New Marketplace
Navigating Payment Reform

The Highest Quality at a Lower Cost? We Don’t Have That Yet

Interview · November 20, 2017

Patrick Conway and David Cutler head shots

David Cutler, PhD, interviews Patrick Conway, MD, President and CEO-Elect of Blue Cross Blue Shield of North Carolina.


David Cutler:  This is David Cutler, professor at Harvard University, and I’m talking with Patrick Conway, who’s President and CEO-Elect of Blue Cross Blue Shield of North Carolina. NEJM Catalyst sent out a [note] to its advisors asking, who would you like to hear from? And I responded, “I’d like to hear from Patrick,” and then they said, “Great, we set up an interview for you with Patrick.” So I’m going to ask you what it is I was hoping someone else would ask you, but we’ll see where the conversation goes.

Patrick, you were one of the fundamental people at CMS responsible for delivery system reform and payment reform efforts, and you made an enormous amount of headway there. What I want to start off by asking you is, what do you think were your biggest successes there, and then understanding that in everything that one does in life, some things don’t work as well as we hoped they would, what things didn’t work out as well as you hoped they would?

Patrick Conway:  Thanks, David, and thanks for having me here today. I’m in my first month at Blue Cross after transitioning after almost 7 years with CMS, and it’s really terrific to talk today. In terms of delivery system reform, the CMS Innovation Center, which I got to lead for a number of years, [and] I think the biggest sort of macro-success I’d call out is Medicare. In 2012, we had 0% of payments in alternative payment models, where the providers were accountable for quality and total cost of care, and by 2016, we were ahead of schedule with over 30% of payment in alternative payment models like accountable care organizations, bundled payments, comprehensive primary care. And really a shift in Medicare toward these alternative payment models, over $200 billion of 30% of payments, actually, providers and patients in these alternative payment models.

I was on a panel with Don Berwick, and he talked about how we’ve seen more positive change in the delivery system over the last 3 to 5 years than he thought we had ever seen, which is kind of him to say. He also said, and I agree with him, we have a lot more work to do. We need the best possible care for every patient every time, and then on your specific [question] what are the biggest successes, I think taking accountable care organizations. You now have heading toward 500 ACOs in the public market, and even more in the private market.

Some results just came out demonstrating increasing savings over time and quality results, which aren’t talked about quite as much and are quite positive in terms of high-level quality and patient experience and improvement — I call out bundled payment. Take hip and knee replacement as an example: JAMA studies are showing improved quality at lower cost, and actually in North Carolina, also showing improved quality and lower cost here at Blue Cross. And then, I call out primary care work. We still have more work to do, but I think we’re shifting the primary care model through models like Comprehensive Primary Care Plus, where you’re investing in population-based primary care.

And then lastly, prevention. I think the example I’d give there is the Diabetes Prevention Program, where CMS built on evidence from NIH and CDC, tested Diabetes Prevention Program, a community-based prevention program for diabetes, demonstrated it could lower cost — that’s certified by our actuaries — and improve quality, sustain weight loss, and now, nationally in 2018, with that benefit and program, we are now available at the community level to Medicare beneficiaries across the country. So those are some examples I’d call out.

Cutler:  What about on the other side? What things didn’t work out as well as you hoped they would?

Conway:  I’d say the challenging things that I struggled with until the day I left were the pace and the pace of learning, and maybe this is my natural frame, from a quality-proven frame, but it still took us too long to get new models launched. We were trying to learn in a much more rapid cycle, and I think got better at that over time, but it just was challenging to evaluate and learn at the cycle you need to, to improve.

I think one of the things, quite frankly, [why] I’m excited to be here at Blue Cross, is I think we can be a little bit nimbler at a state level with Blue Cross in the private sector, implementing programs and learn at an even more rapid cycle. And so those, I think, are some of the things I’d call out that were challenges in areas that we need to improve on over time.

Cutler:  You mentioned four specific types of programs. You mentioned the ACO program, the bundled payment program, the primary care work, and prevention work. When you look at the evidence as part of the debate that’s been going on between some people who say they saved money and improved quality and that’s good, and other people say, yeah, but they saved a little bit of money, they haven’t transformed health care in the way that we thought they would, and quality has gone up, but it hasn’t transformed quality the way we hoped it would. You sounded more like a glass half full than a glass half empty person, but what is the right way to read our success relative to what might have been?

