Patients Beg FDA For Access To MS Drug

Ronald Bailey | March 8, 2006

Multiple sclerosis patients are begging the U.S. Food and Drug Administration to allow the drug Tysabri back onto the market. The biotech companies Elan and Biogen pulled the drug when 3 patients in a 7000 patient drug trial came down with progressive multifocal leukoencephalopathy, or PML.

One 36 year old patient, Heather Smith, told an FDA panel, "I know Tysabri worked for me when all other MS drugs failed. Each patient has the right to make their own choice." Earlier drug trials indicated that Tysabri is perhaps twice as effective in delaying the progression of MS than are older drugs. However, it is estimated that 1 in 1000 MS patients would die of PML induced by the drug. This compares favorably with an American's lifefime odds of 1 in 228 of dying in a car accident.

The drug companies propose to register and closely monitor all patients who take Tysabri. Such a system essentially becomes a post-market drug trial which should become the model for future drug approvals.

Disclosure: I have owned 100 shares of Elan and 100 shares of Biogen for many years. The benefit to me of an FDA approval pales in comparison to benefit that suffering MS patients would get.

Comments

|3.8.06 @ 11:20AM|#

If the incidence of side effects turns out greater than 1 in 1000, then you would be cool with the manufacturer paying tort claims on this basis?

Or is the 1 in 1000 stuf just sort of FYI?

|3.8.06 @ 11:27AM|#

Dear Questioner: All drugs have side effects and in general drug companies should be legally responsible only for the ones that they know about but don't warn patients and doctors about.

Personally, taking a 1 in 1000 risk doesn't sound very risky to me,especially in the face of such a devastating disease, but others will disagree which is why it should be left up to well informed patients.

|3.8.06 @ 11:29AM|#

I am asking what the adverse economic consequences for the drugmaker should be (if any) if it turns out that the risk is really 10 in 1000 after the product has been subject to commercial distro.

|3.8.06 @ 11:32AM|#

"If the incidence of side effects turns out greater than 1 in 1000, then you would be cool with the manufacturer paying tort claims on this basis?"

Not sure what tort you would get if the mfr stated that there was a statistically significant, if currently unquantifiable, risk for this side effect. Anyone who took the drug would be assuming that risk.

I guess you could argue that they didn't do enough to determine the rate, but I don't think that's a very strong argument (although it would work in front of certain juries).

Obviously, these people are aware that there is a high risk, but it's one they are willing to take. Is there anyway you can make the value call on whether it is worth it to live 5 more years in agony vs living 5 less years relatively pain free? I can't. I have no idea what it's like to live with MS. I respect these patient's rights to choose for themselves which they value more.

|3.8.06 @ 11:40AM|#

I respect these patient's rights to choose for themselves which they value more.

I am asking: what if they make their choice based on info from the mfgr that underestimates the risk? Does that undermine the patient's consent (or whatever you want to call it) from a libertarian perspective?

|3.8.06 @ 11:57AM|#

I imagine the presence of a tort depends on whether negligence or fraud was involved. The company is, of course, expected to act in good faith, but the issue here is whether an individual has the right to determine for themselves what risks they want to assume.

|3.8.06 @ 12:12PM|#

it is estimated that 1 in 1000 MS patients would die of PML induced by the drug.

So we can only have access to medicines approved by an organization that figures a .01 percent chance of dying is less acceptable than a 100 percent chance of suffering horribly from a terrible disease.

I hope every single member of the FDA and DEA comes down with one of those diseases that can only be effectively treated by currently illegal drugs. Every damned one of them.

Suffer, you bastards.

|3.8.06 @ 12:18PM|#

Number 6,

If the mfgr seriously underestimates the risk (let's say 10 in 1000 is the real value for the 1 in 1000 probability quoted in Bailey's post), then does that raise an inference of negligence? Or is the burden on the dead patient's family to duplicate the safety research and figure out where the mfgr made its mistakes?

