State governments and statewide HIEs are starting to make progress on moving to digital quality measures based on clinical data rather than just claims data. Tom Curtis, director of state affairs for the National Committee for Quality Assurance (NCQA), spoke with Healthcare Innovation about how North Carolina and Rhode Island are leading the pack by first remapping data to FHIR format so it can be more easily queried. 

Healthcare Innovation: I want to talk to you about states’ roles in building digital health infrastructure. I noticed from your bio that you came to NCQA from the state of Michigan. Were you working on building digital health infrastructure there before you came to NCQA?

Curtis: In Michigan, I had a couple of roles over the years. One of them was leading the state's state innovation model or SIM grant, which was $70 million back in 2015 and 2016 from the Center for Medicare and Medicaid Innovation to initiate or support multi-payer alignment for quality in states. One of our efforts was leveraging the state's health information exchange to gather clinical data from providers and put it into one clinical data repository so that quality metrics could be derived using clinical data to inform where there were really important quality improvement needs in the state. That was the vision. We didn't get to that vision. 

Michigan is still not at that vision, but we did make progress by putting in place requirements for providers to connect with the health information exchange and send data through the HIE, so that we could essentially start to know where this is not happening.

HCI: So you got a close-up view of the challenges of making that happen. It's not like a one-year project…

Curtis: That is correct. It is a definitely a long-term effort in order to really see success and accomplishment in that vision.

HCI: Does a lot of this work involve statewide HIEs earning the NCQA Data Aggregator Validation status and then them playing a central role in these processes?

Curtis: Yes, actually, Rhode Island and North Carolina are the two states we have become most familiar with thus far.

HCI: In all the states where the HIE is playing the role of what they're calling a health data utility, this seems like a next logical step. 

Curtis: That's correct. The current state, even in Rhode Island and North Carolina, is that these HIEs are collecting and sharing CCD [Continuity of Care Document] data, which isn't necessarily a standardized data format. So it it doesn't get us collectively to the level of interoperability that would constitute a successful public health intervention. We're not quite there yet. But what Rhode Island and North Carolina are starting to do is take that CCD data from all of their providers and remap it into FHIR format, which is a consistent data format.

You can't say digital quality measurement is reducing burden when all we're doing is adding extra work to the effort. So it's not at the burden reduction stage yet, but they are remapping it, and the use case for that is they are having clients of the HIE asking for bulk data in FHIR format, so they can query those data. It's almost like speeding up medical record review, if you want to think of it that way. You can just start searching a huge pot of FHIR-formatted data to find out who got the mammogram, who didn't get the mammogram that should have, and how do we focus our care management efforts toward those gaps in care. 

The other use case for remapping data into FHIR is that it helps organize and understand and evaluate data quality. States need to have all of the data for all of the people all of the time. It is difficult to really know how far along we are in getting all of the data from everywhere when it's not in a standard format. This FHIR mapping exercise does help. States are envisioning this helping them understand the universality of clinical data that they have or don't have, so that they know when they can begin using it as a policy tool.

HCI: Also, CMS and NCQA have laid out these roadmaps of moving to digital quality measures. So the health plans and the states are seeing that coming down the pike, and they want to start making progress toward that, right?

Curtis: That's correct. That's another reason that HIEs are filling this gap is to see what the future might look like. I am at the Medicaid Enterprise Systems Conference right now, and what I'm hearing is that one of the barriers is that providers don't have inbound/outbound FHIR data capabilities. There's not a cost-efficient way for the EHRs to have the interface to send data out in FHIR format or even receive it in in FHIR format. So that is another roadblock in actual interoperability.

HCI: Let’s talk about the work taking place in North Carolina first. There’s an HIE Medicaid Service Program, which is an effort between the HIE and the Department of Health and Human Services Division of Health Benefits. They have digital quality measures, health-related social needs screening, and care management data projects all under that umbrella, right? Do you know what the timeline for this is and how far along they are?

Curtis: The care management project I would say is furthest along, and it was one of the first projects they wanted to do. One thing we see pretty consistently across states is that when they've got an HIE tool available, their first go at using that tool to help improve care is to promote better care management. 

The one that we're most excited about at NCQA is the quality measurement piece, and they're probably three to four years into that effort. They started with controlling high blood pressure as one of their key measures, and over a three-year period, they improved the performance rate calculated using clinical data from the HIE to almost become validated and equal to the HEDIS rate pulled using administrative data from their health plans, which is pretty phenomenal.

Now we can start to trust these results that we see with the clinical data. After that, they required Data Aggregator Validation, and the HIE is going through the FHIR Data Aggregator Validation effort, so they're wanting to analyze each of their data streams relative to meeting FHIR format requirements, and then identify data streams where that's not happening, so they can focus more of their work on on where the data quality needs are. I's not just instituting trust in the data; it's also directing them to where data quality improvements are needed, and they're standing up a separate data quality improvement initiative on their own based on the DAV results.

HCI: Does this also allow the the HIE to share care gap reports with the providers?

Curtis: It does. That's more like what Rhode Island is doing right now. They are working with CRISP Shared Services, which is the HIE backbone for eight other states in the country, including Alaska and Rhode Island. One of the shared services that CRISP offers is those care gap reports being provided to the providers on their patient populations.

HCI: Is Rhode Island on a similar path that we just talked about with North Carolina. Are there some differences?

Curtis: There are a couple. They're going along a similar route. One thing about North Carolina is the HIE is essentially owned by the Medicaid agency. So all funding is Medicaid funding. The HIE efforts are focused on Medicaid beneficiaries, and that allows progress in a different way because the state doesn't have to negotiate interests with private payers or commercial payers, right?

The downside is it's not multi-payer; it's just one population and one payer. That HIE was instituted in state law in 2015. One thing we see across states is when the state has a law in place establishing the HIE and giving it authority and giving it secondary data disclosure allowances, that has these downstream implications on how effective the HIE is.

HCI: But as you said, there are limitations, too, because it only involves Medicaid, right?

Curtis: One is that North Carolina was an opt-out model from the beginning. Rhode Island started as an opt-in model and only recently changed to an opt-out model.

The other difference is that Rhode Island is multi-payer, so the HIE is funded not just by Medicaid. There are also commercial payers and private payers that are on board and provide funding to the HIE. The HIE is faced with navigating competing interests in some cases. The state does have a role in facilitating, governance and decision-making. Rhode Island has done this cool thing where in exchange for commercial payer buy-in and support for the HIE, the state is negotiating providing those private payers with access to the Prescription Drug Monitoring Program (PDMP), which had a lot of its roots back in the opioid epidemic days. So that's kind of a unique way that Rhode Island is trying to negotiate that multi-payer alignment around the HIE.

HCI: I understand that Rhode Island is trying to move to electronic clinical quality measurement as part of an overall shift to value-based contracts…

Curtis: Yes. The reason why digital quality measurement using clinical data makes a lot of sense for risk-bearing at the provider level is because it's using provider-level clinical data used to inform care and care decisions for measurement. That is how quality can be measured at the provider level and incentivized at the provider level — whereas with the claims process, there's a disconnect, right? There's a bunch of information that clinicians are using to inform care at the provider level, then they submit a claim, and that is separate and apart from everything they just did with their medical knowledge and expertise. So right now we have a disconnect between how we're measuring quality and how we're providing quality care. Bridging that gap using clinical data is what some believe will be the key to unlocking this clinical level value-based payment.

HCI: How far along is Rhode Island in this work? 

Curtis: Rhode Island released an RFP scope of work for a quality reporting vendor, who essentially would be taking CCD data from the HIE, remapping it into FHIR and then using a clinical quality language (CQL) engine to derive digital quality measure rates from that formatted FHIR data. That is the furthest along we've seen. What that means is they got state backing to spend money on this type of effort to drive digital quality measurement. North Carolina is more on the front end of creating the scope of work to do about the same thing.

HCI: Does NCQA have a role to play in helping these state groups develop their roadmaps?

Curtis: We believe we do. NCQA digital quality measurement teams mostly engage with HEDIS rules and HIE vendors, so we have insight into how they operationalize state requirements and how they are wanting to be innovative and maybe are being held back by the risk-averse state government people. We can use that insight to meet with North Carolina and Rhode Island and share what we’re hearing and what NCQA thinks about these different kinds of timelines. We can give them general reactions to that, but we can’t weigh in on what their RFP is going to look like. We have to set a kind of firewall.

HCI: Do you also explain to them how this work fits with other emerging interoperability standards such as USCDI?


Curtis: Yes, we definitely do. We frequently meet with states and we'll provide that kind of education. We find states that are asking similar questions or that are at similar stages in their journey, and we bring them together, and they start learning from each other. That’s really where states thrive is learning from their peers on how to test something or how to implement something. We brought together Rhode Island, North Carolina, and Ohio recently.

Ohio has the Clinicsync HIE and they're trying to brainstorm on the front end how to work with Clinicsync to  move in the same direction as Rhode Island and North Carolina. From a state perspective, financing is the name of the game for any of this. States have to balance budgets every year, so there has to be a case for this that is bigger than schools and roads and fire and police. It’s a tough sell.