How did you get [interested in] the downside in prescription medications? ...
My attitudes about medications really changed in the 1980s, when I was part of a group of people who had begun to apply modern methods of drug testing -- parallel control trials, new testing methods were becoming available for medications -- and we wanted to find out if we could prevent major complications of heart disease.
We knew that if you had an irregular heart rhythm that you would be at very high risk for sudden death. [In] the pharmaceutical industry, there were dozens of companies trying to develop new medicines, most of them based on lidocaine or Novocaine, which had been found to improve heart rhythm in a patient with a heart attack. So they were trying to develop oral lidocaines. They were changing the molecules slightly to get rid of some of the ... side effects, and finding more and more potent medications that had far more effect on the heart than any medicines had ever [had] that had been developed.
All the science was saying that this is good, we're on the right track, we're going to prevent sudden death. But some wise people said, "Well, we should make sure. We should be positive that these medicines that should be helping are helping. Just by improving the electrocardiogram, maybe that's not enough. Maybe it isn't reducing death. So let's do a study to see if these medicines that make the cardiogram look good actually prevent death."
And we were shocked. The study was stopped after a few months, because once we began, we found that even though the heart rhythm looked better, there were more people dying than had been expected, and more people dying than in the group treated with placebo.
So it taught everybody in medicine that what we called then surrogate endpoints or biomarkers -- this cardiogram that we had tried to improve -- really didn't show true improvement. It began an understanding that you can't rely on these things like just blood pressure and heart rate and cardiogram, because those are surrogate endpoints; they're not the real endpoint you want. If you're developing a drug, you have to look at the endpoint you really want, like death and heart attack, to see if it's reduced.
[There was a] drug that doctors got very excited about. ...
The medical community was looking for an oral lidocaine. ... In the 1980s, the experience in the coronary care unit showed that lidocaine given by vein could get rid of irregular heartbeats and prevent your heart from fibrillating -- that's where your heart just stopped pumping. That experience led physicians to say, "We need a pill to be able to treat patients which come to see us who may or may not have had a heart attack. But they've got an irregular heart rhythm, they can feel the palpitations, they can feel their heart skip. We want to get rid of that, because we think it puts them at risk for sudden death."
A lot of the companies were developing these drugs. Patients were prescribed medicines to make their heart rhythm regular, hoping it would prevent death.
But when the cardiac arrhythmia suppression trial [CAST] was done, we found that even though the cardiogram was looking better and the irregular heartbeats were gone, there were actually twice as many deaths in the people taking encainide or flecainide, another medication known as Tambocor, than the people taking placebo. So Tambocor stayed on the market with major restrictions, encainide was removed, and it stopped the development of a whole class of drugs, because we recognized a toxicity that would never have been found if the NIH hadn't done a prospective, randomized, controlled trial. …
What was the moral of the story there that you all learned?
We learned in this trial that you can't treat the symptoms. You can't even treat the electrocardiogram and hope that you've reduced death in this case. You have to have a study that's large enough and that's independent, because there were a lot of studies being performed by the pharmaceutical industry that indicated a possible benefit, possible reduction in death rates. But until a large enough trial was designed and independently performed by people who weren't doing the study just for the pharmaceutical industry, only then did we really find the harm.
I think that's something that's happened in more recent studies also.
A similar example would be hormone replacement therapy, no?
What sort of assumptions did we make, and what did we find out in the case of hormone replacement?
... With hormone replacement therapy, we were led to believe that estrogens could have a beneficial effect on heart disease. There were some association studies that showed that women had more heart attacks after menopause, when their estrogens were falling. So we assumed that that meant you should give estrogen and you'll prevent the heart attacks. But recent studies with hormone replacement therapy showed that no, you don't reduce heart attacks, even though all the signs would have suggested you should have. Again, until you do the trial, you can't be positive that the benefit is going to be there.
... Why did people think hormone replacement therapy was actually going to reduce heart attacks?
We assumed, because people with low estrogens, menopausal women, had a sudden increase in heart attacks, whereas they previously-- When their estrogen was high, they had very few heart attacks. When they lost their estrogen, their heart attacks occurred. So everybody said, "If we keep the estrogen up, then we won't see the heart attacks." Didn't work that way. The estrogens were replaced, but the heart attacks continued.
