Health data interoperability has long been a goal of health IT executives and policy experts. But it's 2022 — and the healthcare system doesn't appear all that close to getting there.
Becker's spoke to experts from health systems, industry and academia on what it will take to create an open exchange of healthcare information in the U.S.
Here is the first entry in the series, a Q&A with Donna Roach, CIO of University of Utah Health in Salt Lake City:
Note: Her responses have been lightly edited for clarity and brevity.
Q: What will it take to get to EHR interoperability?
A: One area of complexity is that we don't have a unique patient identifier. This has been codified in federal law. We're not allowed to have a unique patient identifier across the United States. If I have a patient record, I can identify it in my system, but if that patient goes down to Florida for the winter, sharing of that data can be very difficult to do.
We've made great strides in the Epic environment. Epic has created some ways of being able to share that data. But if that patient goes to a Cerner hospital or gets treated at a Walgreens, how does that data get shared?
Q: Can you explain the concept of a unique patient identifier?
A: In a regular electronic health record, we have what we call a master patient index, an MPI. And in that MPI, I assigned an MPI number to you, which is unique, but that's in my system.
The state health information exchanges that exist try to do the best job possible, but again, there is no global identifier to give an overall picture of what their care looks like based on movement around the United States.
I've lived in a lot of different states. I have five different MyChart accounts within Epic. What's good about that is I can share that data with those various Epic sites. But those are just Epic sites. I also have a record in Cerner.
And how do you track your children's vaccination record? When I was growing up, it was a paper record. It was a paper card. I don't know that we've made too much headway off that paper record.
Q: Why can't individuals just merge their various patient portals into one?
A: People will say, 'Why can't that record from Cerner just populate into the Epic system?' Well, how they built their system is very different from how Epic built theirs.
We have standards that help formulate some data transfer. But what about the notes? Do the physician notes come across? Well, no. But that standard data element may come across. If I had an emergency department visit, some of that information may come into my record, but it doesn't necessarily mean all the information travels over.
So it is a bit more complex than people realize. If I had any imaging studies, do those imaging studies come over?
Q: How has Utah's health information exchange been working?
A: We were able to utilize that with COVID and give data to the state around our testing process and our vaccination process. But you've got to remember, we are a smaller state, and we don't have that many health systems. University of Utah is one major one. We have Intermountain, which is another major system. They're on Cerner. You have St. Mark's. You've got some smaller for-profits. HCA is here.
We don't have the complexity of a state like Illinois that has five or six major academic medical centers just in Chicago. You've got some big players there, that all built up their systems independent from one another. And so that impacts interoperability.
Q: Is the relative lack of competition in the EHR market good or bad for the interoperability cause?
A: You've got Meditech. You've got Allscripts. You need competition so hospitals are not priced out of the market. Because an Epic, a Cerner can be very expensive to implement, and then down the road, support. It's millions of dollars of investment.
How does a smaller community hospital afford it? A lot of times an Epic won't quote out their system to a smaller community hospital. It has to be minimally at a certain level or Epic will say, 'Partner up with a Connect site.' So that Connect site will sponsor the Epic implementation.
A lot of the community hospitals are starting to shut down because it becomes not cost effective anymore.
Competition is good in order to manage the pricing of it. But competition can also be a deterrent to the interoperability that needs to exist.
Q: What other factors will it take to get to interoperability?
A: It definitely is regulation. There are some great organizations that are trying to make sure that people are aware of what needs to happen. The vendors also need to come to the table, and you see some of that.
What I do get a little concerned about is you've got some disruption in the industry with the Amazons and the Walgreens and the CVSes, and they're not held to the same criteria that I, as a not-for-profit healthcare system, am held to in terms of data and the overall healthcare delivery mechanisms.
So they can skim off the low-hanging fruit. But are they really contributing to the wellness of the community?
It's nice to be able to have a Walgreens close by, and they can offer the 24/7 service. But where does that information go? If I present to my regular primary care provider, and I went and saw somebody at Walgreens, how do they know the complete record unless I tell them or I bring in the documentation? There's nothing I've seen yet that creates that kind of flow of data.
Q: Does competition among vendors and health systems hold back interoperability?
A: [Vendors] all have it on their radar, and they all want to make it happen, but they want to make it happen on their terms. There's no global mechanism. But there's been some great progress made.
We share a lot of data with Intermountain as we're treating patients. There is a lot of cloud collaboration that goes on, but is it perfect? Is it seamless? No. Absolutely not.
You're kind of better off to stick with one health system so they have all of the data. But sometimes even within our own system, we have our own little hiccups and are still working through them. But it's gotten so much better than when I first started. There was no direction. People saw that healthcare data was a competitive definer to their market, and I don't think people think of it that way anymore.
It's really about how do you move the dial on creating better healthcare for your community. And holding back healthcare data really shouldn't be part of that, right? The health data should be owned by the patient, not necessarily the health system.
Q: Why is interoperability good for patients, the healthcare system and the health of the public?
A: If I've already had a test done, or I already have results, somebody can see the complete picture of what's involved in my care and expedite the care. They can see that your blood sugar is dropping, and therefore do something about it.
But if I don't have that complete picture, they would never know you just had a CT scan three months ago. You shouldn't have to repeat a test unless it's medically necessary to repeat the test. So there's value in that.
When you have a more complete picture, you can also start to see where somebody may be developing more chronic diseases or comorbidities. So you start to see that maybe somebody is prediabetic if you have a longitudinal record of how they've been doing over time, or maybe that a growth is precancerous. But if I didn't have that information ahead of time, and now you present with fully stage 3 cancer growth, if I had seen what had been imaged three or four years ago I may have been able to catch it. Or if I had that blood test, I could see the indicators early on. I could have done something.
In the industry right now, they're talking a lot about AI. But how can AI be used as predictors of healthcare if you don't have a complete dataset? AI is great at helping the physician look at, say, early onset kidney disease. So let's start to do some interventions early on rather than wait until you're in full renal failure and on a list for a kidney transplant. The beauty of AI is starting to help predict some of this stuff. I don't think we're fully there yet.