Healthcare Rx Podcast - Episode 4

July 2018


Episode 4: Money-Based Guarantee: The Next Drug Pricing Model

David: So let me welcome you back if you’ve been with us since this morning. Let me welcome you who are joining us for our final panel. And we’re going to be talking about a topic that is boring, and uninteresting, and has no significance to the valuation of equities. It’s the drug pricing. It’s the prices that people are going to charge for their products and value-based contracts.

I am Dr. David Friend. I am one of the co-founders, and chief transformation officer of The BDO Center for Healthcare Excellence & Innovation along with my partners that you’ve seen before today, Patrick Pilch and Steven Shill.

We have two wonderful panelists for our last session of the day. First is Ann Kraft. Ann is the executive director of Purdue Pharma. She is responsible for business development of strategic IP and has a wealth of experience in the healthcare industry, including deal experience, life-cycle extension, intellectual property licenses, and development collaborations. Prior to entering the pharma world, Ann conducted merger and acquisition work for United Technologies, and she also designed jet engines at Pratt & Whitney aircraft. So if you don’t think this is rocket science, we’re ready.

At the same time, to my left, we have Dr. Steven Small. Steven is a professor of Neurology, Neurobiology and Behavior, and Cognitive Sciences at the University of California, Irvine. Steve also serves as the chief scientific officer of the Medical Innovation Institute at UC Irvine. Steven has been a pioneer in understanding the anatomy and physiology of the human brain and its relation to function by direct investigation of human subjects, particularly in speech and in language. So thank you both for being here.

This panel is really one of the hottest and hardest topics in the pharma industry, which is talking about this whole issue of outcomes-based drug pricing. So we’re looking forward to hearing your perspective and your outlook. So I want to read something if it comes up.

So yesterday, Alex Alazar, who is the secretary-designate for Health and Human Services, went before the Senate for his confirmation. He was asked about his priorities. His number one priority is assessing high prescription drug prices. So this is obviously an issue that’s gripping the country.

Now, he previously was at Eli Lilly, and Ron Wyden, the senator, asked him if he had ever approved an increase for drug price. This is really interesting, what he said. He said, “I don’t know that there is any price of a branded drug product that has gone down from any company on any drug in the United States ever because every incentive in the system is towards higher prices.”

Now, if you think about the computer industry, in 1945, the first computer that was built at Harvard was designed for calculations for artillery, and it could do two, basically, additions a second. This computer can do one billion additions per second. Can you imagine if the price of computer power had never fallen down, this thing would cost about $100 billion.

So you see what the computer industry has done with its technology advances, and now we can see what the pharmaceutical industry has done. So this is hugely challenging, and yet the pharma industry, everything with genomics, we are about to have an explosion of innovation. So clearly, the issue is, how are we going to pay for this?

When we are looking at genetic therapies that can be $1 million or more. I think there was one just approved around $750, 800 thousand in the UK. So this is coming.

Pharma Turns to Value-Based Contracts [3:50]

So let me start out with our panelists, and let me get to our first question. So Steven, I’ll ask you this question first, and then I’ll have Ann follow up. So with intense pressure on drug prices, the pharmaceutical industry is turning to value-based contracts, which link the price of a prescription drug to the quality of its clinical or economic performance to kind of ease tension. But the use of these new drug pricing models with insurers is very limited despite an active minority of enthusiastic participants. So from where you sit in the academic world, is this a trend or a fad?

Steven: Okay. Thanks, BDO, for inviting me, and welcome everybody. Before I answer that question, which I will do in a second, I just want to mention my background and how it’s germane to these questions. Yes, I’m a professor of Neurology. Yes, I’m the CSO of an innovation institute. But in fact, I got my PhD in artificial intelligence many years ago and then went to medical school and ended up moving. I worked at the University of Chicago for 10 years, and I built an imaging center there.

And then I moved to the University of California - Irvine, to rebuild the Neurology department and to build Neurology programs, including a comprehensive stroke center, a comprehensive cancer center, and many other programs in epilepsy, and multiple sclerosis, and other things.

So I spent seven years as the chairman of the Neurology department and have first-hand knowledge about how we handled drug pricing and how we handled care in a tertiary care medical center. And it’s on that basis that I can address these questions.

So yeah, I think it is not a fad. And I think the way to start looking at the question, actually—and I’m sure we’ll talk about that a lot during this hour—is what is value-based care? And of course, value-based care, we need to define it. And the way we define it is to relate quality indicators to cost indicators. Okay?

So no longer are we looking at outcomes as whether someone’s better or worse, whether they’re cured or not cured, whether they’re dead or alive, or whatever. We’re now looking at this as a ratio, and that’s a novel concept. But it’s a very important concept because if you look at the numerator and you look at the denominator, you have quality and you have cost. And the question is, how do you measure them? What are they? Can we measure them? And this is where—I mean, this is limiting a little bit.

You said it’s not pervasive yet, this model. Part of the reason it’s not pervasive yet is because we’re still trying to figure out the definitions. We’re still trying to figure out how we can measure these things, and we can talk more about what quality might be, and what cost might be, and how to measure them, and how we define them as the hour goes on.

