True interoperability “will make a 20, 30, 50 percent change in the quality of medicine that we provide,” a health system leader says. But getting there will require ample funding and ingenuity.
Some of the most transformative strategies in health care, including the transition to population health management and integrated care delivery networks, rely on achieving interoperability among stakeholders. Although significant progress has been made in recent years, the obstacles remain formidable.
For this edition of the Healthcare Challenge Roundtable, leaders with a medical group, health system, and health plan shared their thoughts on the latest trends, challenges, and signs of progress in the interoperability push. Participating were Keith Fernandez, MD, national chief clinical officer, Privia Health; Stan Huff, MD, chief medical informatics officer, Intermountain Healthcare; and Kim Sinclair, chief information officer, BMC HealthNet Plan.
How would you define true interoperability in health care?
Stan Huff: The best way probably is just by analogy. With a gaming system, I can buy a different controller than the one that came with my game. I just plug it in, and it’s aware of what system it’s being plugged into. I might need to tell it what my preferences are, whether I’m lefthanded or righthanded. But it works out of the box.
In health care, the analogous situation is that I could have a piece of software, and it can be used in anybody’s electronic health record or any health system. It might need to be configured a little bit to know whether I’m an obstetrician or a cardiologist, for example, but it just works. And it works because it knows and understands the data, and the workflow of the clinicians. That’s true interoperability.
Keith Fernandez: To me, interoperability from a clinical standpoint refers to the ability to exchange information between data platforms in a way that makes that information first, available to providers; second, easily found; third, in the workflow of the user; and finally, able to trigger an outcome to create change for the better.
There are people who judge progress towards interoperability a little bit differently. With many organizations having one or more of those capabilities but very few having all four, they would say if you had two of those things, you’d probably have a good case for having significant interoperability. That’s a much less clinical definition because I’d say if I have two of those things, but I can’t impact an outcome, that’s great but it didn’t change anything that happened with the patient.
Kim Sinclair: I think interoperability in health care is the ability to share and exchange data in a way that makes health care more efficient and meets the needs of the patient.
We will know we’re there when the discussion shifts from what data is available and what system it’s in to how we maximize the data and what we can learn from it. But I don’t feel like we’re there yet. It’s too much of a focus on specific data elements and what you can provide.
What signs of progress are you seeing in the effort to achieve interoperability?
Fernandez: The companies that are succeeding here have a vision of the future that requires having a complete picture of the patient and making an impact on outcomes. That’s what’s driving the most successful groups, the vision of what their job is—not just to sell an EMR or technology, but to sell a platform that can improve the health of individuals as well as populations.
At Privia Health, for example, we look at information a little bit differently. Is the information available, and can we locate it? But also, is it pertinent? One of the important distinctions is that there’s a huge mass of clinical data that may be very, very important in one situation but not another.
For example, you don’t need to know the minute-by-minute blood pressure obtained in the ICU when you are seeing the patient for follow-up in an ambulatory clinic. When you’re looking at a population and you want to ensure that all of the care needed in both sickness and health has been accomplished, you do not want to sift through mountains of other information. That’s the kind of functionality we’re trying to accomplish.
Huff: One of the most exciting things is the HL7 FHIR [Fast Healthcare Interoperability Resources] standard. It’s a standard for how you store and retrieve data in a patient database, which could be an EHR or a quality registry or a report to the CDC. Closely related to that is the HL7 Clinical Information Modeling Initiative. Full disclosure: I’m the co-chair, so I might have a favorably biased opinion of that.
There’s a very new initiative called the Clinical Information Interoperability Council, where a group of interested parties are trying to get together across all of medicine to say: When we share data, this is the logical structure of the data, and these are the exact codes that you’ll use to represent clinical findings like blood pressures, heart rates, serum glucoses, the medications that the patient is on.
Sinclair: In Massachusetts, with the implementation of healthcare reform and the shift to accountable care organizations for Medicaid in March 2018, we’re having conversations—within our own health system, Boston Medical Center Health System; and with our ACO partners, MassHealth and the Massachusetts eHealth Collaborative—to implement ADTs [patient administration messaging] for real-time event notification for ER visits, inpatient admissions, and discharges. Stakeholders from around the state are involved, as there is an important business need to exchange this information.
It may be a small step, but it will have big impact. Providers and health plans are going to have the information they need in real time instead of waiting for a paper census or finding out after a patient is discharged.
What are some of the most daunting obstacles to achieving interoperability?
Huff: The single biggest thing is the lack of incentives for systems to be interoperable. The current fee-for-service payment structure actually incentivizes organizations to not share information—to repeat studies rather than ask for results of a study that was done yesterday in a different emergency room.
