In late October, Healthcare Innovation published a news item about an American Academy of Family Physicians (AAFP) Innovation Lab, study focused on barriers and potential solutions to allow for mainstream adoption of value-based payment models in primary care and how these issues relate to physician burnout. Recently, Steven Waldren, M.D., M.S., chief medical informatics officer at AAFP, and Sara Pastoor, M.D., M.H.A., senior director of primary care advancement at Elation Health, to speak with us in more depth about this research.
For its research efforts, AAFP has been partnering with Elation Health, whose EHR platform serves 30,000 clinicians caring for more than three million Americans, including thousands of small independent practices and large prominent digital health innovators. Elation Health secured $50 million in Series D funding in 2022.
Healthcare Innovation: The study you did with 10 practices found three key themes in terms of success in value-based care payment arrangements: infrastructure, capitation factors and quality measures. For instance, on the infrastructure front, the study uncovered a threshold of financial investment needed to do this work. Did you look at different size practices and what they what they needed to support value-based care work?
Waldren: We weren’t able to look across different sizes of practice, but we find that larger practices often internalize those resources because they can and there's no way smaller practices would be able to internalize those resources, so they hire some third-party service to help them do that — either through their technology vendor or companies like Aledade, Agilon Privia — those types of solutions.
HCI: You found that practices with capitated models experienced less burnout than those in the value-based care models. Was that an observation that was new or surprising, or was that something you've seen in the past?
Waldren: I wasn't surprised to see it. It just seems to make sense that if your payment is prospective, you have more flexibility on how you can care for patients. We did a study that also happened to be with Elation on the direct primary care space. Since they didn't have to have visits to get paid, up to 65 percent of the care they were delivering was asynchronous. So it doesn't surprise me that if you have more capitation, you would see less burden, so to speak.
Pastoor: At this point, prospective payment is a much better way to pay for primary care than the transactional per-visit model. It’s not just that they're getting prospective payment, it's also how much they're being paid prospectively, because there is a threshold below which it's just not enough for the practice to survive. This was a very limited study, but from this testimonial standpoint, we definitely saw that it was really hard for practices to survive if their per-member, per-month payments were too small. Even if they had a large percentage of their revenue from prospective payment, it still matters. So that's why we talked about in the report the quality of the contracts.
HCI: Do you see a lot of practices that are half in fee for service and half in capitated mode and find it a struggle to have one foot in each boat?
Waldren: Yes, that's exactly what's happening. At the recent AAFP conference, one of the value-based sessions was talking about having a foot in both canoes and having to manage both.
HCI: Is one of the trade-offs for getting into the value-based care boat that there's more quality reporting required? Or are some physicians leery of alternative payment models if there's a lack of transparency about the data or not enough trust built into the relationships?
Pastoor: We know that for family doctors, they may have seven to 10 different payers with different quality measures — even if they're about diabetes, they might be different. That just adds a lot of burden. If those are not harmonized, it gets back to the point about the value of the contracts. I think it's also about how much is actually being paid in the bonuses. I think sometimes people ask is the bonus worth all that extra effort?
Waldren: The workflows involved in being successful in fee-for-service payment are very different from the workflows that are involved in being successful in value-based payment arrangements. There are new kinds of work, and there are new competencies, new processes that have to be involved, new data that you need. You don't just flip a light switch. There's a lot of change management that has to happen and the juice has to be worth the squeeze. If the reimbursement that you get for these quality bonuses doesn't pay you to compensate for all of that additional work, then you might decide not to do that. But if you pair those bonuses with prospective payment at a level that is reasonable for the practice, then that might be an opportunity for you to make that leap and make that extra effort. Or if, for example, you give them the opportunity to take advantage of shared savings, that's a little bit more of delayed gratification. You've got to do a year's worth of that work upfront and that transition and adoption of new workflows is a lot of extra investment in the hopes that you're going to get that bonus at the end of the year. But to your point, the transparency is still lacking and so you don't actually know until the end if you're going to get any and how much you're going to get.
HCI: The study found that practices with fewer payer contracts had less burnout. Does this argue for more multi-payer alignment on quality measures? Have we seen some progress on that yet? What are some barriers to more progress there?