Conway:  Yeah, I do have a natural optimistic frame, but I think the best evidence from some of your colleagues are things that like lawyers and attorneys pay for. It really looks at the overall effects, including market effects, which are much better than — I won’t cite names, but people who try to say investments based on growth figures that actually are not an evaluation. The ACO program, just to stick with that example, I think the McWilliams and Chernew paper and some of our formal evaluations, you know, years from now and that, it’s savings, and significant savings that also on the order of one to two points. The ESRD program that just came out was actually a little over 5% savings, but often, that sort of one to two range.

I would say that’s not a double-digit number, but I would also say if our U.S. health system shaved one to two points off trend for the long term — Medicare is in a more sustainable place, private health insurance is in a more sustainable place, so I want those savings numbers to grow over time. I also think sometimes, especially in D.C., we lose sight. A lot of these programs [just] launched — when I started in early 2013 in the Innovation Center, there were just a couple of models.

A lot of these programs really launched over the last 3 to 4 years. I think we need to give some more time to see those numbers grow, and also in the ACO program — I think you know this David — the ACOs that have been in the program longer are on average saving significantly more money and having better quality results, so I’d point to those pieces of data and evidence.

Cutler:  Can you make anything of the fact that the ACOs that are saving more are the ones that are not centered around a hospital?

Conway:  One, I would say we have ACOs of all types saving money. Like most things, there’s variability, so we have an urban, rural, hospital, physician-led, etc., saving money. That being said, you’re correct that the physician-led ACOs, on average, are saving more. I think my hypothesis there — and we did some research on this, as have others — the incentives are completely aligned in that instance. If you’re a physician group, if you keep people out of the hospital, it is in totality good financially and good for the patient.

If you’re a hospital or a health system that still is in a fee-for-service world for a good part of the business, your incentives can be competing, which I think can cause challenges. To take it back, when I was at Cincinnati Children’s, when we had the biggest success was where we could totally align the incentives on population health, from the C-suite down to the frontline nurse to the frontline primary care clinic. And we were able to do that, but in health systems, it can be a little more challenging given there’s some competing demands at times.

Cutler:  You said a couple of things. You said some of these programs will need more time to mature. You also said that it will help if the incentives were aligned more fully.

Conway:  Yeah.

Cutler:  Do you think we need additional action to get from 1 to 2, or 4 to 5% savings to 10 to 15 or 20% savings? And if so, what action should CMS take, or private insurers take, or employers take, or anyone take?

Conway:  A couple of thoughts. One — and you know that I try to think with nuance, so let me describe — so like in bundled payment, you are seeing it, for hip and knee, the results here at Blue Cross are close to 20% savings. Our national results are close to 10%, so I think in a given condition, we’re seeing people successfully align incentives for even a double-digit space, and it’s by definition a little less complex to do that for a singular condition, but I think you’re seeing people align physician and clinician payment, quality measures, and sort of care packages for conditions for double-digit savings.

For population-based payment models: next-generation ACO in Medicare was the biggest one, with the highest level of risk, [and] we’re seeing some people get close to double-digit savings, but it’s not across the board. I think there are challenges,  some of these organizations just need more time, and we need to have more direct alignment all the way down to the physician and clinician level on incentives. I’m sure people have looked at this. The alignment within organizations is often not very tight —  needing the organization, say, for population, but their physicians are still paid on RVUs, fee-for-service.

That doesn’t make sense at all. I think some of this is evolving incentives and systems, because it’s not just incentives, it’s also the systems that allow you to improve. With primary care, for example, it’s incentives, but a lot of it is how do you give them the tools and technology and capability so that primary care clinicians can manage their patients outside the office visit as opposed to just the patient in front of them.

That is with the incentives and systems and tools in place so that those numbers grow to higher single digits over time, potentially double digits. I will end where I started, though. We achieved — as you know, as an economist — if we shifted the trend 2% in Medicare year-on-year across the board, it’s a whole different financial ballgame for the United States.

Cutler:  Give some advice to the Trump administration and members of Congress and interest groups. Given the successes that you’ve seen and the slowness that you correctly noted and where we are now, what should the goal be for the next couple of years? I’m going to assume for this, for hopeful purposes, that everyone wants to achieve higher-quality, lower-cost care, so it’s not an issue where you have to convince people that that’s a reasonable goal.

And leave out the coverage side, so just things about the Medicare side and stuff like that.