There are also other business reasons. Although they protest loudly, there are companies that make a lot of money making interfaces. The fact that each one is different and that they’re doing individual mapping of codes leads to revenue for a bunch of people.
The second thing is the need to coordinate across a number of groups. There are groups that are working on interoperability having to do with quality registries, and people who are working on it to be able to submit data to government agencies. There are activities within every company that vends software, every professional association. To get true interoperability, you have to coordinate across those organizations. Just knowing what everybody’s doing and trying to ensure standardization, while not obstructing the good work that they’re doing, is a challenge.
Sinclair: The money and the time it takes to transition to newer technology platforms that might be out there is daunting to a lot of organizations. Even if those platforms exist and are in place, there is the challenge of getting everyone doing things the same way and speaking the same language.
Many organizations have other priorities, including keeping their business running, that prevent interoperability from being the focus. I think interoperability will have great benefits in the long term, but it just isn’t everyone’s focus right now.
Fernandez: It’s kind of a stark truth that a lot of the stuff we’re seeing that promotes interoperability and the exchange of information is aspirational, and not real. It muddies the conversation a little bit, it diverts a lot of money into places where the money doesn’t belong, and it is exhausting the workforce, including doctors and CMOs. I’d just caution everyone to be very careful when there appears to be a wonderful solution available. Make sure that it’s been tested somewhere in an environment like the one you live in before you buy it.
Generally speaking, it’s all very expensive stuff. In my career, I have overseen millions of dollars spent on trying to make these data sources interchangeable, and I’m going to guess that the vast majority of those dollars did not lead to any conclusive improvement.
I participate in population health forums in a number of different settings, and I think there is a very common perception among population health providers that most of the stuff that they’re sold doesn’t yet work, or works only in a specific environment. It becomes very, very difficult to go to your CFO for another several million dollars when your last request didn’t pan out and there is little proof that the next one will be any different.
Should the public or private sector, or a combination, be leading this effort?
Fernandez: We all need to be in it. But it’s very hard to have a trusting relationship with your most intense competitor—people who perhaps want to take your clients, and suddenly you have to develop a trust that you can work together on interoperability, and you can exchange information safely. Trust requires each group to make sure their competition succeeds.
That’s a tough one, but I think it’s necessary. The thing that binds us in that is the patient. If you’re not really focused on providing high-quality care for your patients, then you’re not doing your job.
Sinclair: The government may need to nudge the discussion along in some ways. But I think if it’s truly going to work and truly be transformative, the stakeholders across the industry really need to be focused on this as a shared goal and priority—possibly working with entrepreneurs from outside the industry who can apply different expertise to help solve the problem.
Huff: CHIME [the College of Healthcare Information Management Executives] recommended that we establish a new government agency that’s charged with creating healthcare interoperability. It would involve private organizations, the standards-development organizations.
The way the government is going about it now is not effective, and it won’t be effective unless they change. It’s just not specific enough, and nobody has authority. It’s not that they’re bad ideas, but it’s not constructed in a way that you’re going to get to the objective.
It would still be a public-private initiative, but a new, properly funded, and properly governed agency would be the very best thing to do. I don’t think it has a chance of getting through the current political process, but that doesn’t mean it’s not the right thing to do.
What level of collaboration among hospitals, medical groups, and health plans is necessary to achieve interoperability? Are you seeing any hopeful signs?
Sinclair: The first thing is that barriers and silos among the stakeholder groups need to come down. This isn’t about a competitive advantage if you’re a healthcare provider or a health plan. It’s just an important component of doing business.
I think the health plans and health systems, as well as larger provider groups and hospitals, are going to continue to make progress. But we need to find a way to get the smaller providers and independent providers involved. These include LTSS [long-term services and support] providers, long-term care facilities, and behavioral health providers. Without these providers, you don’t have a complete view of the patient and can’t deliver the best care or care management.
Health plans, ACOs, and delivery systems, as major stakeholders, need to think more about how to include these smaller, independent providers. We need their information and they need our information if we’re truly going to continue to improve quality and lower costs.
Fernandez: There are many collaborations going on. For example, there’s CommonWell and Carequality. One of them is backed mostly by vendors, the other is a combination of vendors and providers. Both of those seem to be good efforts, and they seem to be highly engaged in this proposition about benefiting the patient. Their goals are not necessarily to corner the market, but to get to interoperability for the benefit of the people we serve.