How could doctors prescribe estrogen to reduce heart attacks without knowing whether they actually did reduce heart attacks?
Many things in medicine are done because we believe it's going to help. We're beginning to learn that you can't do that, that you have to know that they're going to help before you do them. We've had to do that for many years because we didn't have the data, we didn't have the tools. But today we do have the tools; we know how to study these medications. It's expensive, and yes, it will cost the pharmaceutical industry or someone a lot of money to answer these questions.
The toughest part of this is, when do you put a drug on the market? Do you wait until all the answers are in? Well, a lot of people can suffer, waiting for those medications. So that's the hard job that the FDA has to deal with. How much evidence do we need to be fairly certain that there's going to be more benefit than harm? That's the risk-benefit ratio.
When a drug goes on the market, only about 3,000 patients have ever been given that drug. We will never know all the toxicity that can occur, especially the one in 10,000 or the one in 20,000 that can be seriously harmed. Our detection of that will only happen after the drug is on the market and exposed to huge numbers of patients.
Before we get too far away from the story of encainide, how many people were hurt by having those drugs being given routinely?
One of the most difficult questions to answer about medications is how many people are harmed. We know a lot about the types of side effects, but we don't have a system in this country to quantify the harm. We can find out how many suitcases were lost by an airline last month, but we can't tell how many people were harmed by a medication. We don't have a system that can quantify it.
Estimates are that tens of thousands of patients died. One estimate is that 50,000 people died because of the encainide, flecainide and other anti-arrhythmic drugs that were being prescribed to people without any benefit. Some lives were saved by those medicines; more lives were taken. We think 50,000 is a reasonable estimate. ...
Let's talk about CERTs. ... What is CERTs, what does it stand for, and what is its mission?
CERTs stands for Centers for Education and Research on Therapeutics. They are a network of seven universities that work together to be an advocate for the patient. That's really what they do. They try to provide the balance of information to the patient, the nurses, the doctors, the pharmacists on the medications that they need to take, and provide the research, do the research, and change behaviors, so that we get more from the medication. I think their role is really to be an advocate for the patient. ...
I firmly believe that good pharmaceutical marketing and good pharmaceutical drug development requires more than just what's good about a medication. We need to know what's wrong with a medication in order to use it wisely. ...
CERTs came about because we weren't getting that other half of the information about medications. Many of us called upon the government to create a new entity outside of the FDA, because the FDA has a very limited role in regulating the industry, not informing the industry or the public. The public needs information they weren't getting about medicines, the medical community needs it and the drug industry. There needs to be research done that wasn't being done, and then that research needs to be made available in an effective way. We know that simply sending out newsletters to inform doctors doesn't change behavior. ...
So the CERTs were created to get that information that we need, and then try to develop ways to inform in ways that change behavior. That means really that you don't just tell the doctor; you inform the patients, you inform the pharmacists, the nurses and the doctors. ...
What do CERT centers want to find out about medicines and drugs?
The CERTs want to find out why we aren't getting the most from the medications. Sometimes the CERTs are looking at why aren't these medicines prescribed? Why do only half of the patients with a heart attack get beta blockers, which can save their life? We're looking into why women have more side effects and why drug interactions cause harm to women more often than men.
There's another CERT that's studying medications in children. How can we use medications more effectively and safely in children? One of the CERTs is looking at antibiotics. Why are they over-used? Why aren't the safer drugs used earlier, instead of going to the very most powerful medications, which often haven't been studied and have side effects that you wouldn't want?
To what degree are CERT centers developing information that the industry doesn't bother to develop? ...
CERTs will perform the research that the pharmaceutical industry won't do. There's no incentive for a pharmaceutical company to study the serious side effects of medications. They won't do comparative studies when they're going to lose. They won't study the old generic drugs. They won't study why certain populations are harmed and others aren't.
So the CERTs will do the work that the free enterprise system can't justify, but the society needs.
What sort of impact could the kind of research that you do in CERT centers have on people's health and safety? ... How many people's lives could it save? How many people could be healthier?
It's hard to estimate the benefit of CERT. There are many ways to do it. You could look at lives saved. We estimate that from 40,000 to 100,000 people die from their medications each year. That's a huge opportunity to reduce that. We believe that our experience right now with methadone, where we found that methadone is causing deaths-- Those deaths can be prevented. So that's where we're focusing. There are people trying to reduce hospital-acquired infections, and the CERT is looking at over-prescription of antibiotics. That can save lives. It can reduce hospital costs. We can reduce overall medication costs. ...
There's one estimate that we lose $137 billion dollars every year from misuse of medications, either by not taking the prescriptions we were given, or from the side effects or from the loss of work from the side effects. So all that, if you add it all up, is a huge opportunity to save money and lives from these medications. And yes, these are good medications, there's great value coming from them. But there's also a margin that we're losing, because we don't use them the way that we should be using them.
So you're not talking about, at the end of the road, taking dozens of drugs off the market if CERTs really do their job -- or are you?
No, the goal of CERTs is not to get drugs off the market. The goal is to get more from the medicines that we get on the market, to get more benefit and less harm, to find out why they're causing harm and then stop it. ...
What's your policy here about taking funding? Why is that important?
... Because the work that our CERT does can favor one drug or another or could help increase the market or decrease the market, we feel that to maintain our credibility, to maintain our independence, we must try every way possible to not accept any funding from any commercial source. We want to be independent. We may not be able to stay that way, and it limits what we can do, because we turn down other sources of funding. But we believe that it is essential that our data be independent.
I've seen too many examples of subtle influence of funding, where people really believe they're independent, but because they've accepted some funding from a source, it often affects what they do, but more importantly it affects the impact of their message. Once we publish a paper that says it was funded by a commercial enterprise, someone is going to doubt our data. They'll wonder, "Are these data really independent or are they selling something, too?" We want to be as independent as possible, and therefore hopefully our data and our results will have a greater impact.
Is there too much of this science funded by commercial interests such that we don't have reliable information? ...
Our information is not unreliable. It's out of balance, and that's the problem. We don't have too much information from the industry, we have too little from other sources, and that to me is the problem. I think the pharmaceutical industry is doing a great job. But there's other work that needs to be done. It's not being done, and that puts it out of balance.
It means that the public who hears about medications has this unrealistic expectation that they're great, there's no harm, there's no side effects. There may be a few little side effects mentioned in the ads on TV, but they say, "Well, I can handle those." In fact, there are many more side effects. There are drug interactions and there are other things that can happen that can go wrong that aren't mentioned on TV. The public never hears about those. They don't read the package insert. The pharmacists don't counsel them as often as they'd like to, and the physicians don't counsel their patients when they give them a prescription the way they could.
So it's not that the pharmaceutical industry has done too much. It's just that the rest of the world hasn't done their part. ...
I don't expect the pharmaceutical industry to provide all the information, because it won't be independent. If you look back at a lot of studies with the hormone replacement therapy, they were reasonably good studies, but they were designed by the people who wanted a certain outcome. We need studies designed to produce the truth, the real balanced information about medications. It will be a study that will be designed a little differently to focus on finding harm and benefit, and that's not going to happen when you rely only on the industry.
Why not? Why won't that happen? ...
We talk about the pharmaceutical industry. These are people [whose] jobs depend on finding good news. Their advancement in the company depends upon giving a good report to their superior, and therefore they're going to look for the good news. We need people who are paid to find bad news.
CERT centers are not exactly lavishly funded. ... What's standing in the way of government funding this kind of research in a big way? Why aren't we funding more of this?
The agency that funds our research, the Agency for Healthcare Research and Quality, is in a very difficult situation. They want to improve the outcomes for medical care. In this case, the CERT's focus on therapeutics and medications predominantly. But again, if you start interfering with the marketplace, if you start hurting the market, the lobbyists come in, and then your budgets are slashed.
This agency was almost wiped out a few years ago, because they put out some recommendations that the commercial community didn't like. I'm not talking about the pharmaceutical industry now. I'm talking about [that] orthopedic surgeons didn't like the recommendations for how to care for back pain. So there was a huge lobbying effort to shut down this agency.
There's the fear that if we bring out-- If the CERTs and if AHRQ bring out too much bad news, we will offend very powerful groups. I think that's the greatest fear to me -- that the CERTs will be too effective, and then they'll be lobbied out of existence. I hope that doesn't happen.
Why are we just doing CERTs now? Is the pharmaceutical industry out there lobbying against you or trying to be sure that this kind of research doesn't get [funded]?
Well, you know, CERTs is fairly new. But in fact in 1972, I think it was, Phil Lee, who was the assistant secretary for health at the time, wrote a book and called for CERTs-like centers around the country. We started calling for these in the early 1990s. I was told by many people that "This is the worst idea I've ever heard. My pharmaceutical company tells doctors everything they need to know about medications." So it offended a lot of people to say that we need balance.
We finally got the bill approved. There was a lot of battling over it. A lot of people in the pharmaceutical industry openly criticized it and fought it. In fact, it's only a very small part of the funding now. There's seven centers. We only have about a half a million dollars each, which sounds like a lot of money, but one clinical trial can cost $20 million. So that half a million is enough to hold together a group of people to begin studies and to begin to find signals. But the original estimates that we made for CERTs was that every state, every region needs a center that can serve that population. We said at least 15, maybe 75 CERTs, with budgets of at least $3 million in order to be independent.
The other very important aspect of CERT is that you need a team. You need basic scientists, you need clinical scientists, you need pharmacists, you need educators. You need marketing people, because you need to change behavior. So you need a critical mass of a very broad group of disciplines, and they need to be independently funded. That adds up to about $3 million, and the CERTs now have about a half a million. So we're nowhere near the level of funding that we need to begin to do impactful work.
We are making a difference -- I believe that. But I just look at what we could accomplish if we had more CERTs in more sites around the country.
Half a million dollars sounds pathetic in comparison to the amount of money the pharmaceutical companies spend to come up with [a new drug]. How can you possibly make a dent?
Well, one life at a time. We believe that the half a million dollars that we got last year will save some lives of patients taking methadone. We believe that the educational module that we sent out to every medical school in the country will allow them to teach drug interactions better. So we believe it will make a difference.
Sadly, I think we could make a much bigger difference if we had more members of our team with more time to spend on this. Most of people on our team, we only have about 20 percent of their salary. We have to find the other 80 percent of their salary, and we're trying to do it without going to the drug industry. So that means we have to have other NIH grants or state support or other independent mechanisms to fund our salary. ...
I get the sense that sometimes there's not a great deal of intellectual honesty in this whole industry, including among the regulators. Is this a problem?
It is. It's a serious problem, not having independent thought, even in the regulatory side. We have regulators who recommend a drug go on the market, and then they're the same ones who have to make the decision to take it off. That means they've got to say, "I made a mistake, I let this on the market and there's some problem I missed," and that puts them in a terrible situation. ...
We need for the pharmaceutical industry something analogous to the National Transportation Safety Board, so that when a plane goes down they go in and analyze what happened. They may find that the plane was the problem and the manufacturer, or they may find that the regulations were inadequate. But they're independent. When a drug comes off the market, we have no analysis to say, "Should it have ever gone on the market? Was it a mistake at the agency? Was it a mistake at the industry? Was it something that was totally unpreventable? Did the system work?"
I think it's important for people to know that drugs will always have to be taken off the market. That's not a signal that there's a problem. Medications will never be tested completely in our system, and that's OK. But we have to find the problem and we have to see could we have done it better, could we have done it differently. We don't ask the drug company, "Did you make a mistake in developing this drug?" or the FDA, "Did you make a mistake in approving this drug?" It should be another body to make that determination. So independence is not going to be there; we have to add it.
You were under consideration to be commissioner of the Food and Drug Administration. You've got pretty strong ideas about making drugs safer and as beneficial as possible. Tell me about your interview. Who interviewed you? When? What were you interviewed for, and how did it go?
I was very proud that I was considered as a candidate for the Food and Drug Administration's commissioner position. I was interviewed by people in the White House and eventually by Secretary Thompson, and it went well. I think I would have enjoyed working with him. He really cares about health and safety, and I think he recognizes that there needs to be changes in HHS and in the FDA. I felt that I would have been able to work with him.
Unfortunately, it was a very complicated process. It was going to take a long time to go through, and I was given the opportunity to come to Arizona. I'm glad I made that decision, for me personally.
As I look back over my interview and the other interviews, it was very clear that people like myself, who care about drug safety, had become too controversial. Just like Senator Kennedy didn't want to take somebody right out of the drug industry, the drug industry didn't want someone like myself, who was going to focus on toxicity and side effects. So it became clear that I wasn't going to be able to meet a broad enough constituency. I think that's unfortunate -- that people who care about drug safety don't have an opportunity to serve. I think there are good people in industry. There are good people who care about drug safety who could do a good job.
It's going to be-- It has been very difficult to find someone which has been part of the process to lead the FDA, because in being part of the process, you demonstrate that you care about something, and that alienates somebody. You either care about getting a drug on the market or looking at the side effects or looking at food safety. When you do that, you threaten some constituency, and it makes it very difficult to find someone who can not be opposed by someone else.
Maybe we could avoid beating around the bush here. How did Secretary Thompson tell you that you couldn't get this job? What did he say?
Secretary Thompson told me that I was a candidate, that I would be very high on his list, but that it was going to be a long process, that there were other considerations, that I might not be the final candidate, and that it would be a long time before I would know. So because of that-- He didn't say that someone was going to veto me; he did say that I was a candidate that would be acceptable to him, and now he's very proud of that. But he also said that it wasn't just his decision.
Did he imply to you that you would not be accepted by some other powerful constituency, that it would be a problem?
Some people on his staff indicated that my stand on dietary supplements and especially ephedra-containing supplements -- I've asked the FDA to ban these, because I think they are killing people and they must be regulated as drugs -- I think his staff told me that that had really alienated a very powerful part of the dietary supplement industry. ...
It only became apparent over time when I was told that there were other people that I wasn't going to be acceptable to, that others had ranked me further down than the one. Then people like myself, who were subsequently considered, were told at the last moment that no, it's not going to work. It became clear that anyone who had focused on drug safety couldn't make it, just like people who had come right out of industry couldn't make it. Too far on the extremes to be acceptable.
It's a pretty sad day when somebody is concerned about drug safety can't get the job.
... To me, the greatest sadness is that people who care about drug safety and food safety to the level that they get involved and they try to make a difference can't be acceptable to regulate the agency. I think that's one of the main criteria that should be used. Of course the regulators have to serve, but they also have to regulate and they have to protect. To me, protection is first; serve has to be second. You've got to take risks. You can't just be risk-averse and unwilling, because people need medications and they need access to foods, so it's the ability to balance risk and benefit. To work in drug development and focus on drug safety, you become identified with the risk and not the benefit.
Is the balance of the FDA out of whack, in your opinion?
I think the FDA is so grossly underfunded for its mission that it is out of balance because of user fees. User fees enable the agency to hire people to work for the industry. The other budget has been so limited and so cut -- the other budget being that which is there for safety -- the number of people hired at the agency to protect, to analyze data and drug safety, is criminal. The number of people required to study 3,000 drugs that are on the market is far more than the 17 or 20 -- however many they have now. The teams that are needed to do drug safety are infinitely more than what they've got right now.
We don't have a safety system in this country. We've got a good voluntary reported system, but it is not a full system. We need other tools, like other countries. In France, a very small country, they've got 30 sites where they've trained people to go and look at the medical records, talk to patients to find out what happens when you take a medication. So if it looks like there's a problem, they can actually pick it up very quickly.
In the United Kingdom, they have a network of general practitioners who report their findings on every new drug they use, so that you can not only pick up signals, but you have an estimate of how large is the signal. You know that 100,000 people took the drug and one or two people were hurt, 10 people, whatever. You have the ability to quantify the problem.
