Emerging Trends in Health Care

Amol Navathe, MD, PhD, is interviewed by Valinda Rutledge, Executive Vice President of Advocacy and Education for America's Physician Groups, on topics including how behavioral economics and AI can enhance care transformation and the lessons learned.



Valinda Rutledge

Hi! My name is Valinda Rutledge, and I am your host today for a very special guest.

As you know, this is the merging trends. This is every month Apg sponsors a what we call emerging trends, and what it is is A conversation. No slides are required.

We don't have any slides, and it's just a conversation between myself and some leading health leaders in this country. And today it's my special guest.

Dr. Amal, Navathal, and I have known Amal for how many years, probably close to 10 years. We work together early in the value-based movement on bundle payments. He is associate Professor of Health Policy in Medicine at the University of Pennsylvania.

He is also, and I'm so impressed with this of all the vice chair and Commissioner of med Pac. He also Co. Founder in embedded healthcare. It's a healthcare technology company that brings a behavioral economic solutions to providers.

He is has numerous articles that he has written in Jama and health affairs on bundle payments, behavioral economics, and artificial intelligence. So it's with great pleasure that I was able to find a time to be able to join us a Mall. So I wanna thank you so much for this opportunity.

I think you're on mute. We're not hearing anything.

Can you hear me now? Thank you so much for having me, so my first question to you is, What is medpac ? Because a lot of people they they know about the reports, but they don't know what exactly medpac does, what it is composed of.

So could you tell us a little bit about medpac and what it means to be a Commissioner at the medpac.

Amol Navathe

Sure, sure, we're very happy, too. So so medpac is a medicare Payment Advisory Commission. It is a nonpartisan agency situated in the Us. Congress

so primarily medpac works directly with the 3 committees of jurisdiction that have jurisdiction over Medicare. So that is Senate Finance House energy and commerce and house ways and means.

And it's an advisory body. So it conducts independent analysis. Again, it's nonpartisan and independent and some of the work shows up in in these reports that are a culmination of work that gets presented at public meetings.

And then we also provide advisory services, if you will, or kind of advice to Cms to Medicare itself through comment letters, and those are all public as well. So mo, you know, given that it's a tax funded entity. Essentially, everything's in the public domain but it plays an important role in providing independent analysis to help support better, more efficient medicare payment policy.

Valinda Rutledge

So how did you get on that pack? Everyone will want to know. They select the Commissioners. Sure. So the Commissioners are selected by the Gao or the General Accountability Office, which itself is an agency of the Us. Congress

Amol Navathe

people can self nominate or they can be nominated by other folks. Sometimes they're nominated by organizations like the American Hospital Association or or American association of family practice. You know, any. Any organization essentially can nominate folks and then the Gao runs a process where they get feedback. They do interviews. And then eventually, Gao makes the decision.

Importantly, it's actually not like Med Pack itself as an agency has control or direct influence over who's selected to the Commission. The Gao has its own formula, and you know we surmise that it has to do with a nice representation of folks across the health sector. And, you know, different political views and political beliefs as well as different areas of expertise.

Valinda Rutledge

Yeah. So so one thing that I want to talk to you about is this year's, you know, the Medicare physician fee schedule implemented in the final rule, implemented a 3.4% reduction in physician fees schedule

Congress last week slightly mitigated it. And but it was very inadequate. Because if you recognize with the inflationary costs, such as supply and labor costs it just having any kind of cut they were on able to totally eliminate the cut. So I think the cut ends up being a 1.7 2 for the year.

But they were unable to totally mitigate it, and with the increase in inflationary cost. Physicians really are going backwards in terms of reimbursement. You have wrote about the primary care system being in crisis. What do you think are some of the variables within the current system that is contributing to this primary care crisis, that we're getting to a point where there's some parts of the country that literally have primary care deserts.

Amol Navathe

Yeah, yeah, it's a it's a great multi faceted question, Valinda, that I think of great importance to American healthcare. I'll just briefly, note that the comments I'm about to provide are in the context of my own personal capacity. So not in the capacity of of medpac and so maybe I can highlight a couple of things. So so first off, II think you're right that there's been. It's very clear that there's been strong inflationary pressure right on practices. And

and that has multiple potentially negative effects. Right? So first off, obviously, it makes it tougher on physicians and physician groups to maintain their financial situation.

So that itself obviously is not great that can hurt access that can hurt sustainability that can lead to early retirements which can meaningfully affect our supply of physicians. In the first place, which is one of the points that you're making.