It’s a little bit market-specific, and a little bit about how you view your competition. We’ve got hospital systems in some of our Privia Medical Group markets that do want to share information with us. The goal of the system is the same as ours, which is to take really good care of patients.
For that to happen, the ambulatory care provider needs to have all of the information from the hospital, and in a timely fashion. When a patient gets admitted, the ambulatory provider needs to know right away so that they can add information to the care of the patient. When a patient is discharged, the ambulatory provider needs to know the moment that happens; in fact, we prefer to know even in advance of it happening, and there are tools that allow that exchange of information without going through a lot of human beings.
Huff: There has to be that kind of collaboration. We’re seeing that, speaking from Intermountain’s experience, because of some of the reimbursement changes that are coming about with accountable care organizations and healthcare bundles, where more than one organization holds the risk in providing for a population. Those changes create a financial incentive to provide high-quality, low-cost care across organizations. That pushes people together to pursue interoperability.
We’ve got a long way to go because there are a lot of challenges. People don’t trust each other in the relationship; they’re worried about their own business interests. It’s hard, but it’s obviously in the best interest of the patient, and there are some encouraging signs because of those new payment reform strategies.
Where do you see the healthcare industry in five years in terms of interoperability?
Sinclair: The health plans and large health systems will have made a lot more progress. There are many discussions occurring, and it’s a topic everyone is interested in, so that will continue to move it along. Here in Massachusetts there is a lot of focus on this topic, so I am hopeful we’ll start to see results in the next couple of years.
Once most organizations are sharing data, we’ll likely see the desire to expand on the data that is exchanged, as everyone recognizes the value. And then the focus will shift more to sharing outcomes data.
Fernandez: We’ll have pockets around the country where we are achieving significant interoperability, where the government and the vendors and the providers have gotten together and said: Enough is enough, we’re going to make this happen. Organizations like Carequality and CommonWell may be winners in this space in the long run.
We’ll be better off in five years, but we won’t be done. The current uncertainty in government with the withdrawal or weakening of value-based care programs, including a pullback on bundled payments, may slow us down and cause appropriate investment in technology to decelerate. That said, we will continue to forge ahead to serve our patients in the best way possible.
Huff: I can say where I hope we are. I hope that in five years, we’ve solved big areas of interoperability, that we can exchange lab data, we can exchange problem lists, we can exchange medication lists and prescriptions, we can exchange genetic data.
And we’d be working in specialty areas, so that instead of those large domains, we’re working very specifically on the data elements that we need to support cardiologists, obstetricians and gynecologists, physical therapy, diabetes clinics. That we’ve taken care of the core data elements, and now we’re going to work forever on the remaining 1 percent so that we get to the other important but less common elements, and interoperability around those areas.
Are there any additional thoughts you’d like to share about the significance of interoperability?
Sinclair: I’m a member of the HealthCare Executive Group, and we had our annual forum in Nashville in September. The topic of interoperability came up in every panel discussion that we had, from the opioid epidemic to provider directories to population health. This topic is getting a lot of attention, even though everyone may not be labeling it as “interoperability.”
One of the conference attendees actually counted the number of times interoperability came up during the first two hours, and he stopped counting after 25 mentions. That just shows how prevalent the topic is now in everybody’s day-to-day conversations.
Huff: I don’t think people understand the incredible impact that true interoperability could have on health care. People estimate that we’re killing somewhere between 200,000 and 400,000 people a year through medical error. If you compare that to the number of people who die in automobile accidents, we’re killing up to 10 times more people through medical error, and for some reason people don’t see it as an epidemic—we’re not investing even $5 million in interoperability solutions as a country.
It’s a subject that, somehow, we’ve got to approach in a different way to legislators and policy makers, to get them to understand that this is a solvable problem. How much money have we spent on research on seatbelts, shoulder harnesses, collapsible bumpers, collapsing steering columns, padded dashboards? Yet we’re not making even 1 percent of the investment that we make in that kind of safety equipment to prevent those kinds of medical errors.
Interoperability is the thing that will make a quantum change in the quality of medicine that we provide. Better medical schools, zero-harm programs—those things will have a measurable but small impact. Interoperability, where we could share knowledge as executable programs, will make a 20, 30, 50 percent change in the quality of medicine that we provide. And somehow people don’t understand that, and we don’t invest in it.
Nick Hut is managing editor of Leadership.
Interviewed for this article: Keith Fernandez, MD, national chief clinical officer, Privia Health; Stan Huff, MD, chief medical informatics officer, Intermountain Healthcare; Kim Sinclair, chief information officer, BMC HealthNet Plan.
Check out previous installments of the Healthcare Challenge Roundtable: