A recent Manatt Health white paper describes an eight-domain framework for the development of state health data organizations (HDOs) to inform policymaking and regulatory decision-making. In a conversation with Healthcare Innovation last week, Kevin McAvey, managing director at Manatt Health, spoke about the growing potential of state HDOs.
McAvey has been working on building state-level health data analytics capabilities for years. He started as an all-payer claims database (APCD) programmer at the Center for Health Information and Analysis in Massachusetts and was involved in developing CHIA’s cost growth benchmark reporting. He also has participated in national APCD organizations. He has been at Manatt Health for eight years, helping states across the country develop their health data ecosystems.
Healthcare Innovation: There has been an evolution in the health information exchange space in which the HIEs are describing themselves as state-level health data utilities. But are those organizations different from the health data organizations your white paper describes?
McAvey: Health information exchanges are continuing to evolve to expand the types of entities they are supporting. Historically, they have been at the front lines of facilitating clinical data exchange between hospitals, health systems, and physician organizations. But as HITECH funding winnowed and as the private market began increasingly serving these purposes, HIEs are evolving to think about how they can be resources and utilities, as they're increasingly thinking of themselves, to provide value-added information back to participants and new types of population health information back to states and regions and counties. It brings a very population health and public health focus to the work.
State health data organizations are a little bit of a different animal. But I could see these two different types of entities increasingly not only overlapping, but potentially even becoming similar or the same entity. State health data organizations help states to look at the overall performance of a healthcare system. They typically have access to pricing and cost data, which HIEs don’t. When you think about ACPDs, these are complete, or almost complete, records of insured populations and the services that they received within a healthcare system. Many health data organizations have access to surveys for households, individuals and employers. They can bring all of these data together to get a better sense of how the healthcare system, writ large, is ultimately working for the consumer.
HCI: Do you have a sense of how many states already have something stood up that would meet your definition of a health data organization? Are some still in the very early stages of setting something up?
McAvey: What’s really interesting about this is that these organizations have emerged organically, state to state, and so each one looks a little bit different. I think we probably now have over a dozen states across the country that have established some level of state health data organization capability, whether it's formalized in a separate state entity or designated entity, or whether it's embedded within a Department of Health or an Insurance Department. I think states across the country, regardless of political leaning, are moving in this direction.
HCI: Does it sometimes sit in a nonprofit organization outside of the state government?
McAvey: Yes. In Colorado, the Center for Improving Value in Healthcare. Disclaimer, I am on the board of CIVHC. They are a nonprofit designated by the state to serve this role. I'll also call out Arkansas, where those responsibilities were designated to the university, and then the university has a very close working relationship with the the organization that stewards the APCD.
HCI: Do these organizations have to be careful that they're not stepping on people's toes in the state government? For instance, the white paper mentions that an HDO can track the prevalence of health conditions. To me, that seems like something that a public health agency usually does.
McAvey: It's a really interesting question. It’s a negotiation that every state has to undertake to determine how it can best serve its needs, because you have a certain amount of program expertise that might be in a Department of Health to understand what this data means, and a state health data organization, if it's sitting outside of it and is responsible for collecting it, might not have that same level of expertise. So where you draw that line between collection and analysis makes a great deal of difference.
I think one of the things driving states to think about establishing health data organizations — kind of counterintuitively when establishing a new enterprise — is efficiency. If you think about state healthcare and health-related agencies, they have historically established their own data and analytic teams and have to compete with the private market to keep those teams operational, for infrastructure, for staffing, for leadership, for legal support. There’s a tremendous value proposition in thinking about how a lot of those resources can be centralized and yet concurrently advance a lot of the goals of many of the state health data enterprises to improve health. It can also improve the security and privacy by having a consistent standard against which all of this data is collected and protected.
HCI: Can this get into areas that are quite contentious, like hospital consolidations and mergers? If HDOs do the analysis and weigh in on what the impact of a merger is going to be on consumers, I could see where they could potentially get a lot of pushback, right?
McAvey: I think that if state HDOs provide a reflection of how our healthcare system is performing, there are always those who are going to not necessarily want all aspects of the healthcare system to be reflected upon.
HCI: The white paper gives a great current example involving CHIA in Massachusetts during this whole Steward Healthcare situation, because clearly the state had to deal with a crisis with some of those hospitals closing, and some of them needing to be sold and bolstered. It sounds like CHIA has played a role in helping the state figure out what to do here.
McAvey: Yes. I believe state health data organizations are best served as being neutral arbiters of data and brokers of data, and then being able to connect that data to policymakers to best serve the public. And that’s a role that CHIA has historically played.
HCI: The white paper offers an eight-domain framework for how HDOs might be shaped, ranging from like mission and statutory authority to governance and data stewardship. How did you compile that framework? Was it from looking at existing HDOs and speaking to policy experts on whivh elements seem to be most important?
McAvey: It was informed by lived experience, having worked in an emerging state health data organization, as well as my personal work with states across the country, all seeking to better use their data and information to inform their policy goal policy and program decision making, as well as interviews with a state health data organization leaders working across the spectrum of immature to mature organizations. The framework was also vetted with another dozen or more individuals whose opinions I respect across the country, who interacted with state health data organizations and who would be seeking to establish one in their state.
HCI: If a state is just getting started, should they focus on particular use cases to prove their value initially?
McAvey: I think every state has its own health information needs, and the best thing that a state health data organization can do is to understand the data information needs of its public health agencies, of its legislature, and of the public and look to develop reasonable plans to report on those needs. Those might look very different in different states.
HCI: Are there some states that don't have an APCD set up yet? And if they don't, does that kind of preclude doing all some of this work?
McAvey: Progress is being made every year, and this is across political lines. The last three states that have established APCDs are Texas, Georgia, and Indiana, and other states across the country are actively considering them. I think we're moving our way up to more than half the states with an APCD. I think that there's also a geographic and population component of this, which I am fearful of — where we might have states that have the resources to support these health data enterprises and those that don't. And I think that that requires coordinated action to address.
HCI: So if a state doesn't yet have an all-payer claims database, can it still do some of this work?
McAvey: I think that having an all-payer claims database is a cornerstone component of a state health data organization. But some states might choose to establish a voluntary all payer claims database, or a multi-payer claims database. There are also private resources that can serve similar but more limited ends, and states could invest in relationships to get that type of data to support a narrower set of use cases, if they're not yet ready to invest in a full data enterprise on year one.
HCI: You mentioned the worrisome potential of haves and have-nots in this area. Has the federal government done anything to help states do this work?
McAvey: In the Consolidated Appropriations Act of 2022, the federal government had authorized grants to help states establish state all payer claims databases. But those were never funded, which I think is what stagnated some of the development that we were expecting to see to get us over the 50 percent threshold of states.