Conway:  In transitions, you get some time to reflect on stuff like that. It was two weeks before I started this job, so I moved and reflected for like a day — but I worked in the Bush administration on value-driven health care, and then I left government and came back in the Obama administration in the chief medical officer role, Innovation Center, principal, deputy, administrator — I had sort of every job at CMS at some point, or many of them — and served in the beginning of the Trump administration.

I would say we need to continue to focus on testing new models. And when I say new, let me clarify: oftentimes, it’s the next iteration of current models to generate these kind of results. We need to test —  and I’ve said this publicly when I was in the administration and so will say it now — we need to test the next generation of bundled payment models that move to the next level of bundled payment, and that applies to the public and to private sector.

For ACOs, we need to prevent — to give you an example, how do we get consumers more engaged? How do you have consumers and have consumer engagement at a much deeper level? I said that when I was in the administration, I’ll say it outside the administration. I mean, as a scientist and a physician, I fundamentally think that’s a place where we have additional improvement to go.

Specialty physician models: we have some models there. We don’t have enough in the Medicare program. Value, that’s a specialty.

There is an RFI, request for information, that came out of CMMI soon after I left that I’d worked on when I was in the Innovation Center. It calls out consumer directive models, health plan innovation, mental and behavioral health, where I think — I’ll be honest — a long answer to your question, but I think our mental and behavioral health in the system in this country is fundamentally broken, and if we don’t do better and fix behavioral health and substance use, it’s going to cause major negative — it is now causing major effects on our population that are negative and are not getting better. They’re getting worse.

We also called out state-based innovation. We did much of that work in the Innovation Center already partnering with states, and how do you partner with states even closer on their comprehensive delivery system reform at a state level? For example, in a state like North Carolina, how do you really generate the best possible results in any given state?

Cutler:  One sometimes sees on the other side that there’s so many different demonstrations and programs that kind of are getting in the way of each other. For example, if you do a bundled payment model inside an ACO model, who gets to keep the savings from consumer resources expended in hip and knee surgery? Is it the orthopedic provider or is it the ACO? One sometimes sees these calls for pruning, you know, figure out what works, and then go with the ones that have worked and try and jettison the ones that are not working. You seem to be on the other side of that. Can you reflect on that a little bit?

Conway:  Yeah. Clarify a little bit for me. I don’t know that I’m on the other side because I’d say internally in CMS, I would talk about the need for focus. And we actually did do that. In the very beginning of CMMI, for example, the Health Care Innovation Awards, where kind of the purpose at the time was literally funding everything possible across the country as a sort of hundreds of flowers blooming focus, which — by the way, at that time, that even predated me — was considered to have been very reasonable. But I think where we are now is focusing most on what is most likely to work.

The Innovation Center is a top model, for right now it’s in the high 20s in terms of total number of models. I’m not there any more, but I don’t think it should grow much beyond that. I think as things get added, things have to go away and get potentially even more focused, and we were working in that direction. We did have a team actually looking at this overlap issue of as models overlap, it creates new challenges.

One is the one you described: how do you share the money and how do you not double-pay as a payer for the same savings? We had created rules so that we didn’t double-pay, essentially, and then we said the market will work out how bundles and ACOs work together. What I was hearing from people is the market wasn’t always working that out, so how could the payer be a facilitator, or help the market to function, and I think that’s a valid issue to consider.

The second one, which I know you know a ton about, David, is evaluation. It’s increasingly challenging to do these evaluations. The positive is there’s so much change and intervention and providers and payers trying to drive change in the marketplace. The con is, it’s very difficult to find a real control group anymore because you have so much change in the marketplace.

Cutler:  There is no control group. Nothing is in a control group.

Conway:  Yeah, there is no control group. I should say, it is very difficult to try to find — and actually this is good, you can reflect back as a more expert researcher than me, I’m like a researcher who’s retired now or something — our team, I’ll tell you, they would work really hard to do the best possible job on the evaluations and use as sophisticated methods as possible, and we’d bring in external people to help us and have independent evaluators. You can tell me if you think I’m wrong, but I think evaluation is getting more and more challenging over time.

Cutler:  Oh, yeah. Yeah, in fact, people have to use not standard treatment control groups, but sort of treatment and then make some guess about what the control might have been.

Conway:  Yes. Exactly.

Cutler:  Because there is no control anymore.

