Beyond Finger Pointing: The Fight to Lower Health Care Costs

Dan Gorenstein (DG): The debate over how to bring down health care costs often devolves into an exercise in finger pointing.

Hospitals, insurers and drug companies all say the other ones are to blame.

On June 25, I moderated a panel that tried to lean into compromise.

We brought together a hospital CEO, an insurance executive and a pharmacy leader, along with a top federal health official and just for good measure, a leading health economist.

The goal: how these major players could work together to make care more affordable.

The conversation took place at Aspen Ideas: Health, a conference that brings together policymakers, clinicians, researchers and industry leaders to discuss some of the biggest issues in health care.

Today, we’re bringing you that ambitious, solutions-focused discussion.

From the studio at the Leonard Davis Institute at the University of Pennsylvania, I’m Dan Gorenstein. This is Tradeoffs.

*****

DG: The conversation you’re about to hear was recorded live and has been edited lightly for length, clarity and sound quality.

Good morning everybody. Thank you so much for coming. Making it out on the last morning of the conference. It’s raining and you’re here. You’re excited. I know why you’re here. We’ve got an incredible panel. This is very, very exciting. And we are, I think, very true to the spirit of this conference, this session more than many, it seems to me at least I’m biased, of course, but this session more than many is really a very, very big idea.

The issue that we’re talking about today is that one that people on this panel and across the country have been working on for decades. How do you make health care in the US less expensive without compromising access or quality? How do you make health care less expensive without those compromises? People really, we all want that. Everybody in this room wants that. Everybody on this panel wants that. But achieving it has been so elusive. And today we’re going to try to begin to tackle why. What are some of the problems and what are some potential solutions?

So with that said, I’m going to introduce myself. My name is Dan Gorenstein. I am the founder and executive editor of the nonprofit health policy news organization Tradeoffs. We have a weekly podcast called Tradeoffs. Check it out if you’re interested in health policy podcasts, it’s great. More than 70 colleges and universities around the country use our content in their classrooms. And I’m going to go down. Stephanie, if you’d introduce yourself, the quick version.

Stephanie Carlton (SC): Steph Carlton, deputy administrator at the centers for Medicare and Medicaid, also the chief clinical AI officer.

John Love (JL): John Love, I have the privilege of leading Amazon Pharmacy. And if you’re not familiar, I kind of describe it as pharmacy 3.0. So instead of retail or mail order, our pharmacy is in your pocket. It’s a digital forward, experience. We use technology to improve the quality and cost, and then we deliver medications to patients’ door.

Brendan Carr (BC): I’m Brendan Carr, I’m the CEO of the Mount Sinai Health System. Mount Sinai is a large academic health system based in the Greater New York metro region, medical school, graduate school, nursing school and a large series of ambulatory clinics and hospitals.

Katerina Guerraz (KG): I’m Katerina Guerraz. I’m the chief operating officer of Aetna, a health insurance company and a subsidiary of CVS Health. I also am the president of our Medicaid business.

David Cutler (DC): I’m David Cutler, I’m a professor of economics at Harvard. And I occasionally do health policy work in Massachusetts and a little bit nationally as well.

DG: Let’s give it up for this panel real quick and get the energy going in the room. So, as I said, the goal of this panel is to move beyond finger pointing. Today is about a shared responsibility among hospitals, insurers, the pharmaceutical industry, policymakers, to make this system work better for us.

But before we get to the solutions, I want to start with why this problem has been so intractable. David, you’re really the big picture thinker here on this panel. You’ve spent decades looking at all the reasons why health care in the U.S. costs so much.

Why is it so hard to make health care in the U.S. cheaper without compromising care access quality?

DC: Yes. So a couple of points. One is technologically it’s feasible to do. We know you can make health care cheaper and better. So it’s not a question of impossibility.

It’s complicated in part because it’s not just a single thing. So it’s not just reduce this price or reduce this thing. There are a bunch of different facets. So let me give you a few of the facets that are going to show up. And I’ll sort of give you kind of four illustrations.

First one prices are too high. We pay more for stuff than we need to. Example, a drug that sells for $100 a month in the UK sells for $1,000 a month in the U.S. No reason why we have to pay ten times more.

Second is administrative costs are outrageous. In a $5 trillion health care system, about a quarter of the dollar. So over $1 trillion goes to administrative costs of health care. That’s far higher than any other country. It’s far higher than any other industry. It’s far higher than anyone who looks at it thinks.

Third is we do stuff that isn’t always in the patients, in what the patient needs. So there’s a lot of patients who are in hospitals who could be cared for at home, who are receiving advanced imaging when they don’t need it.

Some of its sort of uncertain care, but we have a bad way of directing people. We pay for doing more stuff. So there are a lot of doctors who make money by doing more stuff. The U.S. is among the most medicated imaged populations on the planet, second only to countries that do it worse than us. But we do very poorly at allocating scarce, expensive resources to people who need it.

And then fourth example is we do terribly on preventive care. So we’ve known how to care for cardiovascular disease, risk factors, high blood pressure, hypertension, diabetes. We’ve known 30 or 40 years. Most of the medications for most of those are old. They’re cheap. They’re generic. They cost $0.20 a day. No more than one third of people with high blood pressure have their blood pressure controlled. Same is true for people with high cholesterol. We do a terrible job preventing stuff, and that means that the burden of things like cardiovascular disease, chronic respiratory disease, all sorts of other diseases are way higher than they need to be. So lots of different aspects here.

DG: Thank you, David. And again, I took notes so you all don’t have to, four reasons: prices too high, administrative costs, we do stuff that’s not always in the patient’s best interest, poor preventive care. These are structural challenges. And we’ve got four other panelists, five panelists, can you believe that?Gour other panelists here who can really talk to some of these structural challenges?

