Transcript of BioCentury This Week TV Episode 136
Gail Cawkwell, VP of Medical Affairs, Pfizer Inc. (NYSE:PFE), New York, N.Y.
Martin Marciniak, VP of U.S. Health Outcomes, GlaxoSmithKline plc (LSE:GSK; NYSE:GSK), London, U.K.
Dan Mendelson, CEO and founder, Avalere Health LLC, Washington, D.C.
PRODUCTS, COMPANIES, INSTITUTIONS AND PEOPLE MENTIONED
Jeff Allen, Executive Director, Friends of Cancer Research, Washington, D.C.
Marc Boutin, EVO and CEO, National Health Council, Washington, D.C.
Rep. Anthony Coelho (D-Calif.), Former Member, U.S. House of Representatives
Scott Gottlieb, Resident Fellow, American Enterprise Institute, Washington, D.C.
Kaiser Permanente, Oakland, Calif.
Jonathan Leff, Partner, Deerfield Management
U.S. Centers for Medicare & Medicaid Services (CMS), Baltimore, Md.
U.S. Food and Drug Administration (FDA), Silver Spring, Md.
WellPoint Inc. (NYSE:WLP), Indianapolis, Ind.
Steve Usdin, Senior Editor
STEVE USDIN: Is biomedical innovation making health care too expensive? Patients, payers and pharma say we can have both: cutting edge technology and affordable care. Today, part one of a two part special report. I'm Steve Usdin, welcome to BioCentury This Week.
NARRATOR: Your trusted source for biotechnology, information, and analysis, BioCentury This Week.
STEVE USDIN: The days when society could afford to pay anything for health care are long gone, but too many diseases still can't be treated. So we need to pay for biomedical innovation, too. Can the health care system deliver affordable value and reward investment in new medicines?
In the search for consensus, the advisory firm Avalere Health convened two summit meetings of patients, universities, drug companies, physicians, payers and investors. Many participants are familiar to BioCentury viewers. Jeff Allen from Friends of Cancer Research; Marc Boutin from the National Health Council; Scott Gottlieb of the American Enterprise Institute; and Wall Street's Jonathan Leff of Deerfield Management. Pfizer, GlaxoSmithKline and other drug companies were represented, along with insurance giant Wellpoint, and health care provider Kaiser Permanente. Over the next two shows, key players will discuss their ideas about value and innovation.
In this first of two shows, BioCentury is taking a closer look at biomedical innovation, and delivering value in health care. Today we're joined by Martin Marciniak, Vice President of U.S. Health Outcomes at GlaxoSmithKline, Dan Mendelson of Avalere Health, and Gail Cawkwell Vice President for Medical Affairs at Pfizer.
I want to start by asking all of you, there's this push towards value in healthcare, and it seems to me there's almost a catch-22, because health care payers are saying they don't want to pay for things unless value has been shown in the real world. But on the other hand, you can't get real-world data, or it's very hard to get real-world data, until payers start paying for something and things get out there. How do you resolve that?
MARTIN MARCINIAK: I think you resolve it when you look at the development paradigm. And you can think about a development paradigm and clinical trials that allows you to start bridging that gap. And so typical things that we think about are, truthfully, we think about meaningful endpoints, we think about relevant comparators and we think about target and patient populations. That allows you to bridge over into some of the questions of real-world evidence. Because once you start that, put that paradigm in place, it makes it a little bit easier to think about how the real-world evidence will allow you to bridge into discussions around reimbursement.
DAN MENDELSON: I think one of the things that we highlighted in our report is the fact that we need to come to consensus on what constitutes value. And if you can get groups of patients, and payers, and also manufacturers to come to agreement on what constitutes value, that's half the battle. And one of the things, again, that's very challenging, is the fact that there's so much change in the health care system, what constitutes value is also changing. So manufacturers need to anticipate what value is going to be in the future.
GAIL CAWKWELL: Just to add to that, I think that defining value can be very challenging, because value for patients, it's not all patients. You don't want to leave patients behind who have rare conditions, or rare adverse events, or from unusual groups. I think we live in this hugely exciting time with huge amounts of data available. And the trick is, how do we really leverage that to learn more, to know more, to add the value without falling into the trap of drawing conclusions that are not correct because we just overwhelmed with numbers and stuff.
STEVE USDIN: Sometimes the term value is used as a euphemism for price. What is the difference between price and value?