But I think that’s the issue. Once we get some handles on these things, once we get measurements of quality, once we can figure out what we’re talking about, and we have quantitative, algorithmic descriptions of what we mean by quality, and we assess costs of particular kinds of things that we want to-- that we want to measure, then we’ll be able to show that this is not a fad, but that this is a very important concept. I think my hospital does not know how much it costs to take care of a stroke patient.

David: So Alazar also said yesterday that one of the big problems if you’re trying to figure this out is we are not sharing information between the doctors, the drug companies, and the patients, for example.

So he said, for example, “If we really want to figure out outcomes, we’re going to need all of these different organizations to come together.” So let me, Ann, ask you kind of the same question. What’s your sense? Fad? Is this going to be the way? I mean, investors are investing tons of money on trying to anticipate how we’re going to pay for stuff and how they’re going to get paid for their investments. Well, what do you think?

Ann: Okay. First, thank you, David. Thank you, BDO, for inviting me. And welcome, everyone. Would it be unfair if I went back to your opening comment? Because I wanted to comment on whether—

David: Feel free. Feel free.

Ann:—pharmaceutical prices have dropped over time, and you compared it to computing powers. You held up your phone. The mechanism for dropping pharma prices is really with loss of patent exclusivity and the launch of generics.

So when they say their drug prices have not dropped, if you have an exclusive brand, then it is at a brand pricing. But when generics launch, that really changes the price dramatically. And that brings a much lower price and much more access to the consumer, and the price of the brand actually drops. When the generic comes, the price of the brand also drops.

David: Right. And in fairness, the average length that people actually control the brand is fairly short-lived. Right? It’s like 17 years, and that includes all the innovation and marketing time.

Ann: So you generally get 20 years of patent life, but that includes your development time. And so when the development time can be 10 to 15 years, I’d say companies are lucky if they get five years.

So they really have a very narrow window.

They sometimes have three years. It can be as short as three years. Five years is good. It’s not really longer than that very often.

David: You gave, really, a better answer than he did, that he really should have added that to his answer. But anyway.

Ann: So let’s go back to your question about value-based pricing and the models that we’re seeing, and I think I agree with what Steven said 100 percent. We’re coming along, and we need intelligence. We need data. We need high-tech tools to help us value these.

And I think it’s—this is not a fad. It is definitely a trend. I think pharma has been painted with a brush that is a little bit unfair. Pharma really is here for the patient. We really do care about the patient. We view this as something positive. We want to participate in this. We want to bring value to the patient. We really do, and the days of having an indication and it covering broadly everyone whether it works or not, that is going away. And that’s a good thing from everyone’s perspective.

We haven’t yet figured out how to use this model and use it well. In some indications, it works well. So you see it occurring in oncology. Oncology’s often a leader, but you have metrics there. Has there been tumor shrinkage? Has there been extension of life? It’ll work well in disease therapies where the treatments are very measurable as to their success. So that could apply to diabetes, where you could measure blood levels. It would apply to cholesterol levels, products like that.

My company, Purdue Pharma, we work in pain products. We do have a lot of opioids in our portfolio. Opioids work. It is not really about whether they work, in that case. It’s about managing the risk. And this model of, well, is your patient receiving pain relief? That is just not going to work here. And what we’ll get into as we get on in the hour is, how do we find something where you can really measure the success of the drug, the value of the drug, so that it brings value to the patient without being really cumbersome? Because we do work with the payers, and we are working with them.

Again, in our case, it’s a lot about risk management. And we do work with them and creatively offer different models for how to measure this. But often, they say it would take so much energy and effort to track all these patients, and measure all of these things, and add so much cost in the system. No, that isn’t going to work for us.

So we really need to get down to how it’s going to work. But ultimately, I think we’re all aligned, and I think this is going to be the way with the future, and companies that don’t want to participate aren’t going to be here very long.

David: No, I couldn’t agree more. And just to give you all context, so what’s interesting is I was one of the—I was an analyst 31 years ago in pharma and life sciences, and I was the low-level person who helped organize this conference in 1987. There were 700 people in total in this conference, in this entire conference, and we had it all done at the Hyatt Embarcadero.

We had 100 companies—I know because I had to write a brief description of all 100 - and we had 700 people. I heard this morning that 140 people from China alone just came from one of the law firms. There are now 70,000 people. It’s 100 times bigger, and there are probably 100 times as many companies. Think, 30 years from now, it would be seven million people here. You get the sense.

The innovation that’s been driven the last 30 years is astounding. People are living today who would’ve passed away as a physician 30 years ago. So it’s without question that we as a society have benefited greatly from creating incentives to get this stuff out. Here’s a fact, that the old Soviet Union had a lot of very smart people. They developed nuclear technology. The number of new drugs they developed during their entire 70 years was zero. So if we don’t get these incentives right, we risk really harming one of the most powerful engines we have in the United States. So we appreciate the comments.