It seems like under your view of how things should be it would be pretty difficult for a decedent to prove a case even if it was stipulated that the risk was 10X greater than advertised. Is this how you really believe things should be?

If the risk is underestimated, then wouldn't it be fairer to move the burden of persuasion to the mfgr to show lack of negligence. It is hard not to notice that the mfgr has clear financial incentives to underestimate risks, which kind of makes such underestimation automatically suspect to me.

Timothy|3.8.06 @ 12:27PM|#

Except, Dave, in the American legal system innocence is assumed until guilt is proven. The burden of proof is, and should be, on the plaintiff.

Unless you're proposing that we should, you know, alter the entire structure of our legal system to more closely resemble Napoleonic courts. I mean, if you want to be assumed guilty by the state, well, go ahead; but count me out.

|3.8.06 @ 12:28PM|#

Not sure what tort you would get if the mfr stated that there was a statistically significant, if currently unquantifiable, risk for this side effect. Anyone who took the drug would be assuming that risk.

How about cigarettes?

|3.8.06 @ 12:33PM|#

Hope Peter Venkman shows up with a proton pack and puts you in a containment unit, ghost.

|3.8.06 @ 12:34PM|#

I don't know about Dave, but I guess I am saying that a serious underestimation of the risk f side effects is akin to finding a trout in the milk. To put it another way, I am saying that underestimation itself should be considered as evidence of negligence. It seems like even a libertarian would want to provide strong incentives to drugmakers to either get these risk numbers correct, or at least err on the side that is opposite of their obvious economic incentives.

Is that too John Edwards?

amazingdrx|3.8.06 @ 12:37PM|#

1 in 228 from a car accident? How about from a terrorist incident? Maybe 1 in a million?

Thank gaaawd for the patriot act!!

But by all means, never institute government regulations to require safety devices proven in auto racing, like roll cages, crash tubs, three point harnesses, and explo-safe gas tanks,installed in cars. That would violate the rights of corporate citizens.

As would requiring that critical facilities like US ports be managed by US corporations, that can be fully investigated and held accountable for a mushroon cloud (for instance), caused by something in a container passed through their area of responsibility. Must protect the rights of UAE government owned corporations especially!! They fall under the US constitution, right?

As far as the drug companies concerned in this case, perhaps they did not exersize their right to "free speech" (bribery) by making all necessary "contributions" (bribes) to the apropriate political campaigns? Frist needs your support!!

|3.8.06 @ 12:42PM|#

I would like to put forth the following theory:

amazingdrx isn't an actual human being. He's actually a variation of a spambot that just queries democraticunderground, copies something at random, and pastes it here.

Timothy|3.8.06 @ 12:45PM|#

To put it another way, I am saying that underestimation itself should be considered as evidence of negligence.

If it's evidence then make that argument, and see if it works on the jury (or the judge in the case that the defendant opts for a bench trial). The onus is still on the plaintiff to make that more persuasive than the defense's case. Bring in a statistics expert to ramble on about p-values and F-stats (or autocorrellation or out-of-sample forcasting or moment generating functions, or whatever the hell else), whether it's possible to underestimate risk that badly by accident, etc. That's the duty of the plaintiff, to show that such an underestimation of risk constitutes negligence which, in turn, is grounds for financial renumeration. People, even evil evil drug companies and the purveyors of delicious corn syrups, are assumed innocent...therefore the plaintiff has the burden of proof.

Is that so very hard to understand?

|3.8.06 @ 12:51PM|#

Bring in a statistics expert to ramble on about p-values and F-stats

In my hypothetical the stated risk is 1 in 1000 and the real risk is stipulated at 10 in 1000. In a case like that is there any need for hired guns and their F-stats? My mumbo jumbo detector has gone yellow.

R C Dean|3.8.06 @ 12:54PM|#

Questioner seems to be a little fuzzy on the issue of liability.

People aren't entitled to dig into the nearest deep pocket just because something bad happened to them.

They first have to show, generally speaking, that said deep pocket was negligent or committed some wrongful act.

amazingdrx|3.8.06 @ 12:56PM|#

"I would like to put forth the following theory"

Geek is a shill bot designed by a secret DARPA/IBM reserach program to flood the net with pro-corporate spam.

Apply "The Turing Test". I rest my case.

R C Dean|3.8.06 @ 12:57PM|#

But by all means, never institute government regulations to require safety devices proven in auto racing, like roll cages, crash tubs, three point harnesses, and explo-safe gas tanks,installed in cars. That would violate the rights of corporate citizens.

It would certainly prevent people who don't want to be forced to pay for those features to do so. As with most such rules, the impact would be felt mostly by the poor.

And, of course, every regulation is a barrier to entry to new firms, and a disproportionate burden on smaller firms, and thus serves to entrench the largest "corporate citizens" in virtual oligopolies.

Is that what you want, drx.

Timothy|3.8.06 @ 12:58PM|#

The point is that if the stated risk is 1/1000, and the real risk is 1/100, the burden of proof is on the plaintiff to:

1) Demonstrate that the higher claim of risk is correct.

2) Show that this constitutes negligence.

3) Show that the negligence is grounds for financial renumeration.

My point about bringing in an expert is mainly that clinical trials do not occur in a vacuum and are often published in major medical journals. The FDA certainly has them, and I'm sure they could be obtained for a tort. The burden is on the plaintiff to show that the trial provided to the FDA understates the risk: that can be through demonstration of flawed methodology, incorrect statistical methods, presenting other published studies on the drug, and probably a lot of other things because my list certainly isn't complete.

Why? Again, because we have a presumption of innocence. The End.

I'm off to argue with a brick wall now, I'll catch you on the flip side.

|3.8.06 @ 1:02PM|#

They first have to show, generally speaking, that said deep pocket was negligent or committed some wrongful act.

Telling someone that something is ten times safer than it really is seems like a wrongful act. Akin to, if not, fraud. Caveat emptor, but isn't one entitled to rely on the seller's representations? Or do more lenient rules apply to clients one can bill lots of time to?

|3.8.06 @ 1:05PM|#

The burden is on the plaintiff to show that the trial provided to the FDA understates the risk

How about just showing that the incidence of the side effect has turned out to be 10X more prevalent than advertised? I say this in itself can meet the burden (assuming it is not credibly rebutted). Agree or disagree? Not clear from your last post, Timothy.

|3.8.06 @ 1:05PM|#

"Apply "The Turing Test". I rest my case."

Given that this blog's comment column is practically nothing more than a large Turing Test, I don't see a need to apply it.

And given that your posts are highly erratic, generally incoherent, and don't ever actually offer any sort of rational insight to the topic at hand, I have to question if you're fully human.

Whether this means that you're a spambot, or someone whose mother repeatedly dropped him on his head as a child, well, that's up for debate, I suppose.

|3.8.06 @ 1:06PM|#

Or do more lenient rules apply to clients one can bill lots of time to?

Wow, accusing R C of having a conflict of interest. That's such a new tactic.

|3.8.06 @ 1:10PM|#

If the risk is underestimated, then wouldn't it be fairer to move the burden of persuasion to the mfgr to show lack of negligence.(sic) It is hard not to notice that the mfgr has clear financial incentives to underestimate risks, which kind of makes such underestimation automatically suspect to me.


Here is all the info on the specifics of the Tysabri research and statistics, Ghost:

http://www.medpagetoday.com/Neurology/MultipleSclerosis/tb/2772?pfc=101&spc=230


I'd like to point out that while you'd like to put the blame on corporations for anything bad that might happen to someone taking a highly sophisticated immune drug, the research is primarily being done by respectable research facilities and hospitals, i.e. disinterested third parties who are hired by the corporations to do these studies. You'd like to make it out as though the "big bad corporate world" is trying to make a profit off of sick people, but the truth is that researchers have (or are not legally supposed to have) any financial stakes in the matter whatsoever, and the statistics are the best possible estimates under the research circumstances.

|3.8.06 @ 1:12PM|#

"Wow, accusing R C of having a conflict of interest. That's such a new tactic."

He probably force-feeds corn syrup to kittens, too.

|3.8.06 @ 1:17PM|#

Telling someone that something is ten times safer than it really is seems like a wrongful act. Akin to, if not, fraud. Caveat emptor, but isn't one entitled to rely on the seller's representations?

Yes. And no one is lying. All of the risk information is well documented (as is for any drug) as is required by the FDA. And as far as anyone knows, the calculated risk is still only 1 in 1000, not 1 in 100. No one will ever know if that number may be higher in a general population (unlikely, regardless), if the drug is not made available. What is your point, Ghost? I'm afraid you don't have one.

|3.8.06 @ 1:21PM|#

Yeah, but smacky, if they actually had to use real science in the courtroom how would Dave W. be able to collect vast sums of cash settlements using nothing but shady innuendo?

|3.8.06 @ 1:21PM|#

Questioner,
Think of it this way. Say the risk is 1 in 1000. If you test it on 100 people, you might not see any adverse event. To know if there is one, you might need to test it on 2000 people. Say 2 people get the adverse event. From that data, you can say "the risk as we know it is 1 in 1000." Is the risk really higher than that? Problem is, the only way to answer that question is to expose more people to the drug to see if they experience an adverse event. So, it's entirely possible the risk is underestimated. But the only way to know that for sure is to have more people take it. Does it make sense to you that people who suffered an adverse event would be able to say "the drug company was negligent because they should have tested this drug on more people who were not me before they allowed me to take it?" That's basically asking for safety for yourself by making other people take unknown risks in order to quantify them.

|3.8.06 @ 1:22PM|#

To be clear, I didn't mean to imply that R C regularly receives such accusations. I meant that Ghost frequently makes such accusations.

I may not always agree with R C, but I have never, ever gotten even the faintest hint of shill from him.

|3.8.06 @ 1:25PM|#

Ghost,

The big problem you have is that determining risk isn't an exact science. It takes large studies in humans to make that determination - by which time, well, there are a large number of people who have been exposed to it. As long as the company says that our best numbers indicate 1 in 1000, but we currently don't have enough data to make that claim at 95% confidence, the burden would be on the plaintiff to demonstrate fraud, and not just negligence. Patients have been told that there are risks (and in fact, there are essentially unknown risks with every single drug you put in your body - you may be the 1 in 1 billion person whose genotype is defective for some enzyme that metabolizes the drug, or any number of other possibilities. That's why medicine is called an art and not a science, even though it is evidence based) - and they chose to take those risks, willingly. That's noone's fault other than their own.

Saying we have to wait until we have 95% confidence in our risk assessment is not only impractical, it's impossible.

|3.8.06 @ 1:27PM|#

Does it make sense to you that people who suffered an adverse event would be able to say "the drug company was negligent because they should have tested this drug on more people who were not me before they allowed me to take it?"

LisaMarie makes a good point. More importantly, drug trial participants are made well aware before they enter the study that they are participating in research trials, meaning "trial"="not determined"; "research"="gaining information about patient/drug interaction". Meaning they are notified that there may be possible risk, and some of that risk may as of yet be unknown.

|3.8.06 @ 1:27PM|#

LisaMarie makes a darn good point. Somebody has to take it before we know if it's safe. What makes you so special that you should be protected from risk but the test subjects shouldn't be?

|3.8.06 @ 1:32PM|#

Just to be clear, Smacky sez:

"but the truth is that researchers have (or are not legally supposed to have) any financial stakes in the matter whatsoever, and the statistics are the best possible estimates under the research circumstances."