GAIL CAWKWELL: I'll start off by saying that, I think, value is the value that you're bringing to patients into society. It's that patient who is saying, this has made a difference to me. And that might be an outcome like survival, or it could be things like I could get up today, and because my pain is better I can take care of my kids, I can go to work, I can do the things that are important to me. And sometimes those are easy to measure and sometimes they're incredibly hard to measure.
MARTIN MARCINIAK: I think you're really grounded on the idea of clinical benefit, right? So it's a relationship between the clinical benefit and other outcomes that sort of come together to fill out a value proposition for a therapy.
And so what you want to do is have a solid value proposition that physicians clearly understand, that the patient population has clearly identified, and that there is a, really, sort of a relevant comparator that helps round out the story. So to what Gail was saying, you end up with this integrated approach to healthcare that revolves around the patient, and allows the asset to sort of come to the forefront of what providers may want to use or choose to use.
DAN MENDELSON: And I think value does start with a real focus on clinical utility and whether the patient is getting benefit from it. A payer will see value also relative to the cost of therapy. And that is something that is a reality. And when, let's say, a payer is trying to put together an insurance program, they have to be thinking about what the premium is going to be, and what the co-pays are going to be, and they have to be thinking about making sure that it all works.
STEVE USDIN: And Martin, you used the word comparator, which I think is really the critical thing in here. And that's the thing that's a change from the way the drug companies traditionally have looked at drug development in the past. How do you make that work in a drug development program?
MARTIN MARCINIAK: I think you make it work by, frankly, working cross matrix teams. I think when you look back at pharma in general, I'm sure it's no different for most of the pharmaceutical companies, the cross matrix teams come together really trying to pull in the patient focus, and look at what those alternatives are out there, in terms of treatment paradigms.
So the comparator is one part of it. The study design is another part of it. And then the execution of those studies, successfully, allow us to tell a story that revolves around what the value proposition is for providers, payers, and some cases policymakers. So for us, it's a pretty exciting time to have conversations around patient centeredness. And typically that plays out in different ways in the policy streams.
GAIL CAWKWELL: Just to build on that point, exactly. I think what you said about having a conversation with the right people at the right time is one of the things that's very different now. The right time could be not at a late stage of development but really very, very early. And talking to the right stakeholders, as you're saying, it's the patients, it's payers, it's policy decision makers, it's a wide group of people. And getting that input and figuring out, what do you build into your development program? What you do later? What's the right balance? What's the right weight?
DAN MENDELSON: You know, payers are increasingly demanding more information earlier in the process. And they're also demanding more post-marketing data. So it's not only information of the development program, it's also what happens afterwards, and how do you track it and monitor it? Because when you think about giving information to patients, patients care about what happens in the real world, as well as the safety and the efficacy of the products originally.
STEVE USDIN: Trying to generate that data for payers, does that kind of bias the kind of drugs that you develop, in a way? You know, to things where you can generate that kind of data in a clear way?
GAIL CAWKWELL: I think yes and no. I think in the end, probably -- it's certainly true with Pfizer, but probably true with most companies -- that we try to develop drugs that we believe will really make a difference. If a drug is very similar to something that already exists, it's hard for a payer to see why they should pay for it, it's hard for a doctor to see why to prescribe it. So a drug that really makes a difference, it's easier for the value proposition to come through for everybody.
STEVE USDIN: More than 30 organizations participated in the Avalere workshops in promoting innovation and value. Here are a few of the university and patient groups that participated.
STEVE USDIN: Today, we're assessing the tension between value and innovation in health care. We're talking with Pfizer's Gail Cawkwell, Martin Marciniak from GlaxoSmithKline, and Avalere's Dan Mendelson. Dan, you were saying that it's getting to be more difficult to make those differentiations?
DAN MENDELSON: Yeah, I think we're in a world with the Affordable Care Act, and more people have insurance, which is a really important thing. There is also more integration of systems. And, increasingly, manufacturers are going to be selling to large integrated systems and to managed care payers as opposed to focusing the efforts on physicians. And that's a shift that I do think affects the development process as well as the sales.
STEVE USDIN: And that's kind of what I was getting at with this question. Does it bias the kind of drugs that you develop? Are you developing drugs that are going to satisfy a managed care market, for example, rather than patients, necessarily?