Secondly, it creates this challenge because of our existing fee for service system and Medicare specifically, that pays differently for office based position visits than say, hospital based clinic office visits, and there's some kind of legacy rules here. But the point being that as we have these kinds of pressures placed, it increases the impetus for vertical integration or acquisition of practices. By whether it's hospitals or private equity or other entities, health insurance companies, and by and large the evidence on consolidation has not seen that, you know, has not demonstrated that patients benefit seems like prices go up, and quality doesn't necessarily get better.

So that's a big challenge, and then, now let's bring it to the the really poor point of what your question was is, let's talk about primary care. So you know, over the past decade. We have seen a pretty dramatic decline in the number of primary care physicians who are participating in Medicare building the fee schedule and some shape or form. Some of that certainly has been offset by, I think, a good trend in that nurse. Practitioners and advanced practitioners have started to fill some of that gap.

But fundamentally that doesn't alter this point. That primary care is in in this state of crisis and the fee for service system does put primary care very much in the crosshairs for a number of reasons. I mean, let's take take this in a couple of steps. So one primary care physicians as do a lot of physicians. They do a lot of stuff that doesn't show up, you know, stuff that doesn't show up on the fee schedule. So they're essentially not being paid or reimbursed for a bunch of their activity and when we do have a lot of codes with Cms has added codes for transitional care, management, and care, coordination. And now, telehealth, you know, it's kind of like a death by a million cuts. 

Sure there's these codes that we can bill for. But oftentimes the activity of billing itself is more costly than the payment or the reimbursement that you get from the activity that stimulated the billing. In the first place, right? So II think we need to be reexamining as a nation what is the right way to pay for primary care.

Given that it's such a fundamental core part of producing health for a population and for a well functioning cost efficient, patient centered. You know, healthcare delivery in the long run.

Valinda Rutledge

Yeah. So, as you said, primary care is the cornerstone of our health care delivery system, and you've made some excellent points in terms of how we need to change it.

You recently testified at the Senate budget hearing last week, and if any of you have not read his testimony. It is phenomenal. It's under the Senate budget Committee on Hearings. Senator Whitehouse introduced a new proposal in terms of taking a look at a new way to pay primary care. This new way is a hybrid payment for primary care. Could you explain a little bit about what that is? What would it entail? And has this concept been tested?

Amol Navathe

Sure? So hybrid payments are basically a hybrid of 2 different components, right? One is a fee for service like payment. That may not be the full fee for service payment that we're currently paying under. Say, the fee schedule that's one that's component number one, component number 2 is a fixed per member per month predictable payment that goes or per beneficiary per month payment that goes to the physician or the physician group and and that, for example, can be risk adjusted for the severity of illness, for a beneficiary as well.

And the idea here is to shift the payment system to one that creates predictable revenue that recognizes all the activities that either are too costly to Bill, or don't even show up on the fee schedule.

Stop trying to get there where we have every every activity you know, showing up with the code, but rather shift to a system where physicians are, are rewarded and appropriately compensated when they do in person visits but are also. and kind of imbued with the flexibility, right? The flexibility to say, hey? Sometimes I'm going to do telehealth when it makes sense. Sometimes I'm going to do audio only business. Sometimes I need to see the patient. Sometimes they don't need to see me. It's great if I can have them see the nurse, practitioner, or the pharmacist to really work on their chronic disease management. So the idea behind the hybrid system is to really impart flexibility. To practice is to meet patients where they are and staff in flexible ways, and use technology like telehealth, you know, when it's particularly efficient and when it's particularly effective.

Valinda Rutledge

Well, and I think it would be exciting. And I think, with Senator Whitehouse behind it.

You know, and what I it was my understanding. What he's asking for is teammi to begin to test it, and hopefully we'll be able to see that come to fruition, but I think it would be excellent for us to be able to test the hybrid model.

Any comments about where you think this proposal was gonna go? Cause I thought it was very exciting.

Amol Navathe

Sure, yeah. So you know, hope hopefully that the proposal will move forward. I think you know one of the questions you previously asked is is there are there examples of this being tested, or in the real world or otherwise? And you know there, there are a couple of examples that are worth noting. So I've been fortunate to be a part of a program with the Blue cross blue shield of Hawaii.

Called Hawaii Medical Services Association, where in 2,016, we actually launched a pilot program with about 100 Pcps paying in this hybrid payment model context because of outsize quality gains better adoption of technologies like telehealth.

