've got to be able to produce the appropriate documentation, the appropriate codes, the appropriate information so that you can meet regulatory environments and get reimbursed.
Q: Is the federal government's 2030 goal for improving interoperability outcomes realistic?
A: It depends how you define interoperability. I see more and more data becoming more and more readily transportable over the next five to 10 years, easily. That's already happening.
I think we're making steady and good progress in interoperability. The government's requiring it. Everybody thinks it's a good idea. The patients want it.
We're going to get it done. You just have to understand, it's hard to do. And the pace of technological change in healthcare tends to be slow, because it's such a highly regulated environment and a complex environment. So it's going to take some time, but I think we're getting there.