Capitol Health Associates, a Washington, D.C. consulting firm, is involved in myriad projects and initiatives around healthcare IT and value-based care delivery advancement. One of Capitol Health Associates’ senior leaders is Craig A. Jones, M.D., a principal in the firm.
As the firm’s biography of Dr. Jones states, “Craig A. Jones brings his extensive experience at the county, state, and national level with value-based health initiatives, population health programs, and use of data to guide operations into the design and capabilities of the health system platforms as well as supportive services offerings. He is also in an advisory role with the Health IT Resource Center in the Office of Care Transformation, within the Office of the National Coordinator for Health Information Technology. In this role, Dr. Jones provides assistance to states and federal partners related to implementation and use of health information technology to support the expansion of value-based payment models, delivery system transformation, and population health initiatives.” Among other activities, “Dr. Jones has served on several committees and workgroups including the Institute of Medicine’s Committee on Core Metrics for Better Health at Lower Cost, the Learning Healthcare System in America, and the Roundtable on Value and Science Driven Healthcare. Prior to working in Vermont, he was an Assistant Professor in the Department of Pediatrics at the Keck School of Medicine at the University of Southern California, Director of the Division of Allergy/Immunology Residency Training Program in the Department of Pediatrics at the Los Angeles County + University of Southern California (LAC+USC) Medical Center.
Recently, Dr. Jones spoke with Healthcare Innovation Editor-in-Chief Mark Hagland regarding his participation in the work to modernize the Idaho Health Data Exchange (IHDE), whose website describes the organization thus: “Committed to delivering trusted health information services, Idaho Health Data Exchange helps Idaho communities create lasting system-wide improvements in the value of patient care. Located in Boise, Idaho, our staff consists of an ever expanding, dynamic team whose primary goal is to improve coordination of care and reduce healthcare costs for all. Idaho Health Data Exchange (IHDE) is a non-profit 501(c)(3) company and Idaho’s statewide Health Information Exchange. We are dedicated to meeting the needs of healthcare providers and ensuring that Idaho’s citizens receive the most effective health services possible.”
As IHDE’s website states, “To achieve these goals, IHDE works with a wide-array of stakeholders to actively deliver best in breed technology infrastructure. We provide access to reliable data and information, combining traditional healthcare data with other data sources to help address the medical, behavioral, and social needs that influence the well-being of Idahoans.”
As of September, IHDE had 2,862 total active participant users; had managed a total of 5,462,283 inbound clinical messages; and managed the sharing of records for 3,097,600 total search patients.
Below are excerpts from the recent interview with Dr. Jones.
Tell me about how you and your firm initially became involved with IHDE?
I was working with the Office of the National Coordinator for HIT. I was working with Idaho in terms of state innovation models. They had a transition and wanted to rebuild. So they brought Hans [Hans Kastensmith, who is CEO of IHDE] and myself and the executive leadership team to turn around IHDE.
And, at the heart of this story, identity resolution was one big issue. The turnaround of IHDE was a really great opportunity for us. I’ve been working on health service reform, population health, etc., for 25 years. So when we walked in there, we wanted to see if we could turn IHDE into a driver of value-based care delivery. The providers needed data exchange to do that. And while IHDE was somewhat known, there wasn’t trust in it. So that was our starting point.
And one of the things we ran into at the beginning was that the identity of people hadn’t been resolved; there were large numbers of duplicates. They did have a master person index, but it wasn’t resolving identity issues. And as you know, identity resolution is the first brick in the building.
And that’s when we started to look at the vendors out there to help us with identity resolution. And in the end, there are a lot of groups that have the technology to do the matching that you need; but you’re always left with a number that can’t be resolved by machines. The goal is to make that as small a number as possible. And Greg [Gregory Church] will follow up, but they really appealed to us because of the service end, the ability to take on those tasks, because we had very slim funding and staffing, and we couldn’t be hiring FTEs to do identity resolution. So that led to the first round of work with 4Medica [the Marina del Rey, Calif.-based 4Medica clinical data management and healthcare interoperability software company] that really set us up to do the downstream work.
Were you facing a higher proportion of unresolved identities than other HIEs?
I work with HIEs and data aggregators from all across the country; and like anything else, HIEs vary. If you look at the good ones like the Oklahoma HIE, they had achieved much higher rates of identity resolution than we have. Others might have been similar to us. But if you’re failing in identity resolution, you’re failing on delivery. Because getting people to share and use data relies on trust; it’s not about the technology. And if you can’t get people to trust you, you’ll never build robust data exchange. And we were on the lower end when we came into IHDE, frankly, and if we hadn’t been able to fix this, we wouldn’t have been able to go further.