We actually ended up scaling that program subsequently to the entire State and you know, notably Valinda. When the pandemic hit in March, April of 2020, you know many physician groups, particularly primary care. Practices they were, they were massively disrupted in terms of their financials. Right? Cause. All of a sudden nobody wanted to go anywhere. And there were lockdowns. But in in Hawaii that actually didn't happen, because we had this hybrid payment model that help the physicians keep the lights on. Secondly, they had already shifted over to remote technologies where it made sense. So they're already scaled up in terms of being able to do telehealth visits and remote, patient monitoring. The other technologies that were important there. So we learned a lot from that. You know that initial program. And it's continues today. And it's continuing to be refined and improved

there. Notably, there are also examples in Medicare. So within the comprehensive primary care plus demonstration, there actually is a component of that that is paid on hybrid payment to the primary care practices and and I think the most really interesting thing there is, you know, not only as it feasible and implemented and does it work. But when you have practices that are receiving these hybrid payments through Cpc plus that are simultaneously participating in an accountable care organization model, say, through the Medicare share savings program you actually get outside synergies. You get more savings, better health. And so I think there's real promise, not only for hybrid payments in primary care alone, but also the intersection with that, with the whole shift to value that we're seeing nationally, and that Medicare is leading

Valinda Rutledge

right, and if I remember correctly, at the Medicare position fee schedule, the proposed rule. Cm. Cms. Had asked for Rfi. Some response from stakeholders regarding the use of capitated payments in M. Ssp, particularly the enhanced.

So we may actually see this being tested very rapidly. So let's keep our fingers crossed. I wanna move to artificial intelligence. This is something that we continue to hear over and over again, and particularly as it moves into healthcare. Several years ago you co-led a study at the Va. Using artificial intelligence to identify veterans that were at high risk of hospitalization or death.

What was the outcome of the study. And then what advice would you give providers, as we continue to start moving into artificial intelligence at a broader level in the clinical practice? I'm hearing more and more practices with their Emrs are gonna start moving into artificial intelligence. So sort of what was your experience with that? And what advice would you give people at the clinical practice level?

Amol Navathe

Sure. So so first off, you know. Thanks for for the question for highlighting this study. So what we did. There is, you know, we wanted to really understand.

What are the kind of key characteristics, key subgroups of veteran veteran patients who are, as you pointed out high risk for hospitalization or mortality. So they're a high need kind of high cost type of population.

And what we found is you, you know, not surprisingly, we found that there's a lot of clinical conditions that really kind of define a sub group. So you know, patients with end stage liver disease and cirrhosis, or patients with end stage congestive heart failure, which requires a lot of readmissions and advanced therapies. So those types of conditions were kind of no surprise as their as a way to identify subgroups.

What was more surprising is that there were clear subgroups that were defined, not by their clinical characteristics per se, but more by their social characteristics, so either racial and ethnic groups that they were part of, or an intersection between, the racial, ethnic group and other ways in which they utilized care.

And and that was not something that we predicted upfront. But I think it highlighted to us that, as we think about better managing high risk populations. It's not just about the clinical piece. In some sense we were kind of ahead. This has been a few years ago. We're kind of ahead of this notion that we need to start addressing social drivers and thinking about those elements and bringing that together in an integrated model to really effectively manage these populations. So

I think this was an example. Where, AI didn't tell us, hey? This is the answer of how to fix the population, but it really helped us reveal insights that we wouldn't have had otherwise.

And and so the second part of the question you asked Valinda, is, you know what is the advice for clinicians and practices out there, you know. II would say, it's important for us to recognize both the power and the pitfalls of AI,

and and to be careful and cautious, and how we adopted. And and here's what I mean. So I think there are some really phenomenal uses of AI. You know, one of the ones that that we just talked about, I think, is a good example where it can reveal insights. It can help us recognize patterns, and it can bring to light

things that would otherwise be very hard to discern whether it's from data or just something that's kind of happening outside our 4 walls of our clinic. Or what have you?

I think in those cases is very powerful, and in some sense, because it's not telling us right or wrong, or do this or don't do that, I think we can be a little bit more reassured that we can make decisions using that information that can be an adjunct to the way that clinicians think and do and practice and care for patients.

I think the areas that we have to be more thoughtful and and really more careful and cautious about our when AI is actually telling us to do something right. And this is where there's been a lot of concern about bias and AI, and how it might lead us astray. And so I do think we, as clinicians, with the obligation of doing no harm to our patients. We do that. It is the responsibility is really in part on us to be careful about.