Conway:  I don’t want to totally geek out with you, but we were trying to use like Bayesian methods and we were trying to estimate, without the intervention, what the effects would have been and estimate, well, within the actual innovation, plus market effect. I do remember my research days. I was never as good a researcher as you, but it was exciting to think through with the team the best possible way to do the evaluation.

Cutler:  Are there any payment models that you say, that’s really not delivering what we think it ought to, that’s not the right path to go down?

Conway:  Yeah. You know, I’ll give a couple. Our first primary care model, it was an FQHC model, literally just paid for primary care accreditation essentially without a linked cost. It did not work, and I think there’s a lesson there: pay for hard quality cost outcomes, not prophecies when possible. In the specialty physician arena, it’s an open question, how well we can design specialty physician models that truly improve quality and save money.

In the private payer space, there’s some in oncology, Medicare and oncology care models that we’re waiting for results from, but what about rheumatology, neurology? What about the array of specialty medicine? I mean, I can’t point to a lot of examples of specialty physician models demonstrably proven to improve quality at lower cost, but obviously that’s a large portion of outpatient medicine, but I think that scenario for both Medicare but also private payers to work on going forward. So those are a couple examples.

Cutler:  I want to try and end with a lightning round. I’ll ask you three things and get really quick assessments from you, if I can.

Conway:  Okay. You’re going to try to make this hard on me. All right.

Cutler:  True or false.

Conway:  You know, David, I’ve only been out of government for a month. It’s hard for me to modulate in any other way. But go ahead.

Cutler:  You’re free, you can say whatever you want. True or false: Medicare spending will be lower 5 years from now than they currently forecast it to be.

Conway:  Than currently forecasted. I think true. I think it will be lower.

Cutler:  Okay. Second question: 5 years from now, are we going to conclude that North Carolina has taken the lead in delivery system reform across the country?

Conway:  Yes, and I’ll give — sorry, I know you want lightning round.

Cutler:  No, no. That’s okay.

Conway:  I have worked my Blues colleagues, and we’re all partners. We want to make a lot of progress in North Carolina. Are we going to do that alone? No. We’re of course going to do it with partners, so we’re going to try to work with all the Blues and other health plans, of course. To answer your question on where do we want to be in 5 years, it is a great organization now, and I think there’s a huge opportunity to demonstrate the possible. What I mean is, a health insurance company and health solutions company that delivers even better health outcomes at lower cost and best-in-class customer experience. You know me, I like to set aspirational goals, and so that is an aspiration goal that we’ll work together with the team now.

Cutler:  My final question for you: over the next 5 years, Medicaid will be as important a driver of payment reform as Medicare. Accurate? Inaccurate?

Conway:  I know you want a one-word answer.  Medicaid, by definition, is going to vary at the state level, so I think you will see more space leading in Medicaid and payment transformation, which will be a good thing, you know, inevitably because there’s 50-plus states in Medicaid programs, you’re going to have some that are not as pushing the envelope and innovative. But the Medicaid program and caring for low-income populations is something that’s important to me for my career, and North Carolina is focused on Medicaid right now, and we would be very focused on serving that population.

Cutler:  Here in Massachusetts, we’re setting up an ACO model for Medicaid.

Conway:  I know. I was doing some work with Massachusetts on that. Just to riff on that a bit, on the multi-payer from both sides, when you can get the large private payers like the Blues plan in a state, Medicaid and Medicare all going in the same direction, imagine what’s possible. For any provider that’s 80+% of their business, and that will drive significant change. And for any given payer, if you’re working together across payers, we’d need many, many more examples, and I think at a state level, to do it across a large private payer like a Blue Cross, Medicaid and Medicare, I think you could see some amazing results at the state level.

Cutler:  Yeah, and we’re trying to do that here. I totally agree with you on that.

Conway:  Yeah. When I talk to Andrew Dreyfus and others in Massachusetts, you all are doing some terrific work. And I want bi-directional lines, so we want to learn from what you do in Massachusetts and from the Blues Plan and others in Massachusetts and likewise, you know, in North Carolina, we want to try innovative new work and also hopefully be an example for others of what’s possible at a state level.

Cutler:  I’m sure you know the dynamic at state and local government is very different from the federal government.

Conway:  Yes. I knew that, and I’m learning even more in every day, so it’s an exciting time.

Cutler:  In the final minute or so, what else should I have asked you?