And so, David, thank you for the context to understand why this is so hard. And now with the rest of you, let’s get concrete a little bit here. What’s one specific example of what makes this goal, lower cost, good outcomes, hard to pull off? I’m going to give each of you 90 seconds and I’ve got a clock here, people. So I’m watching. 90 seconds to walk through your own example and I really want to save most of our time here so we can be talking about the solutions, which is, I think, why a lot of the people in the room are here. So, Stephanie.

SC: Yeah, I’d start with we have a system that makes it really hard to be healthy in the United States. You think about it, we spend twice as much per person on health care. We have worse chronic disease outcomes. Something’s wrong. When you think about our kids get two thirds of their calories from ultra processed foods. They spend more time on social media than they do playing outside and getting good physical activity. No wonder we have a childhood obesity problem. We’ve got to get underneath that.

But the system is sort of stacked against us because we think it’s not going to matter. And there’s a there’s a system that can help me when I do get sick. And you don’t build those daily habits that make a big difference on health.

The second thing I point out is that we have a system that is full of entrenched special interests, and for every dollar of inefficiency in our health care system, there is a constituency. And change is really hard, especially when you spend some time in Washington. I was acting administrator at the beginning of the Trump administration, and I remember we had a big choice about what we were going to do with the prescription drug negotiation program that the Biden administration passed. A lot of the pharma lobbyists said, why would you implement it? Just don’t. We had to. It was the law. It was also the president’s priority to bring down prescription drug costs.

So we did it, took some slack for it, and then we doubled down with it for the MFN deals. That wasn’t easy because the politics of health care, that political economy, is hard to overcome. And that really does have an outsized impact on affordability and cost. The last thing I’d leave you with Is that we have a private system, which I’m a capitalist of captain of the capitalist army, private system in America that I believe in. But the incentives are built around profiting off of friction, not profiting off of value and delivering better health outcomes for Americans. I think that was 90 seconds.

DG: I think well, I think it was a little over, but that’s okay. We’re going to with the government. Okay. John.

JL: Yeah. So I agree with the premise. I don’t think there’s any debate. We have a system that is structurally inflationary. But I am an optimist. And I, you know, coming to Aspen Ideas and seeing what we’re working on and the great quality of care at places like Mount Sinai, biometrics that are available, I think even here, you know, saw XPRIZE talking about a standard of, you know, encouraging invention for better health outcomes in the VC panel, Aura and Thrive Capital talked about, you know, creating a program for parents to give care to children with special needs.

So I’m super optimistic we have a lot of solutions, but health care is the one space where the patient isn’t at the center. Everyone else is sitting here today. You chose what you wore. You chose what you eat, where you go.

In health care. It’s a maze and you run through, you know, regulations, triage, trying to get an appointment, seeing a doctor, maybe labs, maybe a specialist, into a pharmacy. And not only is it challenging for the patient to access that care and navigate it and get the right care, there’s an enormous amount of waste and sludge as we all knock into each other because there’s no transparency. So you’ve got prior auth and rejections and things that are happening in the system that are very complicated, and it’s actually more expensive to do it poorly than to do it right.

DG: Well, and we’ve talked about this in the prep calls. Can you just drill down on this like this example of like the pharmacy problem?

JL: Yeah. The pharmacy problem is, you know, a lot of patients don’t even get to see the price of their medication. And so, you know, they will be nonadherent because they don’t know if they can even afford it. They don’t want to go in and have a, you know, a physical conversation about a private health matter. And if I look at it, the cost for Amazon Pharmacy to do a clinical evaluation, make sure that drug is safe, packed correctly and delivered to the customer is a lot less than when there’s a rejection or the doctor prescribes and we’re unclear if that medication is replacing another, there’s a prior auth, we’re emailing each other back. There’s faxes and phone calls. The patient’s calling us. Where’s my meds? They’re non-adherent. Being non-adherent in pharmacy, 27% of medications getting picked up, and it’s $300 billion of cost to the U.S. health system.

DG: Thanks, John. Brendan, what makes this so hard to do?

BC: Yeah. I mean, you just heard one of the most famous health economists on the planet say it’s hard because we’re talking about people, right? People are really, really complicated, it turns out. And so I would say that as its first premise, that’s why it’s so hard.

The second is because we’ve not really defined it. Most, much of what we talk about when that drives people’s health. You heard the food supply critically important, right? You heard about the way that the decisions that you make and the way that you live your life critically important, right? All of these things are driving our health care.

And I’ll give you a dynamic, the way that I think about it for the way we set it up. So we’ve not defined what’s in that bucket. Poverty, education, nutrition, right. All results in health outcomes. Is fixing them health care or is that belong in a different bucket? We have to ask ourselves that. And I’ll just leave you with this. I think about the fact that everybody agrees we need to decrease costs, whether it’s the government or large employers. They get an agent to help them to decrease costs, right? Largely payers, or they run it themselves as the government. On the other side, the employer, which I’m a doc, right, I think of as a patient, they get me as their agent wearing my doctor hat, right?

The economists say this is low value. You don’t need it. It won’t benefit your health outcomes. Right. To save money for the employer. And then the patient comes to me and says, I super want it because I’m stressed out that I might have this disease. And I know that the right way to get it, to make sure I don’t is to get this test or this CAT scan. And I say, no, no, there’s a whole logic flow. You don’t need it. And also it costs money. It’s low value to somebody else, even though it’s high value to you and your anxiety around what you may or may not have. That’s the setup, right? And that is why it is so hard.

DG: So just to just to put a pin on this. You’re talking about the tension between what the guidelines, what what professional guidelines would suggest is sort of high value optimal care versus what that patient actually wants because the patient does not feel good, the patient is stressed. And you know what, doc? Just give me the scan. And that’s the tension because the guidelines say, don’t do the scan. But what what do you do?

BC: Uncertainty is uncomfortable. That’s what they feel. They come to us trying to solve for it. And we have tools to increase the certainty. They all cost money.