Measuring Outcomes [12:49]

Let me ask you the next question, then. If we think about these measures we’ve talked about, how would we measure outcome? Because I think there’s going to be lots of issues. I was at a meeting where a hospital surgeon was saying, “Well, we’re getting really good scores now. Our patients are doing better than average.”

But then he’d made the bad comment, saying, “But we’re self-selecting our patients.” So I said, “If you do heart surgery on people who don’t need it, you get really good outcomes.” Right? Because they weren’t sick. And then you’re going to not take care of the people who are really sick because it might hurt your outcomes. So let’s talk about, what do you see? I guess, here’s the question. And Anne, I’ll ask you first. So what about these surrogate measures? What is the FDA going to look at in terms of provider, consumer, this kind of holistic idea—any sense about what should they be looking at?

Ann: Do you mean for value-based pricing?

David: To figure out what’s worth it, value pricing. In fact, you talked about pain. What is the endpoint? Any idea of how we decide this game?

Ann: I think what we see happening is indication shrinking. So drugs are, when they get approved, they are becoming for a more and more specific, specified patient. That will make the outcome measures a little bit easier because, to your point, you can’t just take the very likely side of the population. Then the indication will only be for that side of the population. So that’s part of what’s happening here.

And then I think this is the $1 million question, right? This is what we need to really develop. Where are the outcomes going? So it’s easy in some indication. In others, we’re still really trying to figure that out.

David: What should we be measuring? I mean, Bob Galvin is at Blackstone now. He was at GE. He has this great cartoon. It shows a graveyard, and he goes, “Where great healthcare ideas go to die.” He said, “If you have more than three metrics, it’s too hard. We can’t deal with it.” When he sees 50 metrics, 100 metrics, he throws up his hands and says, “Impossible for us to deal with.” As a scientist, what do you think really makes a difference, really, that we should be focusing on in terms of what should we be measuring?

Steven: I don’t agree with that assessment, for one thing. I’m an artificial intelligence guy. So just like in the previous panels, if any of you heard the previous panels, we talked about enormous amounts of data and what do you do when you have enormous amounts of data. And what you do is you reduce the amount of data.

And we definitely need quality indicators. We need indicators of quality that are reliable, and that persist over time, and can be tracked, have some relationships so that if you measure it at one point in time, then when you measure it at the next point in time, it’s related to the previous point of time. So you need these kinds of outcome measures.

Now, what are they? In diabetes, can you look at the hemoglobin A1c, for example? That’s an indicator of how good the control is of the diabetes. But then, suppose someone develops a problem with their eye or a problem with their kidney with the diabetes. How do you incorporate that information? And how do you use information when you don’t have a clear indicator of what’s going on? If you’re interested in someone’s quality of life, do you measure tumor size? Or is that an appropriate measure?

So what we have is, we have a whole bunch of possible measures of quality, and we have a whole bunch of—and some of those measures that we have are indirectly related to what’s really going on in the person.
For instance, we have an MRI scan that shows that someone has a bit of tumor shrinkage in a brain tumor, for example. How does that relate to whether their tumor is actually regressing, getting better or not getting better? Is the drug working or not working? Is reducing the size of the tumor—is that enough of evidence to say that, in fact, you’re having an effect of your therapy?

So we need a bunch of these measures, and we need to validate these measures. We need to know that they are good measures. We need to have evidence to support the fact that they’re good measures. And what is the nature of the evidence to support that these are good measures? And once we get these measures, if we have many of them, if we want to have—so we have some single ones.

But then the example you raised is, suppose that we have a number of measures. Suppose that we have 10, or 15, or 20 measures. Can we use machine learning? Can we use some sort of algorithm to reduce that down? And so these are the questions that we’re going to have to ask the next number of years and so that we get for each disease.

And I agree with Ann that, as we get more into precision medicine, then we’re going to have medicines that are much more narrowly prescribed to specific kinds of individuals based on genomics, as we heard in the last panel, or other kinds of omics. And then, in that case, perhaps there will be a reduction in the number of measures that are relevant.

Ann: David, could I answer that just before we move on? We have to be really careful with the measures. We want to make sure that we get the right measures because they sometimes have unintended consequences.
So for instance, one thing we see is some hospitals are measured based on how well did you control a patient’s pain. So at the end of a hospital stay, there is a survey that says, “Well, how well was your pain managed?” And you want the patient to give you a four, or a five, or whatever the scale is and say, “Yes, it was managed very well.”

So if you’re the physician taking care of that patient, you want to make sure you get a good result, you’re going to prescribe pain meds. And you’re going to give them access, maybe, to the PCA pump where they are giving themselves morphine because you want a good score. And that is not the intention—the intention of a score was not to lead to overprescribing of pain meds, and that’s part of the crisis that we see. But the metric that was put in place was to measure how well a hospital is taking care of the patient, and it had this unintended consequence.

Steven: The quality indicator. So in the neurology service, one of the quality indicators we have are these surveys. And the surveys include how was your pain managed, how nice was the nurse to you, and all sorts of measures like that. I don’t find that to be a very good quality measure, but that is a quality measure we have used, and you probably know about this.

