Last month, the Government Accountability Office (GAO) released a report, “Nonfederal Efforts to Help Achieve Health Information Interoperability,” in which representatives from 18 health information exchange (HIE) initiatives described a variety of efforts they are undertaking to achieve or facilitate electronic health record (EHR) interoperability. The stakeholders involved concluded that to date, interoperability has remained limited.
The stakeholder and initiative representatives GAO interviewed described five key challenges to achieving EHR interoperability, which are consistent with challenges described in past GAO work. Specifically, the challenges they described are (1) insufficiencies in health data standards, (2) variation in state privacy rules, (3) accurately matching patients’ health records, (4) costs associated with interoperability, and (5) the need for governance and trust among entities, such as agreements to facilitate the sharing of information among all participants in an initiative.
The report also found that changes to the Centers for Medicare & Medicaid Services (CMS’s) Medicare and Medicaid EHR Incentive Programs would also help move nationwide interoperability forward. Specifically, representatives from 10 of the initiatives noted that efforts to meet the programs’ requirements divert resources and attention from other efforts to enable interoperability.
After the report’s release, a couple of noteworthy health IT experts took to Twitter to share where they thought the GAO research fell short. One of these two people was noteworthy health IT leader Arien Malec, vice president of data platform and acquisition tools for RelayHealth (the Alpharetta, Ga.-based McKesson business unit that focuses on improving clinical connectivity) and former staffer at the Office of the National Coordinator (ONC). Malec recently spoke with Healthcare Informatics Senior Editor Rajiv Leventhal about what he specifically disliked about the report, what the industry can do better to achieve interoperability, and what he thought about the meaningful use final rules released last week. Below are excerpts of that interview.
What about the GAO interoperability report particularly bugged you?
Well, first off, there were a couple things I liked about it. For one, there was unanimity around the need to get consolidation around value-based payments and clarify meaningful use so we’re not criticizing providers. My general critique of the report was that it was a superficially descriptive report in that it talked about a bunch of initiatives but didn’t help get a layer down in terms of where policy needs to go and the drive towards person-centered interoperability. It wasn’t that I thought it was incorrect or not factual or came to terrible conclusions, but I saw it as a missed opportunity to get deeper into driving what would be required for true person-centered interoperability.
What do you see as “true person-centered interoperability?”
By person-centered, I mean that a person has access to his or her complete record, that it’s available to providers to provide care, and that it’s available to the care team to drive improvements in health. This doesn’t just mean around a person-mediated exchange, but the ability for all the members of the care team to have access to the complete record. What’s important about that is there have been a number of initiatives that have driven EHR-to-EHR specific interoperability, meaning I can open up an interface and interface to my EHR. That’s fine and necessary, but it’s not sufficient for person-centered interoperability. It was a missed opportunity to look at the initiatives. This person-centered perspective is the view that the ONC Interoperability Roadmap also has as its definition. You’re starting to see some consensus around interoperability beyond EHRs and interfaces, and more towards what information is required to provide optimal care.
There have been some criticisms that ONC’s Roadmap is too broad and not actionable enough. Do you think that’s fair?
Generally, McKesson’s position here has been that the right role of government is to set a clear destination, put in place the right levers, particularly in terms of payment policy, not micromanage, and not define the ‘how’ too much. Our perspective is that with a clear enough picture and destination, and with efforts like CommonWell and the Argonaut Project, I think we have proven that as an industry we can rise to the challenge and do the work to improve interoperability. I appreciate the picture that was painted with the Roadmap of where we need to go, pointing out some of the federal action. And I’m starting to see HHS double down aligning on value-based payment on a consolidated set of measures, making sure that those measures are more outcome-centered rather than process-centered. Those are all helpful steps. I don’t think that the role of government in this area is to move all of the pieces around to achieve interoperability. In that sense, I appreciated the Roadmap.
How would you state the level of interoperability in the U.S. today? Are we moving fast enough?
I would look at the banking system. It went from paper ledgers to electronic systems in six or seven years. Looking how banking systems evolved from 20 to 30 years ago, we didn’t go from electric banks to ATMs that everyone could access to online banking to Apple Pay in five years. It just didn’t happen like that. Where we are now is disappointing providers and I acknowledge that fully. They have an expectation that they have in consumer electronics that the data is right there and that the usability of interoperability is great. They have that expectation, and they have a right to expect that.