KG: All right, I’ll hit on, I want to go back to David because you are the most famous economist on the planet.

DC: You don’t understand that. That can’t be a good statement. There’s just no way that…

KG: So the friction, the cost in the system is real. Even when we work with a lot of our health systems, we know anywhere from 6 to 10% of their costs are on focusing on things like, hey, whether it’s the payer, whether it’s a different intermediary, trying to figure out why did you not pay my claims accurately? Why did you not pay it quickly?

I’m very confident that’s going to get resolved because it’s costs on our side, it’s costs on the health system side. And I do see some alignment.

The piece that I don’t think we have sustainable solutions yet is when you look at the next three to five years. The three biggest high cost trend drivers of cost specialty drugs. If you think about complex health conditions like oncology or autoimmune. And the third is behavioral health. So even if you look at behavioral health, the biggest category between that is autism.

Today in this country, it’s $460 billion of spend on autism care. By 2030, it’s going to grow to $600 billion. That’s that’s an area where we do not have standards around clinical care. You have a lot of fraud, waste and abuse in that space. So until we figure out all of us actually on stage, how do we combat those three biggest cost drivers? I mean, that’s really that’s the big ticket in terms of dollars in this country.

DG: Thank you all very much for these examples. And, you know, it goes without saying, but what you’ve laid out are these endemic structural problems in our health care system. There are obviously no easy answers. Most solutions, as frustrating as it can be, or modest their iterative, more pruning a branch than cutting down a whole tree.

And I liked to go down the row and ask each of you, 90 seconds again, to describe one concrete, achievable solution that you could implement as a hospital, as an insurer, as a pharmacy, as the federal government to make progress towards this big aspirational goal. That’s, let’s start with Katerina. So we’ll kind of go in reverse order here.

KG: So one example of something we’ve put in place but has not been scaled is so we talk about just taking the friction and the excessive costs out of the administrative costs out of the system. The big piece or the other piece is clinical collaborations between different groups. So one of the biggest problems in this country is we do not discharge patients on time out of the system for different reasons.

So one of the things that we put in place about 18 months ago was we put Aetna nurses on site in 24 facilities across the country working with the care team, but most importantly educating the members on their benefits, what they have access to, having a conversation to say if you’re going home, do you have a way to get your prescriptions? Do you have transportation? Can we schedule your post, you know, hospital visit? Similarly, if someone has to go to a skilled nursing facility making sure the member understands what to expect.

I would tell you just it’s early stages and that’s something that we’ve already seen because the goal is you’re trying to get people not to get readmitted to the hospital, because that means their health went down once they left the hospital. And our job is to help them stay healthy. And we’ve already seen upwards of 15% of reduction in hospital readmissions. It lowers the total cost of care, and it also connects people to more appropriate sites of care.

So that’s one example. I know Brian and I were talking about what are other examples. The sky’s the limit. We should do this with, you know, kids who are in NICU, babies. There’s a lot of different applications, but that requires a different way for all of us to work together.

DG: And can you explain that for the crowd here? What would Aetna and Sinai, as an example, have to do tomorrow that you’re not doing today to achieve this?

KG: Well, I would actually say so again, we’re it’s working with, this is a slow rollout. We want to make sure it works. It would just mean we have an agreement where we’re putting nurses on site who are working with the Sinai care teams to say, okay, how is this patient doing? When do we think they might have to get discharged? And then on the back end, the administrative side of it, we’re clearing, getting rid of the prior auth, making sure there’s a bed in a facility.

There are things like, you know, sometimes patients break their hip because they got up to get water in the middle of the night, and they couldn’t turn lights on because they couldn’t afford to pay their bills. So it’s also social determinants of health. But that’s something where Sinai, I know, because they’re one of our partners, does a lot in the local communities as well. And I think it’s that’s where it’s just getting the two teams to work differently.

DG: And part of what I’m hearing you say when you talk about sort of your teams in the health system, right? I mean, so Brendan, looking at you here, like this is giving up some sort of control. This is really, as you talked about, Katerina collaboration. Can you respond to what Katerina said as this solution, and what would it take for Sinai to get there to, to get to yes?

BC: Yeah. I mean, look, it’s my line is it’s a, it’s a perfect extension of the conversation that we were having before is that a lot of the drivers of, you know, length of stay is a great example. Why are people, why does it take longer to get discharged from the hospital than you might think it should take to get discharged from the hospital? Because, you know, the doctors can say you’re done medically, right? The payers can say, you know, it looks like everything’s done that needs to be done. And now we just have the small matter of the messiness of life, right?

Like that’s what gets in the way. You know, someone can’t go home, they’re not safe. They’re, they don’t have food. They don’t, they can’t get their meds picked up. It’s all the incredible complexities of the fact that we are talking about people. And that’s really, really hard. So, you know, we have forever said in health care that the, you know, there are the incentives are not lined up.

We’ve also been really reluctant to say, I would say, to acknowledge that we have a two tiered system, right. More than two tiers, but we definitively have one, right. Health care is divided by your income and your resources, whether that be family support or, you know, or financials. If we really want to line up the incentives, we have to do a terrible thing that none of us want to do.

We have to say, your hospital stay is now over. We start charging you like a hotel. Who does that hurt? That hurts the people who are the most likely to have not been able to go home on time, if there isn’t on time in the first place. Because life. Right. So, you know, this is this is where we’re going.

And when I, when I come to Aspen and I enjoy every minute that I’m here, but I watch us trying to solve really different problems. We’re trying to solve the foundational things that drive the underlying health outcomes of the nation because people that choose to come here are deeply committed to that. Right. We have a massive mission driven culture at Sinai. We also want to invent the future. And there’s amazing things that don’t exist yet that we talk about here that in, you know, three or five or maybe never, right? Some of them will work. Some of them won’t work. We love to invent the future. We have a massive research portfolio. We think of it as our job as an academic health center to cure all the things. But, you know, but those are widely disparate challenges, and we need to be clear about them.