We don't have that. We know that for every adverse event that is reported to the FDA, about 100 that never get sent in. We only get 1 percent of serious reports ever sent in, and many of those are sketchy. That's why it takes a while, too long a while, to detect these problems. The system works, and one of the disappointments I've had is when I criticize the system, people say, "Well, it must be wrong then." No, it's not wrong. The system works, it's just too slow. It's too cumbersome. We have to do too much else to verify that a signal is real, because we don't have the other tools.
Why is the safety part resisted so adamantly? How do you explain it?
The budgets of the FDA are determined by lobbyists who call for money to be spent in certain ways. The user fee is negotiated to be spent on reviewers. Until recently, none of the money that was sent by the pharmaceutical industry for user fees could be used for drug safety. Now that's changing -- not to the rate I'd like to see it. Some of the money can now be spent for drug safety, but that was impossible until recently. It's only because many of us have screamed that that has to change has it changed.
Not having a commissioner makes it very difficult for them to recruit good people. It makes it almost impossible to retain people, because they don't have a leader. They don't have confidence that their work will be appreciated at the highest levels.
What's the lobbying in terms of safety?
Nothing. They're not lobbying for safety. ... The pharmaceutical industry doesn't lobby for safety. They lobby for rapid review, rapid access to the marketplace. They haven't lobbied for drug safety; no one has. There are groups, consumer groups, that have spoken out, but not with a uniform voice. Large constituencies haven't joined together to say, "Stop the harm."
When the papers came out from the Institute of Medicine talking about medical errors and medical harm, I thought surely there would be action taken and the agency would be given money to increase its safety net. But it hasn't happened. There's a lot of lobbying for helping the agency serve very little, if any -- probably none. No lobbying [is] being done to help it get its safety mission accomplished. ...
... Comparative trials. What's the importance in having clinical trials that compare different drugs, and why do we have so few of those, at least have so few honest ones? There are lots of trials in the literature that compare two competing drugs.
I used to work for a drug company. I began my career working for a pharmaceutical company. I know, when I was deciding how to spend my money, it was to find a trial that would find the benefit. So we would go after-- We'd put our drugs up against somebody that we could beat. That's what industry does. I mean, that's good business. You don't want to stop that. So that's what we have. We have an industry that will do a study to show that their product is better than the other. That's not really a comparative trial; that's a marketing trial.
A true comparative trial will take a group of medications that are potential choices and really compare them fairly to see which does give you the best outcome at the best cost. Those kind of comparative trials are rarely done, if ever. I can't think of any. We wanted the CERTs to do comparative trials. We don't have enough money to do one of those in our CERTs, and if we did one and we found that one drug was better than another, we'd be in trouble. ...
What would be an ideal drug safety system? What should be an ideal?
The ideal drug safety system would be one that had the ability to look at drugs throughout their lifespan, as soon as they go on the market, to have studies designed to look for problems that could not have been detected before marketing. It would involve spontaneous reports for those rare reactions. It would involve a French-type system, where you go in and look at subsets to see what happens.
And then even think long-term toxicities-- I mean, there is a possibility that there are side effects with medicines that will take 10, 20 years. We're now in an era where we're taking drugs for 20, 30 years. We're taking drugs to lower cholesterol. We're starting them in our 20s and 30s. That's never happened before. No one in our society really has taken medications for that long a time before. There's no way we will be able to pick up toxicities unless we have controlled trials where we have groups to compare, and that means we need the ability to do randomized, long-term safety trials. Those are very expensive to do.
We need community-based studies. We need to look in subsets of the population. Those aren't being done now. We don't look at women. ... We need to be doing targeted studies in Hispanics, in Native Americans, because there will be toxicities unique to those populations that will be totally missed if we just look at everybody in the country.
There are many things we need to do. There needs to be a team of people independently funded.
By not doing these things, what's the cost?
The cost of not having this kind of a safety system is we take drugs off the market that shouldn't be taken off the market. ... We lose 100,000 lives every year. We lose $137 billion because we don't have an adequate safety system.
We don't have a system that can prevent the harm. We spend a lot of time testing to prevent harm with medications, and we don't even know it will work. A safety system can identify what will work, what will reduce toxicity.