And so we were wondering at one point, why is it that our quality measures had gone down over a period of a year or two? And this correlated with the opiate crisis, which is very big in Orange County, California. And we did a Lean Six Sigma evaluation of what was the cause of our reduction in quality score. Quality where this particular quality measure is the one we were using, which I agree with Ann, is not a good quality measure.

And we found out it was because we were trying to take care of our patients in a much better way with respect to their dependence on opioids. And so many people who came into the hospital who were dependent on opiates were very unhappy with us. And so that unhappiness increased over a period of a year or two concomitant with our attention and society’s attention to the opioid crisis.

So here we have a “quality measure” that I think is not a—I agree with Anne, not a very good quality measure. So if we were basing our reimbursements for pain medicines or for something else, if this was tied to that particular quality measure, we would be going in absolutely the opposite direction than we want to be going.

David: The wrong direction. Yeah. No, it’s interesting because I ran one of the largest hospices in the United States both as the chief medical officer and the president. And we had four real criteria for our patients, which was medication management, symptom management, palliative care, but the most important was pain management. And the biggest problem I saw was we were not giving people enough pain medication.

And I have not had cancer, but I’ve talked to many people who did, and what I learned from them is that we do not understand how excruciating their pain can be. And we were really under-medicating people. But at the same time, we’ve had this political firestorm that we’re over. And so the docs and a lot of people have felt in a very tight vice, and I’ve been on Capitol Hill in Washington, and you were really in a conflicting place.

And the problem, as you mentioned—it was what was really the measure we should look at, and what really mattered? And again, we’re trying to alleviate suffering. We’re trying to be helpful to people.

So I think to both your points, it is really critical when the policymakers establish policy. To your points, the unintended consequence because the investors are going to kind of—the money’s going to flow where the system says, “Here’s who wins the game.” We’ve got to make sure those goal posts are the right place because, otherwise, we’re going to go ahead somewhere because the game tells us to go there, but it’s not actually helping the people that we want to help, and it’s problematic.

Steven: Let me give you an example of surrogate measures and quality measures in research. So if you look at drugs on the market in neuroscience over the last decade or so, you see that there’s been enormous success in development of drugs for multiple sclerosis, and there’s been a relative failure of development of drugs in Alzheimer’s disease. And I know there are a lot of companies pursuing Alzheimer’s disease and trying very hard to develop drugs for Alzheimer’s disease. It’s one of the things all of us in this room are scared of the most. Correlates with how old we are, right? But we’re scared of it.

David: Some of us are closer than others.

Steven: So what’s the story with these two different paths? And one of the stories that people have identified is that we develop good surrogate measures of outcome in multiple sclerosis trials, and we have not in Alzheimer’s trials.

So it turns out in multiple sclerosis, if you look at an MRI scan of the brain where the person gets a little bit of contrast, you can see how active the disease is by looking at that MRI scan. And that revolutionized the development of trials because, all of a sudden, you didn’t have to ask people how they’re doing, or do exams on these people, or do these other kinds of measures, which were not very good outcome measures because they were variable, they were unreliable, and you could not—and they did not—from time to time, you could not see linear change of the type you wanted.

But with a good surrogate outcome measure, all of a sudden, it was possible to do design trials in this area. In Alzheimer’s disease, we don’t have any surrogate measures. So we have to look at how is the behavior doing, and the behavior can change from day-to-day to week-to-week, and it’s very, very [crosstalk].

Aligning Different Stakeholders’ Priorities [23:33]

David: So let me change the topic, and I’m going to throw this at you because you’re in the business you’re in. So you have so many stakeholders you have to make happy. Investors, payers, FDA, doctors, patients. How do you deal with this dynamic?

Ann: There are a lot of stakeholders. Each one is important, and we do find that their interests sometimes conflict. Right? So sometimes they align, sometimes they conflict. Patients want treatments. They want innovative therapies. They want to treat their diseases and have the best outcomes possible.

Payers, they want what’s good for the patient, but they want it at a value-based price, and that can sometimes conflict.

The FDA, while they want treatments for patients, they also want to manage the safety of the whole population as a group. Sometimes that leads to therapies being removed that work for some populations but not others.
And the FDA is not actually responsible for price, but we hear more and more through Scott Gottlieb. He is bringing that in. He is clearly saying he wants to get innovative therapies through, but he wants to maintain them at—he wants to address pricing. He does.

David: Well, he said high drug prices were, at some level, a national emergency. It was actually said, that.

Ann: Not an FDA mandate, but he is going to bring that into his realm.

David: And as you know, Scott was one of our senior fellows. Patrick and I worked very closely with him for several years.

Ann: I think what we do is we work with each and every one of them, and that’s what we have to do. And you have to listen. You have to listen, listen, listen. We have to understand where the FDA is coming from, the patients, the payers, and we have to try to address their concerns.

Everybody is not going to get everything they want, but certainly, everybody should get one of the most critical pieces that they want. And ultimately, we have to keep the patient at the center of it. This is really about the patient and bringing therapies to the patient.

Using Clinical Data to Determine Outcomes [25:25]

David: Let me ask you a data question, given your analytics and AI background. So real-time clinical data. Let’s talk about this. Real-time clinical data is increasingly critical to determine the outcomes in value. What do you think are the top challenges going forward when it’s going to be with collecting and using clinical data?

Steven: Yeah. I mean, the biggest thing right now, I think, is we don’t know from the medical care standpoint how to make all the data we’re collecting actionable.

So I tell the example of my wife, who bought a Fitbit a couple of years ago, and so she’s got thousands of data points on her heart rate every hour or something—or, every day. And she goes to the doctor, and the doctor puts his two fingers on the radial artery and says, “What’s the pulse?” Okay? So we have enormous amounts of data.

Now, that’s a very simple example from the ambulatory setting of real-time clinical data. But in the hospital, on my intensive care unit that I developed, we developed a neurologic/neurosurgical intensive care unit. We’re measuring electroencephalography all the time. We’re measuring electrocardiography all the time. We’re doing MRI scans and CT scans every day, or two days, or three days.

These are petabytes worth of data, and we don’t know how to use these data other than the way we’ve always used these data, the way you did when you were practicing, or when I was practicing. We look at an MRI scan and say, “Oh, okay. It looks like it’s getting better,” or we look at the EEG—or, the EKG, and we say, “Okay. It looks like the heart rate is okay,” and then these data go into permanent freeze.

And so we need to—and I personally believe because of my background, in part, that we need to take advantage of these data by applying artificial intelligence algorithms to them just to go back from this complicated data set in the hospital to the more simple data set in the ambulatory setting. If I’m a doctor in the ambulatory setting, I’m not going to look at a thousand points from my wife’s Fitbit. I just can’t do that.

But if I have an artificial intelligence algorithm that can take each thousand data points and say, “There are some regularities here in how the pulse is going,” and extrapolate from that, then we have enormous data reduction. We can actually use those data, and we need to do that with the real-time clinical data as well. And we need it to come up with outcome measures based on that.

David: So Ann, I’m going to ask you a “thunk.” Do you guys want to put the slide on—can you put the slide up?

Could I comment on that prior question as well?

David: Please feel free. Feel free. Go ahead.

I think there’s an issue as well. I love the kind of metrics that you’re talking about, getting data and doing it electronically, like through Fitbit. That’s wonderful data, and we can get lots of that inexpensively.

But I think as we move to these value-based outcomes, we are going to have a problem with measuring patient compliance and the cost of putting these metrics in place. So we know that patient compliance tends to fall off, particularly, when the patient is doing well. “I’m feeling really well.” Nobody likes to take their meds, so they don’t take their meds until the point that they’re no longer feeling well.

So if we’re going to measure, if these things working or not working, part of it has to be how well is the patient complying. So then you have to put something in place. And really, it could be a phone call, but I really don’t think a patient wants a phone call twice a day saying, “Are you taking your meds because it’s this time?” So they would see that as disruptive in their lives, and they don’t want it, and there has to be a whole infrastructure in place for how we make sure that there is compliance. But while we’re doing this, we don’t want to drive up the cost.

In order to measure and to be doing additional screening, additional blood tests, additional doctor’s visits, perhaps, additional phone calls, there’s all a cost with that. So while we’re trying to remove cost from the system, we’re adding cost to the system. So how we get to the patient compliance and how we get the data that we need without driving up the cost, I think, are going to be issues going forward.

Steven: One thing that really excites me a lot about value-based drug pricing is that it takes the pharma company and gives them an incentive to work toward compliance.

So whatever you end up with your quality metric, whatever you end up with your outcome, your number, if the drug pricing is tied to the outcome of the patient, then the company that has negotiated that agreement for value-based drug price, they all of a sudden have an interest in the compliance of the patient. So you’ve incentivized the drug company, the provider of these medicines, to actually get involved in compliance.

And as you know, there’s interesting technology available for compliance. I mean, it was in the—I saw the other day—I’m sure many of you saw this, where there’s something you can add to a capsule pill and it will send a Bluetooth signal when you’ve taken the pill to your iPhone. So there are ways to do this, but I like the incentivization of the provider.

David: But we work with the company Exact Sciences, and they have Cologuard, which you may have seen advertised on television. What’s interesting, which I didn’t realize, only about 25 percent of the people who get a prescription to have a colonoscopy actually get it. So the compliance is quite low, and if you don’t have it, your cancer will not be detected.

So Cologuard has a much higher level of compliance. A much higher percentage of people who get the prescription take it. Then there are all these debates as which was more sensitive, which was more specific. But what was undeniable, if you didn’t take the test, it was never going to pick up the problem. So this idea of involving the companies to get there is important. So will you put up the slide?

And I’m going to ask Ann this question in terms of innovative partnerships. So innovative partnerships are a hot trend in the healthcare industry. Think about CVS and Aetna. What are you seeing from the drug manufacturer’s point of view in terms of, are you going to be partnering with other stakeholders to track outcomes as we described, to demonstrate efficacy, trying to figure out pricing pressure, managing cost? But what do you think the future’s going to be for folks such as yourselves in regard to those issues?

Ann: I mean, I think we are going to be partnering. We clearly are. We’re partnering with all stakeholders. We’re doing that now.

So for instance, in our area, the FDA is trying to do a public/private partnership where they’re bringing together—it’s being run by Governor Chris Christie—and they’re bringing together industry partners and government partners, and they’re saying, “What do we do to address the opioid crisis? How do we bring new abuse-deterrent formulations together? How do we bring new treatments? How do we get as much of the treatments out into the public as we can? How do we prevent this going forward?”

And they’re asking industry to put in resources that we have, and they’re willing to use government money for it. We are participating in this, as are many pharma companies. But I think initiatives such as this, as we move forward, are important. And I don’t think—going back to who’s going to do well, and I know you briefly mentioned investors.

Companies that are going to do well, that are going to succeed, are going to work with other companies. They’re going to embrace this. They’re going to work on value-added pricing, and it’s not going to be one party. So I also envision that pharma companies, and we are starting to see this, will work more together to spread the risk, to share the data, to bring us to that next level.

The Role of Government in Value-Based Pricing [33:02]

David: So let me ask you a follow-up. And we talked about investors, people who pay the money, but the biggest payer of all is the government.

What should they be doing differently or better to—I mean, are they the problem? Are they a solution? We know they have tremendous power and influence. I have always said that CMS is the chassis that all of healthcare is bolted onto. If you were the king, the queen, what’s your sense? What should the role of government be in this in terms of using its dollars, do you think?

Ann: In terms of investing?

David: If you think of them as an investor, or as a purchaser. How can they use their economic clout to just make things better?

Ann: So I think the government should be partnering with industry. They should be investing in industry, which the NIH does to some extent. But I think there are a lot of therapies that need help, need funding. I think that’s the first thing. I think they need to rethink their payer model. They have unintended consequences with their payer model as well.

David: Can you give me an example?

So one of the examples is the government gets the lowest price. Whatever price a pharma company contracts with any of their payers, the lowest price is given to the government. So there could be instances where an organization wants to, out of compassion or whatever the initiative, give a lower price to a particular group.

But if that were to happen, that price then applies automatically to the government, who is the biggest client, the biggest patient. So then you can take a huge percentage of your revenue, and it drops down to a much lower price, which is going to perversely incentivize that you will not offer it to that prior group. So I think there needs to be a rethinking of the whole payer model there.

David: What do you think?

Yeah, I have a different take on it, I think, but complementary. I mean, we’re here talking about value-based approaches, and right now, most of what the government’s doing is volume-based, not value-based. And so the government could, in fact, be a leader at value-based drug pricing.

When you’re running a health system or when you’re a doctor—me managing an inpatient-outpatient program, I have to be very wary of costs. I have to be very—I have to pay attention to costs. I can’t just ignore them anymore. It’s not the era where we can just say, “Any drug for any patient. Where it’s a reasonable choice is the best choice,” because there are lots of alternatives.

I think when we were talking the other day about examples, I can give an example of this which I think is quite relevant. I don’t know how many of you are physicians, but those of you who are physicians or even consumers of medical care, you know there’s lots of different choices for medicines to treat more or less the same condition.

And for a health system, if there are many drugs and some are much more expensive than others, we need to select the ones that are more economical for us if the evidence is the same for all of them. We can talk about evidence in a few minutes because developing good evidence is not so easy. It’s not straightforward. It’s not transparent.

But assuming that the evidence is the same for multiple different choices of therapeutics, healthcare systems now will choose the least expensive approach increasingly.

And so we had a problem. Again, we used this Lean Six Sigma approach to analyze problems in our clinical services, and we had an issue where we were using—I’ll give you exact names, and I’ll name the names and the dates. Okay?

We were using Versed to do sedation on our intensive care unit. As you know, many people in neurologic intensive care units need sedation, and we were using extraordinarily high doses of Versed to sedate patients. There was no data supporting the fact that this was perfectly good. It was efficacious, it was valuable, and it was good, but it was about a thousand times more expensive than using another drug. Okay? And another drug which had equal evidence and equal efficacy.

And so what we did, we did a Lean Six Sigma as a clinical program, and we said, “What are we doing here? Should we continue to allow this one practitioner, particularly?” But then students were learning from this one. We’re an academic health center. Students are learning from this one practitioner, and so we ended up with a cadre of people who wanted to use this particular approach.

And through this Lean Six Sigma approach, we reduced cost enormously on our neurocritical care unit simply with that one change. But we’ve done a lot of changes to reduce costs on our neurocritical care unit.

Cutting Costs [38:06]

David: No, and it’s interesting. Atul Gawande wrote about the fact that—he was basically doing abdominal surgery, and at the hospital, The Brigham, where he was in, there with 15 different docs. They just looked at the use of mesh, and the variation was thousands of percent. And he said, “Just bringing together the 10 doctors, who all work with each other every day, to agree on how they should do it,” he said, “that was a colossal amount of energy.” I mean, these are not easy things to do.

Steven: We did that in our operating rooms as well. We went through, and we said, “Why do each one of you 10 need 10 different types of devices in your surgical procedure—for the same surgical procedure?”

David: Exactly. And the challenge that the academic medical centers have, and I remember I was at Penn in the early ‘80s, and we never had these discussions.

But now part of the problem is, we the academic centers, and I’ll put myself in that having been on the board at Yukon, we were the high-cost producer in a world that increasingly wants low cost. So we had to justify. Do we provide quality? How do we make it happen?

I’m going to do a time check. It is about a quarter of. Why don’t I at least ask the audience if there are a couple of questions, and then we can come back more. Go ahead.

Audience member: [inaudible] medicine [inaudible], and so usually they do [inaudible] trails. They [inaudible]. How do you decide, if you’re choosing to save the life, I will buy [inaudible]? How do you compare it?

Rigor and Reproducibility [39:15]

Steven: Yeah. So I wanted to actually bring that up, so I’m really glad that you fed into my desire. So I’ve been a researcher my whole life as well as, more recently, a health care administrator. And there is a—you probably know this, many of you, but there’s a huge push now at the National Institutes of Health at the national level and internationally in scientific communities on what they call rigor and reproducibility.

And this rigor and reproducibility movement, for those of you who don’t know what this is, this is a movement to make sure that the evidence we’re getting is reliable. And it turns out that there are probably tens or hundreds of thousands of journals that are publishing papers all the time. And if you look at the results that are coming out these papers, many of them are not replicable.

Right. It’s a big problem.

So it’s a huge problem. In fact, this was brought to my attention when I read that Glaxo tried to replicate, I think, 60 cancer preclinical trails. Preclinical means before it gets to human study. And tried to replicate 60 of them. Was able to completely replicate three of those 60 at the same level of certainty and replicate maybe half a dozen others at a lower level of certainty and was not able to replicate the others.

And that’s because we’re using statistical approaches that are not rigorous enough. There is pressure, and I take a mea culpa as an academic physician. There is pressure in academia to publish positive results. It’s not possible to publish negative results. It is increasingly so.

My journal now accepts registered reports. Those are studies where everything is done except the experiment. Okay? You write the introduction, you write the rationales, you write the hypotheses, you write the methods, and then you send it out for review. And then if that is accepted at that point, then you do the study, and it’s already accepted for publication. So it could be positive or it could be negative. Okay?

So we need rigor and reproducibility. We’re increasing our efforts nationally as a scientific leader. We are increasing our efforts. I’m on the board of various societies, and we are trying and editing journals. We are trying to increase our rigor and reproducibility so that we don’t run into that problem. And this is the evidence. The quality indicators we need, need to be reliable. Sorry, David. Go ahead.

David: And there’s a somewhat darker side, which is that, as an audit firm, someone could give you their financials, and if they say, “Well, just trust me,” we tend to say, “No. We’re going to audit those numbers.”

Increasingly, I think clinical data is going have to be audited because if the clinical data is part of the reason you get paid, the temptation—either you’re sloppy, or to fudge, or you just flat-out lie—is going to grow. And there have been the scandals from the VA on down where people deliberately misrepresented data, clinical data, because payment was dependent. Now this is very controversial. And again, I think there’s good intention. There’s lots of issues here.

Steven: Yeah. No, I take the upside of that. I mean, I think a lot of it is simply lack of statistical rigor.

David: But there have been some very high-profile examples of big companies in the Bay Area.

Steven: There’s fraud.

David: And the problem is that if we don’t regulate it, the honest player who doesn’t have the best data can sometimes be crowded out by the person who actually makes something up, and they go, “Well, that’s the person that you want, not the—“ You understand what I’m saying, but—

Steven: So concomitant with the rigor and reproducibility movement is a movement for data sharing. Data sharing, as you guys know in healthcare, if you have two different health systems that want to share information or I want to share information with the pharma company, whatever, data sharing is complicated to do no matter who’s doing the sharing. But we need data sharing among all these investigators as well so that it can be audited.

David: Yeah. And one theory, this blockchain is a way to kind of do that. But let’s move on. That was a great question, Anita. Another question. Sir, go ahead.

Audience member: I just have a question. Given that medication adherence costs $683 billion to pharma, why aren’t they building services that leverage software to increase their margins by, say, 80 percent for a typical software requirement? Why are they not motivated to do that?

David: Ann, do you want to take that?

Ann: I’m not really clear on the question. Can you elaborate? You want pharma companies to use software—?

Audience member: So what I’m saying is they should build services around the pill that allow them to change the total cost of care and improve the medication adherence. And then they could have total win for everyone. Right? So the insurer benefits because it’s a lower cost, the health system benefits because the medications are better, and they provide a valuable service that people are willing to pay for. Once we preserve their margins beyond the time or the life of their packing, even if they’re moving a generic model.

Ann: And do you mean software like on your smartphone? Software that tracks?

Because we have lots of software apps that will do that. Like, “How is your pill working for you? Have you taken it today? How are you feeling today?” which generally aren’t all that helpful.