DG: And I just want to press you on this for a second, though, in terms of this idea of this kind of collaboration, since this is a nice concrete example from Katerina. What changes would you have to make at Sinai to like, get that to become reality? How hard would it be to do that?

BC: It would not be hard to let folks that have that are not seeing the messiness of the real world. I apologize, I don’t mean that as a punch, but to say that they’re doing an administrative process. It gets when you get in front of the patient and they say, I cannot go home. There’s no electricity. I don’t have any food right then. Then we have a partner in solving that problem.

KG: Yeah, I would just say, and I don’t think it’s anything we’re not taking anything away from the health system. I don’t and I think you would agree with this. I don’t think it’s the doctors or the nurses job to educate the member on their benefits and what services.

And I think that’s John’s point. We’re, our job is to get rid of all the homework and actually help the navigation. And that’s the piece that this role plays that doesn’t exist consistently in every health system in this country. Again, don’t expect members to be experts in their benefits, but it’s someone who’s helping them and talking to their family to say, and by the way, sometimes it’s other things. Sometimes you can’t get a skilled nursing facility to give you a bed because candidly, if it’s a Medicaid patient, they’d rather have someone in a commercial who’s a commercial patient because they have a higher profit there. So those are the kinds of barriers, the messiness of real life that we also.

But I do think I don’t want to lose the member piece that was John’s point is that’s the big deal.

DG: John?

KG: John. Yeah. John made the point about like the navigation. And we have these like consumers going through these crazy mazes. That’s the big piece where I don’t expect Brendan’s, physicians and nurses to educate members on their benefits and what happens after. That’s why it’s a collaboration. We can take that role on and help them so that they’re not trying to figure out, what am I going to do when I get home?

BC: Can I go one more time?

DG: Sure. Yeah.

BC: But you’re holding the risk, right? Which means that now you’re looking at an additional stay, keeping with the length of stay, an additional day in the hospital, cost versus whatever cost for food and, you know, electricity.

KG: $3,000 to $10,000, depending on their condition.

BC: Yeah, yeah, yeah. Your ability. The tra- I almost said trade off. The trade off there is you could spend, you could since you’re holding the risk, you could spend money on something that costs less, you know? Whereas for me it’s binary. Are they safe to go home or are they not? And why not means 3 to $10,000 for you. Whereas you could say, I can turn your electricity on and get you food.

KG: And David, you made this comment too, that that’s part of the unnecessary care in the system. If people, you know, Brendan’s trying to get more people in to help them in the, in the in the bed. And we have a lot of people, right. Who this, this becomes an exponential.

DC: It becomes. Yeah. And I’ll just add one quick point, which is this shows up in every social program, not just health care. You know, we sort of say, well, people are being displaced by AI or competition from China or whatever. Let’s do job training.

And job training often doesn’t work. Who’s taking care of my kids while I’m in job training? How do I get there? I don’t even know what to do, what am I being trained for? And so on. So one of the things that we’ve learned across the board is that as exactly the conversation was you have to not just present it to people, just say, you know, here’s a bus ticket, go do whatever. And because people are like, I have no idea what to do, you have to actually be able to work with people through stuff. And the places where we’re successful are very high touch in those difficult moments. They’re very high touch. Not in all cases. Sometimes people are fine.

DG: And very high touch is very expensive.

DC: High touch is more expensive than a cheap system, but it’s probably less expensive than, it’s less expensive in the end, because the cheap version doesn’t get what you want, so you’re stuck. But the problem is, no one upfront will pay for the high touch when it’s needed because they say, well, I can conserve on that without recognizing what it means on the back end.

DG: And I just want to point out that this is supposed to be the solution section. And like, we can’t help but sink back into the problems, right? Because like, that’s what this is. John. Solution, buddy.

JL: Solution.

DG: Help us with the solution. Betting it has something to do with Amazon.

JL: Well, I would just say, in the seat of a pharmacy, there are really two concrete things that I think we’re accountable and responsible to do. And we’re positioned to do this. And it relates a little bit to what you guys were saying.

One is access and one is affordability, because the messiness of life, health is not distributed equally, and access to health systems is not distributed equally.

But pharmacy fundamentally at its core is delivery of a clinical service and a physical product. And so we can help with the idea that, you know, you don’t need to miss work or miss a kid’s soccer game or get on a bus because we can deliver those medications to patients. And so on the access front, we’ve spent an enormous amount, and we continue to invest ahead of the curve around putting medications and pharmacists close to customers.

So by the end of the year, we will have 4,500 cities and towns that will get medications in under 24 hours delivered to their doorstep. And our pharmacies are open 24/7. So you don’t have to sacrifice or trade off something else in your life. You can go see it at any time. You can see the medication, you can look at research, and we’ve proven that even just making prices transparent, improved adherence, we didn’t change the price. We actually did nothing structural to the price. But people having the confidence they could afford their medication helped. Getting people medication quickly right after receiving treatment helps them get on therapy, you know, making it easier to get your medications and not trade off other things in life helps people be adherent.

DG: Well, one of the words that you brought up first and other people have echoed the word is friction, right? And it seems like one of the things that Amazon, based on some of the conversations you and I have had, is Amazon Pharmacy is really trying to identify where some of the frictions are and trying to minimize those frictions.

And to that point, just really quickly, because I want to get to Stephanie here and move on with the conversation. But what’s one really concrete thing that you guys are doing that reduces the friction?

JL: Well, I think price transparency is one. And so we share that-

DG: Walk, walk, walk the room through how that works.

JL: Well, how that works is not only have we created discount programs for cash and by the way, on Amazon Pharmacy, twice the national average of consumers use cash because we’re reaching people who are underinsured, uninsured and couldn’t otherwise don’t have an insurance instrument.