We have accomplished a lot, though. What’s good is that there are the 15 or so initiatives that were surveyed [in the GAO report]. We have made progress, but are just not moving fast enough from a provider perspective, and that’s a perspective I absolutely agree with. CommonWell has announced that it’s trying to get to 5,000 provider organizations signed up by the end of the year. Cerner folks are talking about mass deploying, and folks are talking about that at McKesson too. A number of organizations are rising to the challenge. We are in that place where we have a lot of electronic records and are starting major initiatives to connect them all, but we are in the middle of it and looks messy right now.
Whenever we have this interoperability discussion, the CommonWell/Epic situation seems to be at the thick of it. Is this a big problem?
First of all, CommonWell right now represents something like 75 percent of acute care systems, with the major hold out being [Epic]. I don’t believe that Epic needs to join CommonWell for it to be successful. If you look at how banking networks evolved, Visa started out highly associated with Bank of America, but then you had MasterCard because someone had to create a competing network, but now we don’t care, we just swipe our card and it works. I think we will get there in healthcare. That might mean that we put in place a SMART on FHIR [Fast Healthcare Interoperability Resources] app for Epic hospitals who want to join CommonWell to use. It might mean that Epic still doesn’t have a nationwide record lector service, but they have the ability for two hospitals to query each other. It might mean that Epic deploys a nationwide record locator service and we figure out how to bridge them. There are a bunch of options for getting there; I am not worried that Epic not joining CommonWell means that CommonWell cannot be successful. CommonWell has enough folks in already to drive success. If you look at any other network that has grown in a vertical sector, and financial services is a great one, it’s not surprising at all.
What do you make of this notion around information blocking?
Information blocking is defined as an EHR vendor that refuses to open an interface. I have not seen that anywhere. But there are other definitions of information blocking. I have seen it where the cost and complexity of the interface and the ability of the people involved to deliver on the interface is harder than the system can bear. It can be too hard, though people are trying. That happens. We have cases where because it’s so complex it costs a lot, and the cost is too high. That happens. And we do have cases where there are local areas with more of a monoculture where the EHR vendor has done a lot of work to make sure that it makes it easy to connect to the EHR, but it creates a strange incentive to not do cross-vendor interoperability. Instead, in these situations, you convince your sister system to buy that same EHR. I don’t blame the vendor involved, but it happens, and we need to make interoperability better, faster, and cheaper in order to get where we want to go. The major issues when people complain about information blocking are associated with the total cost of care of interoperability, and that’s exactly what our work in Argonaut and CommonWell are intends to go after.
So what can providers and vendors do to achieve true interoperability?
I have a general perspective; I don’t make a distinction between leading and following, I make a distinction between leading and accepting the world as it is. If you’re not leading, you’re accepting the world as it is, and if that’s the case, you can’t sit back and complain about it. Start getting involved in trying to improve the world if you don’t like how it exists. That’s the number one best practice. If you’re a vendor organization, I would say join the Argonaut Project, join CommonWell, implement SMART on FHIR, or if not, do something to improve the state of interoperability. If you’re a provider organization, there are the same kinds of initiatives to participate in. A number of providers are participating in Argonaut offering to be demonstration sites and get technology up and running. For providers, it’s about defining a clear strategy where you move into and understanding where you have gaps in the world that impede your ability to get there now. Many organizations want to get to value-based care and payment, and they need to recognize that the total cost of interoperability is an obstacle. Rather than sit back and complain, they need to work with vendors and with these initiatives that are trying to reduce the cost. Roll up your sleeves and get involved somehow.
Finally, what were some of your quick takeaways on last week’s meaningful use final rules?
In general, I wasn’t surprised by them. The certification requirements represented that ONC listens and [outlines] a more achievable set. They also doubled down on APIs which I hope creates a consensus to move to FHIR-based APIs that align with the Argonaut work.
I am disappointed that CMS dropped the view/download/transmit patient engagement measure to one patient [for Stage 2]. I believe that getting to a threshold percentage drives systems to change their workflow to enable better patient engagement. That was a missed opportunity. But I applaud them to make measures simple in general.
I do think that on this stage, providers will be massively confused between meaningful use and the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA). Starting in 2017, eligible providers will have to demonstrate achievement of the value-based modifier in meaningful use and the other components of MACRA. People will get confused about that. What do they do? Meaningful use, MACRA, or both? CMS needs to clarify that and fix it. I think the MACRA shoe will be the second shoe to drop. I appreciate Sen. [Lamar] Alexander’s view on the Merit-based Incentive Payment System (MIPS). It should be simpler at a time when it’s too confusing. But there is also the understanding that we need to keep moving. It’s the debate of being too fast and too slow at the same time. At some point, you have to decide which of those worlds you are most concerned by.