If the mark of a good leader is the ability to evolve, as many believe it is, then Chuck Christian may be one of the greatest. Throughout his career, he’s been a staunch advocate for state’s rights, and a firm believer that healthcare is provided locally, and therefore should be governed locally. But as someone who has been heavily involved in advancing data sharing (both as Director of Executive Engagement with the Indiana Health Information Exchange, as well as an active member of CHIME and HIMSS), he has seen the challenges created by the lack of an overarching consent model. And so, he believes perhaps it’s time to make some compromises.
Recently, we spoke with Christian about why he believes the patient-centered data home model could be a difference maker, the need for better education and more transparency when it comes to data access, and what he hopes to see from ONC’s proposed rule. He also provides his thoughts on CommonWell and Carequality, the obstacles still facing smaller facilities when it comes to federal regulations, and what he believes is the key to effective advocacy.
- Continued growth of PCDHs – “The data is following the patient wherever they seek care.”
- HIEs & the concept of “regional hubs”
- CommonWell & Carequality – “How do we support population health using that type of model?”
- Vendor adoption trends in KLAS’ Interoperability report
- Direct Messaging in rural areas – “We use the tools we have.”
- DirectTrust: “It works well, but does it work for everything?”
- Thoughts on TEFCA
LISTEN NOW USING THE PLAYER BELOW OR CLICK HERE TO SUBSCRIBE TO OUR iTUNES PODCAST FEED
The Carequality platform has been filling those gaps, but it’s going to be around a very small and defined number of use cases. I’m waiting for a deeper explanation or conversation around, how do we support population level health using that type of model?’
They’re kind of like Apple. If they can control both ends of the transaction, it works great. But it used to be if you tried to get data from a facility that doesn’t use Epic, it didn’t work well. I think that’s getting better too, because we’re getting better with the standards.
Interoperability isn’t like the Lord of the Rings. There isn’t going to be one ring that controls all. I think that it’s going to be more like Six Sigma, and it’s going to depend upon the need at the time and the market.
I’ve tried to put myself in the position of people in the federal government who are having these conversations. What they’re hearing is, ‘I don’t have a standard way to connect to the highway. I’ve got this vendor who wants to do it this way, and this vendor who wants to do it that way.’
To get to a broader community, people want one way of doing it. I understand that. Particularly if you’re a physician practice — you just want to plug it in, but the question, is what do you plug it into?
Christian: One thing we’re seeing is that the Patient-Centered Data Home continues to grow. We continue to gain members and the transaction rates continue to increase, but it’s only going to happen based upon where people need the data. And access to that data is controlled by the patient walking into a facility that is not in their home HIE space. Basically, the data follows the patient wherever they get care.
It’s kind of simple, but it hasn’t been simple to implement, because there are a lot of questions we need to ask. Does everybody need to be connected to everybody? No. Let’s do a model that looks like the federal banks. I believe there are 13 Federal Reserve Banks that move transactions back and forth. That’s how banks communicate data.
We’ve adopted a similar model where we have hubs that are regional. Right now we have four, and I think that’s all we’re going to need, unless we get overwhelmed by additional HIEs. I don’t think there are enough HIEs in the country to overwhelm the current model we have, but there are some white spaces we need to deal with where there are no HIEs. I’ll use the State of Illinois as an example. There’s a small one in the middle of the state that’s not connected to a lot of facilities. They’re servicing their market right now, but we know the grand majority of the population in Illinois lives in the Great Lakes area and in Chicago. In those areas, they’re using things like Care Everywhere and Carequality to move data around, but — and this is my opinion — all of that is predicated upon a single-use case, as is the patient-centered data home. It’s predicated on the idea that a patient shows up, a clinician needs data, somebody asks for it, and it gets delivered.
In many of the regions that don’t have HIEs, that’s what they fall back on — ‘I’ve got to get to the data. How do I do that?’ That’s where CommonWell and Carequality come in. The Carequality platform has been filling those gaps, but it’s going to be around a very small and defined number of use cases. I’m waiting for a deeper explanation or conversation around, how do we support population level health using that type of model?’ Part of the answer is that ONC is funding a new standard through HL7. It was originally called Flat FHIR. I don’t know what the official name is where they could make one FHIR request and extract population level data that would support an ACO, but we don’t have enough time for me to state all the questions I have about that.
Gamble: Right. You mentioned CommonWell and Carequality. What do you think this means in terms of being able to share patient data? Is it something you can realistically see in the future?
Christian: You know me, any opinion I give you is going to be biased. But as you know, KLAS recently released the 2018 Interoperability report. I’ve known those guys since the beginning. I’ve sat on their advisory board for four or five years, and I consider Kent, Scott, and Adam (I never had a chance to meet Lawrence) to be personal friends. One of the things they mentioned in that report, which I thought was interesting, is that they’re seeing a lot more vendor adoption as far as CommonWell and Carequality, but not all of them have the connections and the development on their roadmap — at least, they’re not willing to talk about it. The next level of work that needs to be done is to implement this in each one of their facilities.
