Pharmaceutical liability

Derek Lowe ponders whether FDA approval should pre-empt liability:

Arguing will now commence on whether pre-emption is a good idea or not. I've thought for some time that all approved medications should be labeled as "investigational new drugs", and that everyone taking them agrees that they are participating in a post-approval clinical study of their safety and efficacy. (I suppose that's my own form of pre-emption). But there's room to argue if the FDA is ready to take on the full responsibility of drug approval, without the option of later redress in the courts if something goes wrong. (Counterargument: that's what they're supposed to be doing now. . .) And all of these schemes have to make room for new information turning up, or for outright fraud (which is most definitely in the eye of the beholder). Personally, I'm glad not to be a judge.

I wonder why the FDA doesn't do the clinical trials itself. Our drive for liability stems from our fear that the pharma companies know something they aren't telling us--so why not give the responsibility for testing this stuff over to a presumably neutral arbiter?

Comments

Having worked on clinical trials for a medical device, I have a pretty good idea why the FDA (or more to the point, Congress) doesn't want to get involved. They are hugely expensive operations, there are literally thousands of them going on at any time.

It would require thousands of government workers to run clinical trials for every drug and medial device under development. Every medical specialty and sub-specialty would need to have its own department, with domain specialists as integral parts of the team.

The neutral arbitration that is required is that which is being eviscerated: vigorous oversight of the methods and results of the trials themselves. FDA auditing staffs have been slashed and when an action against a manufacturer is called for, the career professional has to get approval from a political appointee to proceed. The latter requirement is new to the FDA, implemented by the free market thinkers of this administration.

To me, this is one area where hard-core, anal-probing, rigorous, regulatory agencies and not only a good idea, but a moral imperative. The profit motive is too strong, and the financial risks too great, not to have an enforcement mechanism that punishes the living crap out of any company that crosses the line.

Posted by Kevin | April 8, 2008 11:14 AM

Having worked on clinical trials for a medical device, I have a pretty good idea why the FDA (or more to the point, Congress) doesn't want to get involved. They are hugely expensive operations, there are literally thousands of them going on at any time.

It would require thousands of government workers to run clinical trials for every drug and medial device under development. Every medical specialty and sub-specialty would need to have its own department, with domain specialists as integral parts of the team.

The neutral arbitration that is required is that which is being eviscerated: vigorous oversight of the methods and results of the trials themselves. FDA auditing staffs have been slashed and when an action against a manufacturer is called for, the career professional has to get approval from a political appointee to proceed. The latter requirement is new to the FDA, implemented by the free market thinkers of this administration.

To me, this is one area where hard-core, anal-probing, rigorous, regulatory agencies is not only a good idea, but a moral imperative. The profit motive is too strong, and the financial risks too great, not to have an enforcement mechanism that punishes the living crap out of any company that crosses the line.

Posted by Kevin | April 8, 2008 11:14 AM

Having been through the wringer of relatively rare cancer, and spending the last four years researching the genetics of hematopoietic malignancies, the last thing I want to see is some new, massive regulatory anal-probing, punishes-the-living-crap out of people system that does even more than the FDA _already_ does to discourage things like more "exotic" treatments. (For example, 'tailored' gene-targetting drugs that have effectiveness for "only" several hundred or a few thousand cancer victims annually.)

Posted by mmokc | April 8, 2008 11:19 AM

The article in the Times about preemption was a dual editorial disguised as front-page news. They made great shakes about a case, the birth control patch, that is not eligible for preemption under existing laws. They also listed a litany of drugs, e.g. Vioxx, that save lives and have been denied to consumers due to flashy trial lawyers persuading gullible juries with fancy powerpoint presentations.

Oh, remember how SSRI's supposedly increase the risk of teenage suicide? An epidemic! Right--except that when the FDA included a black-label warning about this, zipcode by zipcode teenage suicide skyrocketed. That's blood on the hands of the Times and other hysterical anti-big-pharma outlets.

Personally, I think that drug companies should be required to accurately and transparently report the efficacy and safety of any drug on the market. After that, it should be buyer beware. So long as the FDA is set up as a nanny-state apparatus, dedicated to controlling what we can put in our bodies, rather than the informational service it should be, I think advocates of preemption make a fair argument.

If the FDA would step aside, though, then let a thousand lawsuits bloom.