And so we have Prime RX, which is a discount program on branded and generics up to 80% off. We created RX Pass. Our chief medical officer worked with our clinical team. This is 60 common generics for $5 a month. That program measured, we published a JAMA study, 27% of patients had more medications on hand to stay adherent and save 30% of dollars out of pocket.

We also work directly with pharma manufacturers, where they’ve got discount card programs. In the industry. On average, 15% of those get claimed because patients can’t navigate the system. That’s a source of friction. We ought to clip that for them. So patients have saved over $200 million on women’s health, diabetes medications, insulin and GLP-1s. And so those are concrete steps that we work and we’ll work with, whether it be a payer or a provider or a regulator to integrate it so that the patient feels agency in their their visit.

DG: Stephanie, solution.

SC: Yeah. First of all, say solution again, market capitalist, I believe we can drive affordability with market based solutions. I think about that in two ways.

One, you got to get the demand side right. You’ve got to make it easier to be healthy for consumers. You’ve got to make it easier for patients to navigate the system. That’s really confusing and complex. I want to double click, though, into the supply side and get tactical there. Where how do we drive down costs in the system?

And I think the answer or the philosophy is really simple. It’s we’ve got to get to a system that pays for outcomes. It’s not did you see a patient in your office? It is. Did that visit result in that patient getting healthier? It’s simple. Breaking it down. How do we implement that in the very tactical things that we’re doing to drive policy?

At some level, it’s taking on sacred cows every day. It’s taking on legacy systems. We set payment rates for Medicare Advantage a couple of months ago. A lot of, you know, we like the Medicare Advantage program. We like the idea of private plans and Medicare. But we wanted to get to a point of parity where, you know, we believe Medicare Advantage plans can deliver better outcomes at the same payment rates as original Medicare can.

And we’ve looked at the data. We believe it’s possible, and we want to encourage those plans to reduce costs, to deliver better outcomes. And we took a lot of criticism for where we set the rates that were a lot lower than expected.

I think another area that we’ve really focused on is how we pay hospitals and how we pay for procedures. More specifically, we don’t really care if if a procedure happens in a hospital or an outpatient or an ambulatory surgery center, if it’s the same procedure and you get the same outcomes, it should be paid at the same rate. And that has been a philosophy we’ve implemented with our site neutral payments. More to come there, but I think it’s starting to change the results that we demand from the system and that patients deserve. I think there’s a broader philosophy that isn’t just our philosophy.

My friend Liz Fowler ran the Innovation Center for President Biden, talked about her goal of having every senior in an accountable care relationship by 2030. We share that goal. We want to double down on it because we believe that results in better value and better health outcomes for patients. And it’s those kinds of tactical choices that move the needle towards better outcomes.

DG: What, a version of the question I asked earlier, what needs to happen tomorrow in order to have more people in those accountable care organizations? What’s not happening today? What needs to happen?

SC: Yeah, we’ve gone through, when I was… accountable care relationships can include Medicare Advantage. So we’ve already done some tactical things there. It also includes the spectrum of things that are accountable care relationships. We’ve just released our new version of accountable care organizations, which moved from reach ACO to lead. That took …

DG: I’m so sorry to interrupt you just because I’m realizing you and I are talking jargon. Can you explain ACO to folks real quick, I apologize, I want to include the people in the room.

SC: Yeah. It’s a tool that Mark McClellan actually started. It was sort of enshrined in the Affordable Care Act. There’s a lot of bipartisan support for the concept where instead of paying per visit or per volume, we start to say, here’s a capitated rate. We’re going to hold you accountable for quality and outcomes. And if you save money, you get to keep it as long as quality is good. If you go over, you assume the risk of it. And that has grown gradually. And we’ve just released our new version of lead ACO that makes those improvements very tactically.

DG: When we come back, our panelists discuss the potential for artificial intelligence to lower costs and answer questions from the audience.

BREAK

DG: Welcome back. We’re listening to a conversation I moderated about health care affordability at Aspen Ideas: Health on June 25. 

On the panel were Mount Sinai Health System CEO Brendan Carr, Aetna COO Katerina Guerraz, Amazon Pharmacy’s John Love, CMS Deputy Administrator Stephanie Carlton and Harvard economist David Cutler.

So moving from solutions to perhaps a meaningful solution from all the different stakeholder perspectives here. Artificial intelligence, the continued march of technology, that’s something that your predecessors did not have access to.

There’s a lot of hope being poured into AI, as we’ve been hearing the last couple of days in various ways here. And at the same time, I certainly remember the electronic health record revolution and how that was going to revolutionize. Oh, so much was going to change. Everything was going to get better because we were going to finally have electronic health records for patients.

So, and yet still here we are, right? We still have these problems. So in your most honest heart of hearts. Not aspirational, but realistic and pragmatic. On a scale of one to ten, how likely do you think it is that AI and other tech is going to help us make this big leap forward on making health care cheaper, better and easier to access? Stephanie, start with you. We’ll just kind of go back down the line again.

SC: I get 90 seconds, right?

DG: [Laughs]

SC: I’m going to start with a five. And let me caveat it because there’s a big asterisk there. It could go either way. Our history is technology often makes health care more expensive. But if you get the payments right, it can have a deflationary effect.

And so the potential of AI really depends on how do we pay for it. What are the incentives we’ve put in place, do consumers trust it. And if we get those things right, I’m a huge optimist. I think we get to a 12. Some of the reasons I say that if you think about health care, our productivity rates far lag every other industry in the United States. Part of that is we’re highly regulated. We care about safety.

Our labor costs are about half the cost of health care, and we spend a lot on administrative costs that aren’t adding value and improving health outcomes. Those are things that AI can really help with if we get those incentives right. And if we get the change management piece of those things right. So I’m a huge optimist there. That’s on the supply side and how care is delivered.