GAIL CAWKWELL: No. And I think one way to look at it is, you said manufacture, and I'd say most pharmaceutical companies, certainly Pfizer, we're really an innovator. So we're starting right there at the beginning in saying, what are the drugs that patients are going to need, that payers are going to pay for, that doctors are going to prescribe? And gone are the days where you could just take a molecule and say, this is interesting. Could we develop it into something meaningful? Instead, we're starting with, what do we need to develop? What is going to add value?
STEVE USDIN: And do you have a formal process for looking at that?
GAIL CAWKWELL: Yeah, we have a very formal process right now that involves, as we talked about, getting input from a lot of stakeholders and identifying a few key differentiators that we call a medicines differentiation index. And those key differentiators have got to be there for us to want to develop that medicine.
STEVE USDIN: And is that the same at GSK?
MARTIN MARCINIAK: Yeah. At GSK, we have a program called Medicines Vision Program. It's important to note that even though we were talking about payers and providers, at the center of all this is the patient still. And so we're looking for the differentiated value that means something to a patient who is going through the process.
So when I talk about a meaningful endpoint, a targeted population and relevant comparators, the patient is at the center of this conversation. It starts early on in early-phase development. It continues through the 3B4 program 'till it's eventually marketed. So you get that differentiated value by bringing those voices to the table early on.
DAN MENDELSON: I would only add that we've developed a number of these programs not for these two companies but for others. And I would say that the way that we've seen them change is that, increasingly, they are more focused on the needs and the interest of commercial payers. So it is always true that the patient is going to be at the center of a development program, but you also have to anticipate the fact that you are going to have to sell this to a managed care payer who is going to be thinking differently. And that is something that, I think, over the last 10 years, we've seen more of the focus shift over towards those payers.
MARTIN MARCINIAK: I think the imperative is to generate value. You can look at different ways of creating that value. At the end of the day, there's a lot of opportunity there.
So I hear your point about the managed care environment. But the fact of the matter is there's a lot of opportunity in other therapeutic areas that might not be as visible to others, including rare diseases, oncology, and places like that without differential value. It's still very important as it relates to the patients we serve.
STEVE USDIN: I wanted to take a step back in some of the nuts and bolts of what you talk about in the report about generating the data that can be used. One of the things that you talk about is the idea of creating a unique patient identifier that you can use to track patients. What's that about? Why is that important? And how would you go about doing it?
DAN MENDELSON: There's more of a need to get data not only during the development process but also afterwards. And the unique patient ID number is really a way of tracking the patient experience after the clinical trial is over. And so we live in a health care system right now where data are not continuous.
We don't have good access to really following the patient experiences because of how fragmented our healthcare system is. And I think all of us, who are in the drug development world, are looking for more information. And patients need that information to understand what the value of medical therapies is going to be.
GAIL CAWKWELL: If I could just add to that, I think one of the exciting challenges but also big challenges is this overwhelming amount of data now with electronic health records and other systems that now capture so much. And there's this promise of your ability to learn and to learn about a drug, to learn about a treatment not just in a big population but in very specific sub-populations that may not have been part of a clinical drug development program. The problem is those systems are really not where they need to be to be useful. And the analytics, it's the patient identifiers, but it's also how do you look at that data? How do you make it into something that's meaningful, that's useful, that will really tell us more about how to use that medicine, how to treat that patient?
STEVE USDIN: We're going to continue talking about that in just a moment. But first, in addition to GSK and Pfizer, here are some of the other industry players that joined the Avalere process.
NARRATOR: Now back to BioCentury This Week.
STEVE USDIN: Is innovation making healthcare too expensive? We're talking about it today with Martin Marciniak, Dan Mendelson, and Gail Cawkwell. Martin, you wanted to pick up there about methodology.
MARTIN MARCINIAK: Sure. When you're thinking about the health infrastructure, the data infrastructure, one of the challenges particularly there is the methodology -- is the methodology keeping up? The particular issue is, can you conduct studies in real-world data? I think the answer is yes. I think the methodologies are tough to get a handle on, particularly when you're thinking around complex issues such as bias and how you control for that.
GAIL CAWKWELL: Yeah. I think when you do a clinical trial you randomize patients so patients are about even in the different treatment groups. In the real world, a doctor made a decision to prescribe one medicine or another. And those reasons could be obvious. It could be based on age or a stage of an illness. But they also could be very subtle reasons that may be very hard to pick up in an observational type data set. Now people like Martin have very sophisticated ways to analyze data to do a great job with it. But there's so much data out there now that it's also very easy for people to just look at the data without applying these rigorous methods.