Conway:  I’m still a practicing physician on weekends, and trying to do that still in this role. At the end of the day, it’s all the macro changes that we need, and I care about macro economics and macro changes and macro quality measures, but at the end of the day, in North Carolina, I’ve been interacting with our members, including a member yesterday who was a breast cancer survivor who literally thanked me for the amazing care that was delivered to her across the board from Blue Cross North Carolina.

My mother is a Medicare beneficiary, and I won’t bore you with the long story because I’ve told it in other formats, but she was hospitalized, and her care coordinator helped coordinate her care, keep her from being hospitalized again. At the end of the day, it’s how do we have a highly reliable system each and every time in this country that delivers the highest possible quality at a lower cost? We don’t have that yet.

We have pockets that are great, but we have too many times where a person or a family falls through the cracks, and so my goal at Medicare, and now my goal at Blue Cross, is how do you design and test it so to deliver that highest, best health outcome, higher quality care, best-in-class service experience, and at a lower cost. That’s what I would leave you with as a summary statement.

Cutler:  I’m tempted to conclude with how they do it at the end of Wait Wait . . . Don’t Tell Me! “If any of these come to pass, we will thank you very much.”

Conway:  David, thank you, and thank you to NEJM. It’s been a great partnership over time with all of you.

This interview originally appeared in NEJM Catalyst on October 31, 2017.

Call for submissions:

Now inviting expert articles, longform articles, and case studies for peer review


A weekly email newsletter featuring the latest actionable ideas and practical innovations from NEJM Catalyst.

Learn More »

More From New Marketplace

Hospital Readmissions Reduction Program (HRRP)

Medicare’s Hospital Readmissions Reduction Program was established by the ACA to combat excessive and costly readmissions. Learn about HRRP penalties, criticisms, and strategies to reduce readmissions.

Conceptual Framework for Evaluating Specialty Care Partnerships

How to Engage Specialists in Accountable Care Organizations

Should an ACO insource or outsource specialty care? Here’s a framework to help leadership decide.

Illustrative Examples of Health Policies, Possible Goals, and Relevant Evidence Base

Evidence-Based Health Policy

Having a clear framework for characterizing what is, and isn’t, evidence-based health policy is a prerequisite for a rational approach to making policy choices.

U.S. and Canadian Prices of Some Generic Drugs with U.S. Prices That Recently Increased by 1000% or More

The Price of Crossing the Border for Medications

The health and safety risks faced by the many Americans who cannot afford medications necessitate consideration of alternative strategies to provide less expensive medications.

Economic Investment and the Journey to Health Care Value — Part III: Health Care Purchasers

Early successes suggest that value-based purchasing programs can both transform employer-based health care and have a powerful and lasting impact on the economic strength of U.S. businesses.

Single-Payer Health Care Is the Favored Outcome of Future Payment Reform

Survey Snapshot: Deep Frustration with the Current Payment System

Many NEJM Catalyst Insights Council members are frustrated with the pace of value-based payments and expect single-payer health care to gain traction — though maybe not soon enough.

Value-Based Payment Models Payer-Provider Contracts Value-Based Arrangements

Economic Investment and the Journey to Health Care Value — Part II: Health Care Payers

Payers’ broad scale of investment in value-based arrangements makes a compelling case for the importance of sustained efforts to identify effective value-based payment models.

Four Principles for Navigating Payment Reform

The changes needed in health care are happening way too slowly. Health care stakeholders must insist on value in what they pay.

Economic Investment and the Journey to Health Care Value — Part I: Health Care Providers

Early evidence suggests that value-based payment and care delivery can transform our health care system, but providers must increase the momentum for this positive change.

Shift Toward Value-Based Payments in the Industry and at Organizations Is Accelerating

New Marketplace Survey: What’s Next for Payment Reform?

As health care reform struggles to gain traction legislatively, health care professionals report that payment reforms continue to move forward at a moderate pace, and indeed are essential to achieving the Triple Aim.


A weekly email newsletter featuring the latest actionable ideas and practical innovations from NEJM Catalyst.

Learn More »


Value Based Care

222 Articles

New Marketplace Buzz Survey: Who’s to…

This survey of NEJM Catalyst Insights Council members shows strong opinions about the impact of…

New Marketplace Buzz Survey: Who’s to…

This survey of NEJM Catalyst Insights Council members shows strong opinions about the impact of…

Platforming Health Care to Transform Care…

Health care leaders need to focus less on ownership and control of the delivery process,…

Insights Council

Have a voice. Join other health care leaders effecting change, shaping tomorrow.

Apply Now