The other part that’s really exciting to me is it’s a tool for patients to radically transform how they experience the health care system. It helps them ask the right questions about is this costing what it should? Is this doctor actually going to make me healthier? How do I negotiate with my insurance company? Do I need to? Can I go to basic questions, or have a fun conversation earlier with John about some of the things that Amazon is doing with AI and some of the questions, I’m a doctor’s daughter. I grew up called my dad whenever I needed to. This is a tool that makes that easy for every patient in America to ask questions at, you know, midnight when you’ve got a kid crying and you’re wondering what’s going wrong. It really can change how patients experience navigating their own health. If we get some of those things right about how to pay for it, how to regulate it safely and responsibly, and how to drive adoption in a way that is based on trust from patients because it works.

DG: And just real quick, because we don’t really have time. So I’m indulging myself here. But you, you do a five, right? Like, which is pretty good, but it could be a 12. I heard it, I heard I heard all the way up to 12. But, but given how sort of bullish you could be, five is pretty low. And I’m wondering, is it because of the pessimism you have in our ability to figure out the pay, figure out the regulation, figure out the adoption? Why? And I’m not trying to dampen the bullishness. I’m trying to understand the realism of why the five.

SC: Yeah, I think because I think history has taught me to be humble. When you think about our experience with technology that often makes care more expensive.

If we get things right, we released a payment model for the first time paying for AI in December, but we would only pay for it if it improves health outcomes. And we are launching that model this summer. Those are the kinds of things I’m really bullish on. If we can get those right and we have the innovation center to test it moving forward. So I’m optimistic and we’ve got a great team that’s working on those things. Abe Sutton runs our innovation center and has spent time in the private sector, as well as in the policy world, and gets that kind of balance of real world versus the wonky DC policy.

DG: 1 to 10.

JL: I’m going to give a single number. I’m very optimistic. I’m an eight or a nine. And I’ll tell you why. One we’ve always had technology driving innovation. We’ve got X-rays and MRIs. You know, our doctors have been using software for quite a while. Digital health records.

I think AI is a very promising new tool. The way we see areas that’s helping us right now today at Amazon Pharmacy. And some of those are just back office. It’s working on the friction and the sludge. We’re using data entry. This is transcribing a fax or e-scribe or something written in doctor shorthand so that it’s clean and organized so that a pharmacist can make a clinical evaluation and a patient can understand their directions. That now happens 94, 95% of the time. The AI can do it. It does it faster than a human four times the accuracy. And so that’s a huge cost time saver and it improves quality.

We’ve got AI working on rejected claims. Every PBM has different requirements and things prefix, suffix, codes, diagnostic. So when we get a rejection back which is blocking a patient from access to their medication, it often comes with reject code in a short piece of language that AI can go look and say, oh, I can go find that right piece of information, revert that back, get a paid claim for the customer.

So that’s saving doctors and pharmacists time on the call and faxing each other back and forth, trying to resolve basic administration. So those are obvious. I think the natural language creates an interface for people to engage in their health that we haven’t had before. If we walk down the street and ask them about ACO reach or whack and gross to net pricing and pharmacy or the complexity, It’s incredibly hard to navigate.

So creating a natural user interface, we see people engaging in their health now in a way that’s incredibly hard to do. If you’re going to go read, you know, 500 pages of health content. And then I think it assists providers and clinicians, things like open evidence. And we’ve seen it in even low acuity care where a human is treating a patient. But there’s something buried in free text on page 200 of their health record that is a counterindication. And the AI can help just serve that up. So the clinician, you know, makes a more informed decision. So those are all happening now. And then the last thing in the future is I think we’re going to be able to match AI augmented or lead care at a much lower cost, which will take cost out of the system for highly protocolized, low acuity care.

BC: Yeah, I got it. I’ll go fast. We’re as a health system and I personally incredibly bullish on this, but I’m a six. And I’ll tell you, you just heard a framework emerge right? There’s the framework here is there’s administrative stuff. We can create a lot of efficiencies. Parentheses. Massively displaced. Workforce. Civil unrest. Let’s just put that down there. Right. To remember I’m the realist at the at the pointy end of the spear that that employs those people that are doing things that we’re calling administrative. I think sludge was the word.

JL: Well, data entry, they’re literally looking at one screen and typing information into another screen.

BC: And paying their mortgage and sending their kids to school. Just saying. Right.

JL: The jobs will change.

BC: But we got to figure that out. But there’s administrative back office stuff. There is this incredible democratization of information and empowerment of patients, and that blurs right into clinical delivery on the on the provider side.

So my ring, my watch, my medical record, all the things that, you know, my phone knows about me. It’s impossible for me to know what to do with it until it gets synthesized and becomes knowledge on the AI front. Right. And that’s true for my doctor, too. The last edge of that clinical side, as we think about this continuum is are we going to trust totally virtual doctors? I have seen some demoed versions that are pretty incredible. Are we ready to say we have an access problem in this country? Lots of people have no health care. Let’s give them something that is maybe as good as a doctor. Maybe better. We don’t really know. It’s a natural experiment. Let’s see what happens.

Last thing I’ll say is I don’t have a lot of confidence that our political process right now, or trust in government or sophistication of regulations on artificial intelligence is going to deliver us to the promised land. I would love to be a 12 not a six.

JL: I, by the way, I do think that’s a challenge, which is the science is moving fast and regulation tends to move fairly slowly. So that’s just a very difficult thing to build a framework to create a safe piloting of what you just talked about.

BC: Which is what pulls me down.

JL: Yeah, that’s a challenge.

KG: So my score is a six. We have deployed…

BC: Payer provider agreed.

KG: So a six because AI we’ve deployed and we continue to deploy a lot of tools. What I think is going to be best at is improving productivity. So we’ve done it with all of our clinical operations. Our nurses aren’t taking an hour an hour now to collect all the patient information, to do an outreach and talk about gaps in care and connecting them to care. That’s been so that they can do more outreaches. Do I think it’s going to take so much cost and completely make jobs obsolete? No. I think AI is about augmenting the workforce.