Audience member: I’m not just talking about software in isolation. I’m talking about a total service that impacts the total [crosstalk]

Steven: You’re talking about a sort of an ecosystem around the drug. And my argument is that that will be incentivized by changes in how we compensate. And in particular, value-based drug pricing will, in fact, incentivize companies to do that kind of thing. What’s the incentive now? You think that there’ll be more margin if they do that. I don’t know. I’m not sure that’s clear.

David: Let’s take another question. Any of the other questions from the group? Go ahead, ma’am.

Pricing Expensive Therapies [45:08]

Audience member: Yeah. I just wanted to know the passports—how governments and pharma companies should go about [inaudible] gene products and cell therapy products.

The reason why I’m asking is I’m European, and in Europe, we approved, in 2012, Livera. And that was a very [inaudible] product and everyone was cheering big-time when it got approved. And the European Medications Agency gave a positive opinion, European Commission, authorization.

For them, basically, the reason those [inaudible] was in practice of the National Public Health Authorities. And what happened is that this was a product that was—the price was foreseen at the level of €1 million, so very expensive. But it was supposed to be a one-off treatment all down. So in terms of value-based, that would have been a great value.

The problem was that when it got approved, it was conditional approval because the long-term effects are uncertain. So you have an issue with the uncertainty of the evidence. So with gene therapies and cell therapies now getting momentum in time and coming to the market, I see that there’s a big issue and a problem to get patients access to these very [inaudible] products.

Ann: This is an issue when you have these very high-priced models, but I think for the most part in the US, there has been access to them.

What I would actually like to see happen and going back to your question of what the government can do, I’d like to see the US government working closer with the European agency. So often, the European trials are not accepted as the US trials, and there has to be a duplication of effort here.

I’d really like to see the EMEA and the FDA get together and have one plan where a company can develop a product and then it is approved in both territories. I think this would help bring down the cost. I think this would help bring down the price.

I think there is an issue. The US is wrestling with this issue, as is Europe. We saw this during the election. When we go to more of a universal healthcare system, as we are moving in that direction, there has to be some expectation of, well, then there’s a budget attached to it. So what happens to those $1 million therapies and one patient? And there may be some of that falling onto moreso the consumer.

David: I mean, we’re talking about having drug mortgages for people. Someone mentioned, I think, the duration. How long does a member sign up for insurance companies? The cell phone, I had to take the three-year plan, the five-year plan. Perhaps we’re going to have to have folks sign-up like that. If you’re using Medicare, you kind of sign up for the duration of your life, but under 65, there’s lots of churn issues. Do you have any comments, Steve, [inaudible]?

Steven: Yeah. I don’t know if you were here for the last panel, but the CEO had a microbiome therapy. It was a very complex therapy, and she had a money back guarantee. So if the procedure, the intervention, did not work, then the patient could have that procedure for free as a re-do, sort of, of the procedure. And so these are all parts of these value-based models.

And so maybe for some of these drugs, if they’re extremely expensive and they don’t work in the first trial, you’re automatically enrolled in the second trial. Or if they have long-term side effects, then they have to be compensated.

We want our patients to get involved in these decisions, and in some of these value-based drug pricing models, patients don’t have a big role. But we could get patients much more involved in this value-based decision making so that, if they got better, not only would the company that made the drug and the insurer benefit, but the patient could also benefit.

David: So let me do this. We just have a couple of minutes left, so I’m going to ask each of you to think for a second if there’s any kind of closing thought you want to give. And it’s been a great discussion. There’s been a lot—if there’s something you want to re-emphasize or that we happened to miss but let me give each of you maybe a minute or two just to give me a closing thought or something you’d like the audience to kind of walk away with.

Steven: I think we need to keep the patient in mind. I think we need to keep in focus the fact that the whole goal of this enterprise is to improve the health of our individual patients and the health of our population more broadly. And we need to come up—I think we’ve been using the word disruptive all day today. We need disruptive ideas in order to make sure that we can improve the overall health of our population. And apropos of this particular panel and the day, we need to do that. At least in the United States, we need to do that at a better cost level.

David: Thanks Steve. Ann?

I’m going to go back to something that’s similar to Steve’s opening comments, actually. The panel, we really were here to talk about value-based pricing right now. I think we are moving in that direction. I think there’s a lot of data, a lot of intelligence and technology that is coming along. I think as it comes along, it will help us move to personalized medicine, it will help us move to these outcome measures.

I think we envisioned this 20 years ago. It is moving far slower than we expected it to move. Actually, I’m really frustrated with the lack of the iPhone as having really anything meaningful to do with our healthcare. It still cannot measure your temperature, it cannot measure many of your body functions or how your health is, but as we get there and as we get the data, I think we will see more and more that it will help us move, not just to the value-based, but to the personalized medicine and a much more effective healthcare system, which I think will bring down the cost of the whole system. So I’m looking forward to that, and I’m optimistic that it will help us.

David: Okay. Well, first of all, I want to thank both of you for a wonderful panel. Please give a round of applause. You’ve been a wonderful audience.

If you enjoyed the show, please visit iTunes to subscribe, rate, and leave a review. Join us next time for another edition of Healthcare Rx.