Some of the vendors are easier to work with than others. Epic was really early at the forefront in terms of sharing data with Care Everywhere, but they’re kind of like Apple. If they can control both ends of the transaction, it works great. But it used to be if you tried to get data from a facility that doesn’t use Epic, it didn’t work well. I think that’s getting better too, because we’re getting better with the standards about the documents.
I had a conversation with a friend of mine who works at an Epic facility and I asked, ‘if we send you a CCD, what format can you use to ingest it?’ He said, “If you work off of the Carequality standard, they’ll be able to bring it in.’ And I thought, okay, now we’re going to the lowest common denominator of CCD formats. But even with CCDs, it depends on what the care transaction is (and that’s my term). Is it a referral? Is it a transition of care?
And so you can create the CCD sections differently. You can include or exclude sections depending on that, but the one thing that stays constant is the PAMI piece — the problems, the allergies, medications, the immunizations. All of that is codified data that can be ingested pretty quickly, but somebody still has to reconcile that inside the EHR framework.
That’s where my bias comes in. If you’re in a facility and you query out to these other facilities and get back four different CCDs from four different facilities where this patient had care, you’ve got to review and reconcile four different CCDs. But if you ask for the same data from an HIE that has a clinical data repository, like ours does, we can hand you back a multiple-facility CCD that maintains the purveyance of where that data comes from. It’s been normalized, it’s been de-duped, and you have one source.
Interoperability isn’t like the Lord of the Rings. There isn’t going to be one ring that controls all. I think that it’s going to be more like Six Sigma, and it’s going to depend upon the need at the time and the market. I’ve had conversations with folks who are supporting physician practices out in the hinterlands of Montana and South Dakota, and direct messaging works really well for them. That’s what they have, and so they make it work. It would be like if I had a skateboard and a bicycle, which one would I use? I’m going to use the bicycle. It’s easier to ride, I probably won’t fall of it if I’m careful, and it’ll get me farther. We use the tools that we have, and we need ways of sharing those successes — what’s helpful, what’s not helpful, and how can we make it better.
IHIE is a member of DirectTrust. I knew Dr. David Kibbe really well, and I got to meet their new CEO [Scott Stuewe] when I was at ONC. I’ve worked on a couple of advisory groups with them and it seems to work, but does it really work for everything? It could, but the world’s got to be rewired. And so in some ways, interoperability is like healthcare. It’s been growing organically within the markets over time, and to undo it and put in something new will take time and cost money.
Gamble: It seems like we’re still seeing this mentality with any announcement regarding interoperability like ‘okay, this is going to be what drives it forward.’ It’s just not that simple.
Christian: I’ve spent a lot of time theorizing on this, and I think I understand the answers in TEFCA (Trusted Exchange Framework and Common Agreement) — or at least, the ‘TEF’ piece of it, and what the intent is. I’ve tried to put myself in the position of people in the federal government who are having these conversations. What they’re hearing is, ‘I don’t have a standard way to connect to the highway. I’ve got this vendor who wants to do it this way, and this vendor who wants to do it that way.’ Most health systems I know of don’t have one EHR. They have an EHR and they may have other modules. They may have a separate PACS system. They may have a separate lab system. They may have other ancillary systems that they’ve integrated internally. But to exchange data externally, if you’re a Cerner shop, it’s been CommonWell. If you’re Epic, it’s been Care Everywhere. If you’re communicating with another Epic site, it’s Carequality. But to get to a broader community, people want one way of doing it. I understand that. Particularly if you’re a physician practice — you just want to plug it in, but the question, is what do you plug it into? There’s a lot of work that has to be done there.
In the first version of TEFCA, one question I had was that the queue-ins are going to be required to do a massive amount of work, because there are a lot of requirements. If you’re going to a queue-in, you need to be able to do a litany of things, and a lot of the things they’re required to do is gather up all the data from everywhere the patient has data if it’s a broadcast query. Bringing all that data back in, de-duping it, normalizing it, and turning it around in one CCD back to the requestor — that’s a lot of work.
As far as I know, I don’t believe that you can normalize and map data in real-time with just technology. We do it every day, but we’ve got people that are lab techs, that are rad techs, that know the terminology, and have worked in the bench and in the ancillary departments for years. They know how to map these things into a standard set of terms. I’m was an X-ray tech for the first 14 years of my career; I can give you a dozen ways that a chest x-ray could be ordered. You can’t give a computer system a dozen ways to display those kinds of results. Physicians don’t want to look at 12 different lines. They will look at one line and get the result.