I think the bigger thing we haven’t talked about is the single most transformational thing. And it starts with your EHR comment, Dan, is interoperability. The faster that everyone’s actually sharing data seamlessly and exchanging it, we’re going to be able to identify trends faster. We can connect people to care faster. It’s all about the data.

DG: Thank you. David?

DC: I’m going to give an eight.

JL: I was counting on you David. Thank you.

DC: I gave, I started off with four types of issues and I think three of them AI can really address well. So it can address issues of preventive care because you can reach out to people very well. It can address issues of what care does this patient really need? Because it’s very good at doing prediction and it can address the administrative costs extremely well.

The one of those that it doesn’t address very well is the price. If the price is 1000 in the US and 100 in the UK, that just doesn’t work. But I want to make one point, which is to try and wrap some of this together, which is in health care you cannot think of technology by itself. So there are three things that have to go together.

One is the technology, second is the people. And then third is the economic incentives. And you screw up any of those three and you won’t get it. So, for example, the technology can indicate who is at risk of being rehospitalized. That’s where you need the people to then help you out. By the way, that’s what I would like to do with some of the workers who are very organized, who are no longer coding things from one screen into another, has moved them into helping people navigate what’s truly going, what’s what’s truly going to improve their health. So I want the people, the technology and the economic incentives. If my profession of economics doesn’t do our job well, what we’ll do is we’ll create bot wars where his bot gets paid, tries to figure out how to get paid more from her bot, and her bot tries to figure. And so in milliseconds, they’ll be changing their coding frames all the time.

KG: Can I just make one other comment? Because one thing also, just about all the work I know we’ve done in AI. AI doesn’t actually do a better job if you don’t have the right processes and systems anyway. So unless you’re ready to redesign it, that’s you just can’t throw AI and thinks it’s going to improve everything.

DG: So a question that I would have, I don’t want you to answer it. But what, what kind of like model or demo could we do to create some sort of like SIM city kind of attempt at watching this? David. Have you given some thought to that?

DC: Yeah, we ought to be. Actually, I’m glad you said it because we could actually do demonstrations like clinical trials of this in the way that we do clinical trials of any technology. And we could take a health system like Mount Sinai and a payer like Aetna and say, look, let’s just put this together and figure out…

Gorenstein: This is the origin story. Right now Aetna and Mount Sinai.

KG: We already have like three ideas from this morning.

DC: I knew them before they merged. But you could actually do that. And you could see what the outcomes you get are.

JL: Sign me up. We carry the interoperability flag.

DG: Okay, I have one more question, but I’m not going to ask it because I want to get to your guys questions. We have people with the mics. It’s awesome to see folks asking questions. The folks with mics are going to come around.

Audience Member: As a clinic. He has a clinical psychologist. I think the only thing we didn’t talk about is the patient. What is my responsibility to this problem as the patient? What do I need to do? What do I need to do better? What do we need to train the patient to do better? To make this system work better?

DG: Who wants to take that one first? 

KG: I can take one example. And I’m just going to use because you’re a psychologist, behavioral health. So I’m going to get to the patient. So one of the other big problems in this country is people think like they’re trying to get an appointment with a therapist. They’re trying to get an appointment with a primary care physician. It looks like in the directory this person takes new patients. And by the way, they’re both in Burbank and Santa Monica. But then when they go to the call, that’s not true, right? It doesn’t happen.

So our job is, and we’ve done this with member transparency tools is we we have had to work around that system because that relies on a lot of different provider groups to give us their real time data where we just now look at claims and we tell a patient, hey, this doctor actually just saw six new patients. And we know that from claims data in the past two weeks, we have a pretty good idea that they probably take a new patient. And also, by the way, they’re not in Santa Monica. They only practice in Burbank. And we do that with a lot of just data scraping. So that is to give the the patients responsibility is we have to give them that information, accurate information, and that takes the friction out from them to make a decision.

DG: Anybody else?

BC: I would go so much bigger. And you know, I talked about the conversation we’re not having as a country. We tried over the last couple decades. Right. Some of the architects of that conversation in the room, the country does not want to have a serious dialogue around whether or not we’re going to have skin in the game, whether or not we’re all going to expect to have everything all the time.

I told a story of someone visiting from another country, got in a cab in New York City, talked to the cab driver, cab driver had two cancers and a transplant had been cured from all of them. He looked at me and said, does a cab driver in America think that they should deserve to get treated for two cancers and have a transplant? Because where I come from, in a caste system, that is like incomprehensible, right? So, you know, the American expectation of health care is inconsistent with what we have agreed we will spend on it unless we create incredible breakthroughs using AI. Maybe that’s your 12.

DG: Stephanie real quick. And then we’ll get to the next question.

SC: Yeah. I just build two things. One, the choices that we make every day about getting outside, exercising what we put in our bodies during meals, like the healthy choices that we make, not only for ourselves, but for our families. That has an outsized effect on our health.

And I think the other thing is go experiment with different tools. I mean, it’s fascinating what you could do. Like any tool you pick with AI right now and like what it starts to teach you and tell you. And experimenting and just becoming more educated consumers, that gives us greater control and pushes the system to be more accountable to us.

JL: And patients need to demand that it behaves in their best interests. Like you are a consumer. I think a lot of people just say, oh, health care is what it is. And I’m just in this malaise and we need to keep pushing the standards.

SC: Yes.

Audience Question:: Hi. So my question is, for somebody who’s probably not up on that stage is what about the medical legal costs and over utilization? We talked earlier about over utilization of imaging labs, etc., based on patient need for medical anxiety. Where does that fit into this conversation?