Waldren: I would hope that actually happens. What I've heard from my colleagues here at AAFP is that there's a lot of great discussion around let's align on these measures and have a core set of measures, and everybody thinks that that's great. But then they add two or three extra ones on top of that. If you have seven payers that are doing that, it defeats the whole purpose. Also, we can't really measure the things that we really should be measuring, like continuity and comprehensiveness and coordination and access — those things that we know drive down cost and increase quality.
Pastoor: We can add another layer to that which is: are the payers going to communicate to the practice, saying: Of all of our beneficiaries who are attributed to your practice, here are the ones who need care gap closure for mammograms or for colorectal cancer screening or for diabetes. Let's say that you've got five payers and they're all aligned on a core measure set. You’ve still got five different platforms that you need to log into to find out the patients care gaps and understand what the status is and manage that stuff. So there's still an extra layer of complexity that needs to be solved beyond the problem of not having a harmonized set of quality metrics across payers.
HCI: Can you talk a little bit about the work that CMS and CMMI have done on primary care models including the upcoming Making Care Primary. Has there been a gradual evolution and fine-tuning of the models to set the practices up for success or are there still things that they need to do to get those right?
Pastoor: I definitely think that we have seen positive evolution in those models. CMS and CMMI are learning and evolving those models in the right direction. I like that they are offering upfront investment to practices that don't have experience with value-based payment to help them hire additional staff, invest in technology, and develop those new processes and competencies so that they can get over that hump. I also liked that they are beginning to build in social determinants of health in their risk stratification program, because we know that so much of poor health is determined by those socio-economic factors that need work, but there's only so much that a PCP can do, so if we're going to pay primary care physicians to take care of those patients, they're going to require a lot more resources.
I definitely think that we are moving in the right direction with prospective payment, with upfront investment, with, risk stratification, and offering them this opportunity to share in the savings that they create. To Steven’s point, we really have an opportunity to measure primary care in a much better way. My favorite way is called the person-centered primary care measure and it has been fully validated by the National Quality Forum. It has been accepted by CMS into their MIPS pathways, and it could be deployed to every primary care practice today, and we're just not doing it. We're not seeing uptake. Payers are not wanting to do that, because I guess it's just too hard to change maybe.
HCI: Dr. Waldren, I saw you speak at the National Academy of Medicine meeting about the potential for AI solutions to help with easing some of the administrative burdens. Could you talk about some of the promising use cases for AI?
Waldren: In our report, there were several different kinds of administrative burdens that are not just in value-based care, but fee-for-service as well. What we've seen is that leveraging these AI assistants for documentation, and now with the ambient documentation piece that we're seeing, 60-, 70-, 80-percent reductions in the amount of documentation time. One of the key things there is to make sure that it's well integrated in with the EMR so that that flows into the rest of the workflow.
We've seen some chart review type of AI that's able to summarize large records and especially those that are connected to health information exchanges. Even with the best-designed EMR, you still have to go and find the information as opposed to pulling that out specifically for that case.
We're also excited about some of the EHR inbox tools. They're a little bit too early for me to say that they are going to work, but what I've seen has been very impressive and we just had one company at our big annual meeting and the docs loved it. So the question is, does it really work in practice, which is one of these reasons we're doing these types of studies is to talk with practicing docs to make sure that these things do really truly work in practice.
HCI: So the EHR inbox tools route messages to the best person on the team to respond?
Waldren: Yes, they can do that. The feature set that I saw looks at the amount of time that it thinks it's going to require you to disposition the message. So if you’ve only got five minutes, you don't open up a message that is going to take 18 minutes. Or if the message is about renewing a diabetic medication, you’ve got to know the hemoglobin A1C and when was it last done? When was the last time the drug was filled? When was the last time I saw them? Do they have their appointments scheduled in the future? It surfaces all that information.
HCI: Sara, is Elation working on tools like that?
Pastoor: We are looking for any opportunity to reduce administrative burden and enhance clinician efficiency through the use of AI, so we have begun that work already, and we're excited to start piloting some of that stuff soon.
HCI: Are there other things that the AAFP Innovation Lab and Elation are working on now or want to study?
Waldren: When we looked at the literature for peer-reviewed studies, there just wasn't a whole lot out there at all. And if so, it was case studies even smaller than ours. So I would like to continue the review of these types of innovations that we found in the study, and scale that up to larger cohorts. I think making this transition to prospective payment is a critical thing for family medicine and primary care to be successful, not only as practices, but also for our patients.