Posted by Smokey's Bandito | April 8, 2008 11:43 AM

The Hundred Year Lie "dismantles many misconceptions and offers readers practical solutions for making their lives healthier, including an entire chapter on self-detoxification. Capped with a comprehensive bibliography, Fitzgerald's book is a well-crafted and thought-provoking text."
Library Journal
http://www.hundredyearlie.com/


"...A must-read survival manual for the human species and planet Earth."
Terrence Cafferty, MSME, NASA Consultant


"a damning treatise on how food and medicine are destroying the public's health. He provides devastating evidence on how chemicals are damaging our health even though it was supposed to improve it, sort of a case of science of nature. Very scary...enlightening...Well worth a read."
The Union Jack

the FDA is FOS

Posted by MEH | April 8, 2008 11:58 AM

"I wonder why the FDA doesn't do the clinical trials itself. Our drive for liability stems from our fear that the pharma companies know something they aren't telling us--so why not give the responsibility for testing this stuff over to a presumably neutral arbiter?"

It is far too expensive for the FDA to do. And since they would be working with limited money, that would mean that far fewer drug trials would get done than now--practically guaranteeing that many perfectly good drugs wouldn't get approved until years after they would under the current system.

Posted by Sebastian Holsclaw | April 8, 2008 12:11 PM

An open invitation for corruption of FDA and officials...

Approve the stuff, and land a cushy job after your tenure at FDA....

Sick.

Ever head of checks and balances?

Posted by D | April 8, 2008 12:22 PM

D,

they take the checks and hide the balances..

Posted by MEH | April 8, 2008 12:38 PM

To expand a bit, my point is that the big pharma and the device manufacturers want to eliminate effective oversight altogether.

If they can simultaneously get pre-emption and gut the FDA, the cost-benefit analysis for exacting safety standards demands that safety be cut. And trust me on this, the CEO's and CFO's of every one of these companies will cut expenditures on safety and clinical trials the very next day after a pre-emption ruling.

Fear of monetary punishment is the only language that is understood and by the time the market provides it (if at all), hundreds of people may have died.

Senor Bandito posted the following:
"Personally, I think that drug companies should be required to accurately and transparently report the efficacy and safety of any drug on the market. After that, it should be buyer beware."

Really? And what if it turns out that a drug is suspected of causing pancreatic cancer 15 years down the road? Should the company get a free pass on that one and continue to sell the drug? If the link is tenuous and the drug a blockbuster, there is no way on God's Green Earth they would pull the drug. Unless the FDA has the ability to order new clinical trials (which it currently doesn't) or force a recall, I think you could expect the Tobacco Institute defense.

And for mmokc's point, I do understand that those waiting for treatment for rare and exotic diseases want a faster turnaround. But there are reasons why experiments on humans are not done early in the development cycle and only after research shows that benefits outweigh risks. One only has to look at Thalidomide to see the hazards of rushing a product to market.

And for those who think that companies won't obfuscate to say that their products aren't the cause of death, I direct your attention to thirty years of debate on the smoking/cancer link.

Posted by Kevin | April 8, 2008 12:56 PM

Complete transparancy in the trials themselves should be the main goal, and an achievable goal. I would not favor having the government run the trials since that is a guarantee that the trials will be more costly and slower than they are today.

As for pre-emption, I don't see the legal basis. What we need are better judges for civil courts so that real harm caused by the drugs can be assessed, and the remedies be given to the actual victims. What we have now is a complete fraud, in my opinion.

Posted by Yancey Ward | April 8, 2008 1:13 PM

Kevin wrote: "Really? And what if it turns out that a drug is suspected of causing pancreatic cancer 15 years down the road?"

What if, what if? If the lay public and doctors have open access to the clinical trials that produce that suspicion, and they choose to use/prescribe the drug anyway, then obviously they think the benefits outweigh the risks.

Now let me play the "what if game". What if an over-regulated pharmaceutical industry stifles innovation, leading to fewer drugs being released, thus causing untold millions of people to die? What if what if?

Posted by Smokey's Bandito | April 8, 2008 1:32 PM

I do believe that Smokey has missed my point, but perhaps I was unclear. The hypothetical link to pancreatic cancer wouldn't be contained in the clinical trials used to approve the drug, as trials generally don't run that long. I am talking about a risk that is revealed after approval, one that the FDA cannot ask for a another clinical trial to determine the validity of.