You know, over over interpreting or relying too much on these kinds of technologies which can sometimes make it just seem easy. They can, you know, cut through some of the cognitive work and make it seem like, Hey, I can just do this, and I'll be done. But but we have a responsibility here. So I think, what's my, my last point here, Valinda, is, we should just be very careful about how we're using the AI AI. I would not say.

AI is good or bad, I think if we segment it in terms of what we're using it for, however, it can be a really helpful guide

Valinda Rutledge

in terms of whether we should or shouldn't use or trust AI in the short term. Right? So can you give us some examples that you see, from organizations you're working with, or some research projects in which you think the use of AI is is truly has an opportunity to transform care

cause. I think people are very excited about okay, the potential. So what? How can we operationalize it?

Amol Navathe

Sure. So I think there are several places. So you know. For example, AI does a very good job of going through. And what we would say and Techno speak is ingest, a lot of data, right? So it can basically read a lot of information whether that's text information and clinical note or Ehr information or information information in insurance claims or something else. Right? It does a very good job of reading that and then surfacing things that are likely important, based on historical at from historical significance, meaning what's happened in the past.

and so it can be a great adjunct to revealing barriers to things like adherence to medications or vaccination. Other prevention behaviors. So I think these are the types of applications I'm really excited about where it reveals something to a clinician or a practice that would take hundreds and hundreds of hours for er to for us to pour through a chart to really understand. But the AI can do it in the background, and then boom! Pop it up and say, Hey, look! It's an insight, right? So so one of the examples could be. Now tell me if I'm all based. But

Valinda Rutledge

you if you could. If AI could have access to how much that they have filled in the anti hypertensive drug and you also have some of the readings from not only the primary care, but maybe if they went into specialists 140, you know, because a lot of times your blood pressure is taken. If it's elevated a little bit, they sort of say, Oh, it's because you're in the doctor's office, you know, so they sort of can take a look at that, as well as how much you filled to be able to sort of say, this person is not filling it on a regular basis. So there may be some issues in terms of being able to afford it is that something that you can see can happen for the future. So you don't have to pour through all kinds of different data.

Amol Navathe

That's absolutely right, you know, especially in some sense, the more disparate that data is, the the further that data is from a single source, and the more varied it is, I think, the more and more strength you have of of using AI to go through and create the type of insights, Melinda, that you're pointing out. I think that's exactly right.

Valinda Rutledge

So then I want to turn to behavioral economics. So we have used behavioral economics in the development of the insurance design for patients. In terms of center incentivizing them to have healthy behaviors eliminating copays on certain things. So we've seen that work.

You've done a lot of work in terms of developing incentives for physicians to develop, to deliver a value based care. And that was part of your new company embedded, which is not so new anymore. Tell us a little bit about that, because that is the new field. We certainly have seen the success in using behavioral economics for patients.

And that has been behind some of the things that insurance companies or employers have done in terms of making employees make healthy decisions and get preventive and etc., by waiting some of the co-pays. What are some things with your new with this company? And what can we do in terms of supporting physicians in terms of incentives for them to be able to support value based care.

Amol Navathe

Sure. So so the spirit of behavioral economics, this is gonna sound glib. But I promise I don't mean it this way is to recognize that physicians are humans, just like patients are humans. And so we're not robots. We're subject to a lot of the same kind of cognitive decision errors that other humans are and so how can we best support clinicians and physician groups in addressing some of those kind of repeatable patterns that we tend to do things that we could make a better decision if we had the right support. So that's the kind of basic construct of why we wanna try to apply behavioral economics to physician decision making in the context of patients has been used, for example, to triple smoking cessation rates. Right. So there's ways to use either financial incentives or non-financial incentives like showing people data on their peers or getting them to commit in advance before they undertake a program and identify a peer who's gonna be? They're gonna be accountable to. So some of these are are really just decision making strategies that are on, how we behave that can deeply influence, how the decisions that we end up making.

So how can we use this in the context of physicians? Well, you know, one of the things as a mission based company that we were very interested in doing is seeing how can we help physicians accomplish their goal of taking great care of their patients, but doing so in a more cost, efficient manner that saves their patients money most importantly on out-of-pocket expenditures. That also saves the whole system money. And how can they win in the process?

And and that's basically the spirit of what embedded healthcare is trying to do. And so what we found is that by bringing new types of data, helping physicians set goals in very

and very deliberate, intentional, targeted ways around how they're gonna pursue more cost, efficient practices, whether it's referring to higher quality specialists, or, you know, doing colonoscopies or knee replacements in the appropriate setting of care.