STEVE USDIN: And it took decades, starting in the 1960s, for people to develop the methodology for looking at randomized controlled trials. And one of the things that they figured out is that you can't just do the trial and then start rummaging around in the data and find a statistically significant result and say, ha, we got the truth here. But that seems to be precisely what people do with observational data today.
DAN MENDELSON: Well, I think that one of the big challenges is that, as we look at more, different, and varied sources of data, there are also a lot of challenges in communicating that information to patients. And one of the things that we focus on in the report is the need for the Food and Drug Administration to take a step back and think about how the data from all these different sources can and should be communicated better to patients, and how to ease some of the restrictions on how that information is communicated by companies that are developing these medicines.
STEVE USDIN: Isn't it part of the responsibility, though, of the companies that they have to change so that the data that they generate is apples to apples, so that everything isn't unique and handcrafted and not comparable?
GAIL CAWKWELL: It's really a hard challenge. In the end, there's a lot of people who need that data, right? The regulators need it to see if the drug is safe and effective, should they approve it. Payers need it to see should they pay for it. Doctors need it to see how should they prescribe it. Should they include it in treatment guidelines? And patients just want to know, is it going to work for them? Is it going to be safe? And to try to find something that answers everybody's questions can be challenging, and also very hard to build into a development program.
In some cases, you really just have to wait until it's being used in the real world to generate the data after approval. And sometimes people feel like they wish they knew it before. But that balance between what do you do before approval and after approval is one we're looking at all the time.
MARTIN MARCINIAK: I think the real issue is in a clinical trial, you're getting efficacy, right? And so that's where the story starts. When you move it out into the population of patients who are taking it, you're now talking about effectiveness. And so you're looking at the clinical benefit in a large population of patients. And then you want to bridge that over into talking about the economic value, or the value in general, both clinical and economic, it brings to a patient, a provider, or a policy maker.
And so the challenge is, as you move across those three domains, the methodology is not necessarily always uniform. And you're really trying to focus in on the relevant story that you'd like to talk to those individuals about, the most important of which is the patient and the provider relationship. And so getting that right stands at the forefront.
DAN MENDELSON: Gail was talking earlier, also, that some drugs work in some sub-populations but not in others. And it's very common, also, when you go in to talk to federal regulators at CMS, who are trying to craft the Medicare program -- they're trying to craft the Medicare program -- they will ask specifically, have you tested this drug in my population in the Medicare program, those over 65? And it's very difficult to test many drugs in that population. And so the payers are getting more specific about asking questions that are specific to the populations that they serve.
STEVE USDIN: So the real question -- I guess what it really gets back to -- is, are we going to be able to create a learning healthcare system so that all of this information, especially what happens in the post-market, is going to be able to influence clinical decisions in real time?
GAIL CAWKWELL: I think the answer is absolutely yes. But we're not there yet.
STEVE USDIN: How do you engineer the idea of quality measures into drug development and the way that drugs are assessed so patient's views are incorporated into what's happening and what's being paid for?
GAIL CAWKWELL: Well, besides working at Pfizer, I also take care of children with arthritis and rheumatic diseases. And I think that gives you a good example -- or in the pain area in general -- about how complex things can be. If you look at the world of chronic pain, patients are not always diagnosed with pain. They may be diagnosed with an underlying health problem. So to even know what's going on with them and to then capture information about qualities and benefits can be very complicated.
So the systems -- it's very easy to talk in generalities. When you get down to the specifics, and the fact that, say, in the rheumatic diseases the difference between one child with lupus and another could be night and day -- it could seem like two totally different diseases. So how do you do that? How do you measure that? In the clinical trial world, to prove a drug works, you often need a fairly similar population. In the real world, we need to learn a lot more.
STEVE USDIN: And Martin, that brings up this whole idea of heterogeneity, and, again, this idea of, does needing to develop drugs to satisfy these metrics that payers generally want to know -- they want to know whether these drugs really provide value -- does that bias you, again, against developing drugs that are going to address heterogeneity, they're going to work for very different people with very different circumstances?