And if "lay people and doctors" had access to raw clinical trial data, virtually none of them would be able to make sense of it. Biostatistics is a dense field, one that even people with higher math in college are unlikely to penetrate. Most physicians have some stats, but not nearly enough to get through this kind of stuff. Clinical trials documentation runs to millions of pages and even the summaries are hard to understand. I have degrees in math and computer science and helped contribute data to several clinical trials and I haven't the faintest idea how to tell if the statistics have been manipulated.

As it is, you can read clinical trial results yourself now, if you have the time. Medical journals publish the results of trials all the time. There are multiple thousands of trials every year, both in the U.S. and Europe and if you know where to start, more power to you.

And I ask this out of curiosity, how would you know if an industry is over-regulated? By industry complaint, death rate, number of drugs released? I err on the side of caution, that new drugs have to prove their efficacy and safety first before release, rather than assuming safety until proven otherwise.

"Untold millions"? Let us pull back on the hyperbole, please.

Posted by Kevin | April 8, 2008 2:03 PM

Kev,

I am proposing that the FDA's main function should be in reporting the efficacy and safety of drugs. I object on moral grounds to the idea that they have a right to regulate which drugs people end up using. If someone wants to take cyanide, then it's not up to society to stop them, it's their choice--but they should be informed of what it is, so they can make a choice.

You'd like to "err on the side of caution," but the level of caution that would be required to block the release of the hypothetical drug that causes pancreatic cancer 20 years down the line would also block many other drugs that would more than offset the lost lives.

So, without hyperbole, I can say that you support a policy that would lead to millions suffering and dying. Further, you would restrict the options of all in accomplishing this goal. Good show! :P

Posted by Smokey's Bandito | April 8, 2008 4:44 PM

Cinderella, are your identifying with the Greenwald criticism of you? This is not a good idea. The studies are not done by the drug company anyway. They're done by and large by medical school faculty. One of the powers of the chairman is to decide who gets to do the studies, they generate money which that person can use to do research which is how I managed to get this done; but I digress. For 'fair and balance' perspective the New England Journal of Medicine will publish any data that has been tortured to the point that John Yoo would blush if it makes a drug practice look bad. The data submitted to the drug companies and then the FDA is thus largely produced by the equivalent of the ideal Caesar's wife (just ask Freddie).

Posted by Michael | April 8, 2008 5:40 PM

Smokey,

Your objection may be on moral grounds, but they come from first principles I disagree with and saying that your reasons are moral doesn't make them logical. I can also say that you have twice completely managed to ignore what I have said.

Please note that at no point did I say that I would block the release of a drug for 20 years. In fact, I said exactly the opposite: that after a drug has received regulatory approval and has been used for 20 years, new risks may appear. There is no mechanism that the FDA has to ask the company producing the drug to conduct a study to verify the risk.

As I understand your position, it should be up to the companies to decide which drugs should be marketed and people to decide which drugs to take. The FDA would only provide information and not approval. There is a reason this is not the system we have today, it's called Thalidomide. That is the moral reason I have for desiring vigorous drug regulation. That's is a concrete example of a tragic outcome from lack of regulation whereas you have only a speculative "untold millions".

What if the drug in question was a widely prescribed diabetes drug that caused catastrophic liver failure instead of the deliberately small example I picked (so as to avoid hyperbole)? Think of how many millions would be affected then.

To me, this is where the libertarian formulation breaks down. Please tell me how an uneducated person should be expected to evaluate clinical trial information? How they are supposed to do anything but take the word of a doctor who is being swamped with drug company money to prescribe whatever wonder drug the big pharma pitch girl (and they are almost always girls, usually former college cheerleaders) shows him that month?

Lastly, can I assume from your lack of response on the availability and usefulness of clinical trial data to the lay person that you agree with my position?

Posted by Kevin | April 8, 2008 7:57 PM

Kevin,

If there is no test to predict the damage caused by the drug 20 years down the line, there is nothing the drug company could have done not to harm these people (short of stopping making drugs). Why should they then be liable for those 20-years later deaths?

Posted by Joe | April 9, 2008 5:49 AM

I didn't say anything about liability. I merely want someone to have the authority to order a new clinical trial and the ability to remove the drug from the market if trial results market.

Assuming that initial approval trials were conducted honestly and above board (and that's a big assumption for some companies), then no liability would attach.

Posted by Kevin | April 9, 2008 7:50 AM

Should read "if trial results warrant".

Posted by Kevin | April 9, 2008 7:55 AM

Kevin,
There is one large problem with your plan (it's an especially large problem with the particular disease and number of years before that disease appears that you chose for your example, but I think the problem would be present to different degrees regardless).