What was the process like, as you and your colleagues went in to rework data management processes?
They got access to all the records from our HIE platform vendor; and it was 14 million records. Now, remember, that’s records, not individuals; there can be 10 records for an individual. So the first step was going through the process and linking the records. So, OK, this person has five records; let’s link them. So they ended up with about two-and-a-half million identities. Then they ran that through their service in the MPI [master person index], and that led to about 2.2 individuals. So from 14 million to 2.5 million to 2.2 million. And that service that’s now working every day, lowered potential duplicates by about 10 percent.
What came next?
4Medica then sets it into a day-to-day production-level process for us. We have about 100,000 records coming through the HIE every day. That goes through the MPI process with 4Medica. And now we’re at about a 3-percent potential duplicate level. And that 1- to-3-percent level gets down to about where the high-performing HIEs are. And of the small percentage that the system can’t figure out, 4Medica can resolve a number of those just by human beings looking at the records. Back in the 1990s, while I was working out of USC Medical Center, I built a population health database for Los Angeles County, and every morning, I had to come in to resolve things, and if you can’t resolve identity, you won’t get far.
Can you describe other areas in which your team has been making progress?
I can talk about two broad areas. One has to do with the exchange platform, which is delivering records and messages to EHRs [electronic health records] across the state, populating the HIE portal. So it’s exchanging records across organizations and sites, for care management. And we’ve restructured a lot: things like standardizing data and completing things with standardized coding; we have a terminology engine to support that. And we’re mapping and flagging data, to understand the organization, the site; if you don’t get that patient or person to provider to organization to service relationship clearly established, then it limits the usefulness of the data downstream, whether for care management or for things like analytics and quality improvement. And if you can’t do that, you can’t subsegment your population. So we built the core components you need to establish those relationships. And then our HIE vendor has an MPI identifier inside it; we had to work with 4Medica to create accurate identification inside the portal.
So now when people look inside the portal, they have much more complete, accurate data to use. And we have a large number of people using the portal every day. I’ll check the dashboard. It’s grown steadily over the last year, since we put a number of these things in place. A lot of HIEs had huge drops during COVID; we had a brief blip, but then it went back up. We have a good user base, and it’s growing, and we’re told it’s because we have reliable data.
And in addition to sending the improved data to the exchange platform, we also send it into a FHIR-enabled enterprise data warehouse. And we’re able to aggregate large-scale clinical data; but also with social needs data. And it’s integrated into our system. And so if you need to do an assessment, you can go right into the social needs database and generate closed-loop referrals. We’re working with Aunt Bertha, which is the social needs platform. There are several big social needs platforms, and Aunt Bertha is one of them. And we’re working with them and 4Medica; we’re going to have an identifier system driving both the clinical and the social needs data. And once you get that identity resolved between Aunt Bertha and 4Medica, we can merge the data in our data lake. And we can combine clinical and social needs data for population health work and population health improvement. We have an analytics platform, and we feed data into that. We convened a payer group, which had never been done in Idaho; and they asked us to be a convener between them and the providers, and we have a collaborative setup and they’re working with secondary data for population health. And that depends on delivering data to them that they can trust; and the first step in that trust was identity resolution; the second, being able to link all the different data types. And then can you deliver it in a form that they can use? And can they use it for value-based payment or quality measurement? It has to be complete, reliable, trustworthy data. So that’s what IHDE has done—to put all those steps together to deliver complete, aggregated, trusted data. And, because we’ve made progress in all those areas, we’ve been able to build trust among providers and payers in the state.
What will the next two years look like for IHDE?
We’re continuing on this roadmap and continuing the integration of all these different types of data. And it means a consent management system that allows us to exchange highly sensitive data, with granular control. Probably first-level consent within six months for trusted exchange for sensitive data among behavioral providers in the state. The same with the integration of the Medicaid and commercial-payer eligibility platform, so people can get eligibility here; it will be part of what’s available in a patient’s record. And payers really appreciate that, not just providers. If you make it easier for them to do their work, they can work more easily with providers. We have an image exchange function, that will be elevated to provide large volumes of exchange for health systems. And we’re working on a number of public health use cases. For example, someone wanted to understand the rate of myocarditis in Idaho following the onset of COVID. So we’re working on an ongoing series of requests in that regard.
So the roadmap is one part of this. And another is looking at regional levels of efficiency; we’ve been talking to the HIEs in Washington, Oregon, Montana, Utah, and other states.