MARTIN MARCINIAK: I don't think it biases against it. I think it speaks to the underlying challenge that it's just very difficult to do this. And so you can think about different approaches to identifying a patient population, phenotypic gene or biomarker, but none of those approaches are terribly easy to do. And so we work very hard on our clinical trial programs to look for opportunities to really specify the best patient for this given therapy. And so when you loop it back together, you think a little bit about the question of heterogeneity, and then you think about quality. We play roles on quality forms as well in steering committees looking at resource and economics. So that's part of our role.
GAIL CAWKWELL: Just to add to that for one second -- I think it's, again, right patient, right medicine, right time. That adds up to quality.
DAN MENDELSON: And then you have to then translate that into the health care system in payment. And it's a challenge, but something that we have to do. Because we have to be very clear with clinicians, as well as with payers, about what is quality and how do we measure it. And sometimes, in the case of pain, for example, you ultimately really have to ask people about their pain control. It's not something that you can necessarily assume. And so some of the metrics that we need to be developing are really about communication back and forth with patients to make sure that the therapies that are being delivered are actually working.
STEVE USDIN: To learn more about what's being discussed today, download Avalere's report on delivering value in healthcare. Find it on BioCenturytv.com. We'll hear some final thoughts about healthcare innovation and value up next.
STEVE USDIN: What can be done to reduce healthcare costs and promote innovation? Let's get some final thoughts from Gail Cawkwell of Pfizer, Martin Marciniak of GSK, and Dan Mendelson of Avalere.
Gail, I want to start with you, and I'm going to ask all three of you the same question. If you could just do one thing that would help to better align the value and the promotion of innovation, what would it be?
GAIL CAWKWELL: One thing is tough, but I would pick transparency of how people are analyzing data. What are the methodologies they're using? People come out with a new piece of data, and you just don't know, where did this come from? How did they generate it? Did they use methods that make it valid data? Are the conclusions ones you can count on or not? I'd like transparency.
I'd like transparency as well, bridging off of what Gail was saying. I'd talk about the transparency of the data itself. And so one of things that GSK has focused on recently is transparency around our clinical trial data, making the data available for those assets that have not been brought to market as well as those assets that have been, making that patient anomized [sic] data available to third party researchers, to really sort of answer scientific questions. To look at the data from a different perspective, and frankly, drive innovation and value.
STEVE USDIN: And so there's some controversy around that. Some of the other pharmaceutical companies have opposed it, for example. And on the other hand, some of the patient groups have said, you're not going far enough.
MARTIN MARCINIAK: I think GSK had made a strong statement about where we feel the marketplace of ideas should go, and the transparency of data is core to who we are in terms of our strategic approach.
DAN MENDELSON: Yeah, I would just add, I agree that transparency is very important. And I think that another thing that could be done would be to bring quality-based payments into the healthcare system for providers, and then also to integrate the patient viewpoint into a lot of these discussions.
I think too often, both policymakers and ultimately the payers are not listening to patients, understanding exactly what it is that they need, and really bringing that into the discussion, as well as into the payment systems.
STEVE USDIN: Can you maybe elaborate a little bit more on what you said about quality, bringing quality payments into the system? What does that mean in terms that people can understand?
DAN MENDELSON: Well, instead of paying just when someone comes to the hospital, or comes to the doctor, we should be paying on the outcome. And that is a very different way to be thinking about things. Instead of just paying a managed care company when they sign up an enrollee, we should be making sure that they are doing the things that they need to be doing to actually improve patient health.
And those quality-based payment systems are absolutely vital to make sure that innovation is continued, because ultimately that is what should drive innovation in the healthcare system.
STEVE USDIN: Gail?
GAIL CAWKWELL: Just like to add to that part of it by saying, if you look at something like chronic pain. It's very hard for payers, or hospital systems to get their hands on it and understand, where's the value, where are they losing money, how could they treat patients better? Partly because of how things are coded. Partly because of how the systems are based.
STEVE USDIN: So it's really about generating better data that's going to allow them to do that. That's this week's show. I'd like to thank Martin Marciniak, Dan Mendelson and Gail Cawkwell. Please join us next week as we bring in the patient perspective and get down to the bottom line: paying for value.
Dan and I will be joined by former congressman Tony Coelho, the author of the Americans with Disabilities Act, and by Scott Gottlieb of the American Enterprise Institute, a former deputy commissioner of FDA. Remember to share your thoughts about today's show on Twitter. Join the conversation by using the hashtag #BioCenturyTV. I'm Steve Usdin, and thanks for watching.