A drug patent lasts 20 years. The patent is usually filed when the molecule is first invented, so it's typically another 5-10 years to go through preclinical and clinical trials. These trials are expensive (around $897 million per successful drug, accounting for failures. That's a number from Tufts from 2003, I believe), but they're worth pursuing because you have monopoly pricing power until your patent expires.

If after 20 years of being on the market, the FDA tried to force somebody to perform a clinical trial (the particular type you're asking for is known as an outcomes trial, which can have direct costs of ~$100-200 million), the companies (most of them generic at this point) would simply pull the drug. They wouldn't be able to get a return on the money they invested in the clinical trial. If you forced the initial drug discoverer to do the trial after the fact, they'll have to take such possibilities into account, so you would simply be raising drug development costs.

I don't think forced after-the-fact clinical trials is the way to go.

I should mention I'm an engineer at a big pharma company.

Posted by Chris | April 9, 2008 7:57 PM

1. Good point, at 20 years (I originally had it at 15, but somewhere along the line it expanded) costs would ballon to make it impractical. You are correct that most effects would show up long before that point, which I believe actually makes my arguments more on point.

2. If they pulled the drug in a case like that, fine.

3. During patent protection, blockbuster drugs bring in a billion plus a year, making an outcomes trial worthwhile. And pharma companies are continually trying to find ways to expand patents, such as the devious little "combine this and that" formulations.

Engineer to engineer, what changes would you make to the system? Pretend that the pointy-haired demons above you don't exist.

Posted by Kevin | April 9, 2008 9:41 PM

Kevin,
In bullet point 2., you said it would be ok to pull the drug if the trial couldn't be completed. I disagree. The drug definitely works for some segment of the population, but some possible link to a serious side effect appears after 20 years. My gut feeling is that such a drug provides a therapeutic benefit to patients and it shouldn't be pulled from the market. If I misunderstood the original example or properties of the hypothetical drug, please correct me.

As for 3., you are exactly correct. Lipitor was not the first statin to market (in fact, it was almost killed by Warner-Lambert because they thought the statin market was saturated). However, it turned out to be a very good molecule and Warner-Lambert (later Pfizer) ran extensive post-launch clinical trials that led it to become the world's best-selling drug (the huge marketing push didn't hurt either). So long as there is patent life left on the drug, the pioneer company will probably have incentive to run trials (especially if it's a blockbuster).

As for the overall system, the whole problem has to do with information. Acquiring information about the drug (via clinical trials) is extremely time-consuming and expensive. Once you have the information, it's extremely difficult to analyze and even more difficult to distribute. As was mentioned earlier, patients and doctors, frankly, are not fully equipped to make informed decisions about drugs.

Additionally, the information is collected and controlled by a less-than-impartial party (the pharma company). If Ford told you they did a study that proved their truck was better than GM's, what would you think? That being said, the people working in the pharma companies are much more technically competent and faster than most of the folks at the FDA, so I'm loathe to turn over clinical trials to the FDA (even leaving my other libertarian sensibilities aside).

So some features of my system would look like this:
The New England Journal of Medicine requires that clinical trials be pre-registered with the journal. If you didn't pre-register, you can't publish your results when they come in. This is an excellent step to prevent negative results from being swept under the rug.

Payers (such as insurance companies etc.) are demanding more head-to-head comparison trials of new drugs vs. old drugs. In addition, they're demanding to see the pharmacoeconomics of new drugs to demonstrate that the new drug really is worth the price. These trends should continue. The insurance companies are actually equipped to read clinical trial data and have the leverage to negotiate on price (or even leave a drug of the formulary). In my system, the insurance companies are the groups that really impose market discipline on the pharma companies.

In my system, the FDA would allow drugs onto the market after limited trials (say, phase II, once the drug has been proven effective). That's much cheaper, so it would spur competition amongst pharma companies; as it stands, only a few companies even have the ability to actually bring a drug to market; small innovative companies still need to go to the big guys if they want to reach market. Further clinical trials might be demanded by the insurance companies (see above).

FDA approval would not shield the companies from liability. Drugs aren't really that different from every other product on the market (everyone seems to think they're so special). The threat of liability is a way to ensure discipline after receiving approval (otherwise, it's a free pass). As a separate issue, I think we need to reform how complex technical cases are litigated, but that's for another day…

Posted by Chris | April 10, 2008 10:24 AM