DC: I can take a little bit of it. So the economic incentives are incredibly important. In addition to the patient and physician uncertainty type issues. Less liability these days because most states have made it far more difficult for people to receive compensation. But certainly the economics, a lot of healthcare organizations will say, I can’t really afford to invest in prevention, for example, because A, I’m not paid for it and B, it cuts back on the things I do earn money from, which is the acute care visits and stuff. So the economic incentives writ large, of which that’s a part are hugely important here.

BC: Which is why I would just say sort of listening to Stephanie’s comment from before, when we start taking pieces of it that we don’t like, that don’t make sense, right? We chip away at the example you used was site neutrality. Should you get the same bill if you’re treated in an outpatient center versus at a hospital or you know, the other examples that are that are policy talkers right now are the 340 B drug benefit program. I mean, there’s lots of conversation around whether or not for profit health systems deserve their and earn their not for profit status.

It really does beg the question, right. If if if we’re going to chip away at these pieces, we do have to zoom out and say, well, how did we end up here? Right. And there are foundational payment models that have created these, these strange Band-Aids that have been put on the system. And so, you know, you, you’re asking a massive question, a massive question. It begs the question around whether or not we can start to chip away and fix this system, or if we should just be, you know, starting over.

DG: Next question.

Audience Question: Two provocative comments that I want to panel to comment on. The first is about patent law. Look at look the GLP-1s. We allow drug companies to get new patents if they synthesize a new chemical, a new molecule, but it goes after the same target. I don’t think that’s novel. And the consequence is to raise pharmaceutical prices, where instead, we should encourage pharmaceutical companies to develop drugs where we need them the most. That cost the country the most money, like an orphan drug addict for the most for the most expensive diseases. And the second point that I want to make is I know of no other industry other than the insurance companies in health care that add as little value as they do. And I’d like you to comment on that, because prior authorization denials, they just increase our costs.

KG: Yeah. So I’ll start on prior auth, which obviously has been in a headline. So I think across the industry, the majority of payers we were talking about this morning would tell you that that’s something that hasn’t worked for a whole host of different reasons.

Just a couple of things. One is prior authorization, different rules. We have federal regulations for Medicare and Medicaid that require us to have it. And we have large employers in this country, self-insured employers, who also require for different reasons.

I will tell you, the industry as a whole has made some big commitments in terms of. And that’s something Stephanie and the administration has been very involved with driving is to make sure that there is some standard, there’s more standards. Number one is reduce the number of prior authorizations. And number two, create standards so that from a health system point of view and from a provider point of view and from a patient point of view, you’re still maintaining there’s some level of still safety and quality of why these exist, but at least it makes it takes all that friction out of the system. If all the payers are using the same questions and the same clinical documents.

DG: Anybody else?

JL: I would just add on that topic, there is a material lack of transparency. Your doctor doesn’t know what medications are on your formulary and might be clinically indifferent to a variety and with all best intentions, be running into prior auth and cost overruns and things. So it’s not always just even ill intentioned, it’s just complexity. And so major payers and PBMs do not publish which meds are on prior auth and which costs. So both the pharmacy and the providers and everything are all banging into each other. And these things are happening and causing a lot of friction and a lot of pain and a lot of cost unnecessarily, because we don’t have transparency and interoperability.

DG: That’s just, we’re at time, but I see there are a number of people. I will stay a couple more. We’ll stay for a couple more questions here because it seems like people are interested.

Audience Question: I would say that on the 250th anniversary of America, that the cab driver should expect that his two cancers and his transplant would be paid for. That he would not die in this country as he would in his home country, and that we are smart enough to find solutions for that.

If there was a patient on the stage, perhaps we could contribute to that conversation in a meaningful way, rather than just being seen as being greedy consumers of wanting everything.

I would say that in working with patients and patient advocacy organizations, we can think of hundreds of ways to reduce care, to pay for things that we use to improve our outcomes that are far less expensive than what our insurer would cover far down the road. And so I would encourage everyone to partner with patients, and I would love to see a future panel on how each of you from your domains are partnering with patients to come up with solutions to these problems that we do all share responsibilities for.

DG: Thank you for the comment. Is it just really quickly, is there one, do any of you have a nice concrete example of where you are partnering with patients?

SC: I would just say real quickly, I started my career off as a labor and delivery nurse working directly with patients, and it’s really formed my philosophy and how I go about policy making, because at the end of the day, it is all about the patients, and I want them to be able to access care because it’s affordable and I want them to be able to access care that works. So I appreciate your point. It’s really powerful.

BC: Patient family advisory portfolios across our organization.

DG: Explain what that is.

Brendan Carr: An opportunity for people that are patients to interact with people that are taking care of them and trying to build systems and structures. I mean, you know, I mean, to me, this feels like this is 100% in the DNA of what we do at the pointy end of the spear. At the end of the day, we also have to balance a PNL. And this is why we I said before we run a high school, you know, what are we doing running a high school? We’re a health system. But they have such profound behavioral health challenges. They can’t stay in high school. So we partner with the Department of Education. You know, that we have lots and lots and lots of patient engagement. We often have to find compromises that to them feel inadequate because we are responsible to, you know, run the business responsibly.

DG: David, do you want to say something and then we’ll wrap it up.

DC: I’ll just say on the on the technology development end, the patient advocacy groups have been enormously important in channeling physician researchers into where to go and saying, no, we don’t need this. We need this instead. And so it’s actually made a meaningful difference in a number of conditions in what the nature of what folks work on are. In addition to how easy it is, many patient groups have put together registries of patients. So it’s easy for clinician researcher clinicians to get groups of patients to run clinical trials with or do other kinds of testing on. So it’s actually been enormously valuable for the health system.

DG: I think ending this on patients is an appropriate place to stop this conversation. Let’s give it up for this panel. Thank you all very much. Have a good day. Enjoy the rest of the conference. 

Thank you for listening to this conversation, recorded live at Aspen Ideas: Health.

We’ll have links to the video of this panel in our show notes.

I’m Dan Gorenstein, this is Tradeoffs.