You know, site of service selection or some other clinical decision that they're making by bringing data to the front that they don't see otherwise by helping them set these targets, they can be much more effective in accomplishing their own goals. Which is really fundamentally important piece. So this is what are your values. And how do we help you achieve your values where patients benefit in the process?

Valinda Rutledge

Right? So I've got one last question for you.

and that is, you know, a lot of people are feeling that the value based movement is slowing down that. We're struggling in terms of coming up with different ways to do risk adjustment, particularly incorporating the social determinants of health to incorporate some of the savings that we've had before, and not keep having the benchmarks. Do you know, I mean incorporate the savings. So you know, your threshold just keeps higher and higher, or you know you're so efficient, and the benchmark is set on historical. And so it just becomes really difficult for you to participate. There's lots of different reasons out there, but we're just seeing some people sort of. Say, this is exhausting.

And are we really making a difference? So tell me your thoughts on that, because I think we're getting to an inflection point in this country, people are saying, if we're gonna continue down this journey show show show. Show me the results. Have we made a difference in terms of the Medicare beneficiaries and your thoughts about that. How can we reenergize the stakeholders?

Amol Navathe

Sure, yeah, II think you know great points. II definitely agree with you. We feel like we're getting to an inflection point. I mean, we are, you know, 1314, 15 years now past the Portable Care Act and the start of Cmi and many of the programs. So I think you're right to call this out.

You know, I'm very much an optimist. and I would say, you know, change in healthcare, especially in a you know, multi trillion dollar system. That's 20% of our economy. This is, you know, this is true. The metaphor of trying to turn the Titanic here is is very apt, right? It's really hard, I think we, as a community, should be really proud of the results that we've gotten.

I think scoring this, based on you know how many dollars we have saved in the short run is not always the right way to look at it. Rather, I think, if we look at how practice patterns have changed

under a variety of different models, whether it's aco type programs like Nextgen or Mssp. Whether it's under bundle payment models or advanced primary care models. There's a number of different examples where we have shown time and time again, that if you change the incentive structure and get physician groups and hospital groups and other organizations that that want to participate want to make those changes we can really materially change practice patterns.

And that's really hard work.

And I think to some extent we're stuck in this very challenging game around the being counting right? The accounting of this weather savings are coming or where they're coming from. For a couple of reasons. Right? One is, we have created this pretty fundamental shift in mentality

for provider groups and hospital systems and other delivery organizations. And that's spilling over all over the place. I think we've seen this material shift, you know, not to assign too much to a trend here. But Medicare spending growth is actually flattened in the past several years. Take the pandemic piece out of it for a second right that that hadn't happened for decades.

I think at least part of that is related to the value-based care movement.

And when we do micro evaluations of this model and that model, this model and that model, it's actually very hard to tease out. Well, you know the group that you're comparing to guess what they're probably participating in some other model, or at least they're affected by some other model. And so we tend to understate, I think, what the true effect at a very large system perspective we are getting. And so I'm very bullish on value based care. I'll also in value based payment. I'll also note that it's also notable to me, you know, in the past decade, plus that it's it's not like, there's some other strategy that is particularly physician centered or delivery system centered, that has emerged as Hey, you know, what if we don't do value based payments and value based care. Let's just do that.

And I think I think there's a reason for that. I think that we collectively as a nation, we're onto something. We're a leader in the world. The world actually looks to us for the delivery system reforms that we have started. And so so I'm you know, Valinda, I'm still very optimistic. That doesn't mean that it's not hard and change not gonna be hard doesn't mean that we don't have to roll up our sleeves and and keep trudging along. I think you're absolutely right that we have to be very thoughtful about the design, so we're not penalizing the efficient.

and that we're being very clear about setting the right target. So as you did derive more efficiencies, you continue to win. Not that we, you know. Take the wind out of your sales. I think we those are very real practical pieces, but I think from a philosophical perspective.

I think you know absolutely we should double down on value based payment, because I think we have a lot more to be proud of than sometimes shows up in a Cbo report, or shows up in a single model evaluation, or or whatever that is

Valinda Rutledge


Well, I would like to thank you so much. This has been a great conversation, and I know how busy you are, and I'm so thrilled that we were able to pull you away from this very important research. You have, as well as your role in Medpac. So I really want to thank you very much

for everyone else. Our next guest, which will be in April, Norma will be sending out. The invites will have Dr. Patrick Conway, and I think all you know him and Amal. I'm so excited that you joined us, and you're certainly more than welcome to come back anytime. And I look forward to having you here again, and I look forward to next month's conversation with Dr. Patrick Conway. Thank you again.

Amol Navathe

Thank you.