Assuming there's a "don't" missing in the first phrase, I think Smacky's wrong. They generally do (doctors get well-remunerated for being a part of a trial, even if it is only the big gala affairs they are invited to at the kick-off and close of the study. They generally want to make their customers happy, so they can get repeat business), although the use of double-blind conditions is specifically targetted at eliminating this problem, along with others. The double-blind study is a powerful tool, but it can be (and has been) manipulated. You can't prevent that, all you can do is make sure that there are consequences for those who do (manipulate, that is). Which is why I agree with Smacky's second sentence regardless of my disagreement with Smacky's first sentence.

|3.8.06 @ 1:33PM|#

LisaMarie makes a darn good point. Somebody has to take it before we know if it's safe.What makes you so special that you should be protected from risk but the test subjects shouldn't be?

More importantly, since when are researchers expected to be psychic? As quasibill noted, science is evidence-based. If people like "Ghost" are going to frivolously scream "lawsuit" every time some medical advance is being tested, nothing will ever be accomplished. Bottom line: if you personally don't want to expose yourself to risk by taking a drug that has been researched and shown to be fairly effective, don't take it. No one is forcing you to take it. But it is not your business to make that drug unavailable to other people if they want to take it.

|3.8.06 @ 1:34PM|#

Smacky,

You misunderestimate me. Let me be more explicit about my concerns and my agenda. I think a lot of what goes on at the FDA is nonsense. That stuff Doctor Rx said about bribes is not incoherent. It is probably close to the secret dealings of the secret doctors type truth. Government agencies get corrupt. Especially ones that deal largely in secret for big money stakes. Libertarians (at least ones not beholden to companies who like the bribing game) should be able to understand a natural tendency of government regulators to get corrupted in this manner.

I now propose a significantly different model to the present system to stem the FDA corruption game, while still insuring that patient's have an accurate idea of the risks of their medicines before agreeing to take them.

My proposal is to change who makes the ultimate decision of whether a drug is safe from the FDA to the patient. However, the patient can only make that decision if they know the risk. So my proposal is to simply require drug manufacturers to identify all side effects in their drugs and put down a conservative estimate of the risk of each side effect. If the published risks are accurate (or erring on the "safe" side) and inclusive -- well, that disposes of FDA safety testing and disposes of the problem that Bailey's company is having with its new product launch.

The quid pro quo is that if the drugmaker does not identify side effects, or if it underestimates risk, then it pays in tort for all the damages occasioned. A big incentive to get the risk numbers correct and complete so that patient's really can assess the risk of a medicine relative to an auto risk or a terrorism risk. that is what we think should be happening after all, isn't it?

So, Smacky, since I am proposing to cut waaaaaay back on FDA nonsense, I think you can see I am not exactly anticorporate (antigov't, yes, antitrust, sure). But if patients really are trying to evaluate risks as Bailey suggests and I agree they should, then these patients need a reliable basis to work from. I propose exchanging the FDA for a consumer labelling requirement enforced with some serious teeth. The sad suffering people with MS deserve no less, I think. Maybe the pain is so bad that they want to take a drug with a 99% fatality rate? Great -- just make sure that goes on the label. And send the FDA back to looking for spyder eggz in the BubbleYum.

|3.8.06 @ 1:37PM|#

I'm way outside of my field of expertise here, but why not rely on products liability law instead of all-or-none regulation? If there's a defect in the design or a defect in a warning, then there's strict liability--no specific negligence need be proven (I'm grossly oversimplifying, I'm sure).

|3.8.06 @ 1:40PM|#

qb: (One of my faves here!) Under my plan, the company is free to overestimate the risks as much as it wants. This solves the confidence problem. No trails? Just list every side effect known to man at 100% incidence. As the confidence increases, the risks slide down like a smooth morning movement.

t.: same way they did when they invented HIV during the polio trials -- use poor people outside the jurisdiction of the tort court. and make sure your old high school class stops wasting pigs so they are available for these trials as well.

|3.8.06 @ 1:41PM|#

They generally do (doctors get well-remunerated for being a part of a trial, even if it is only the big gala affairs they are invited to at the kick-off and close of the study. They generally want to make their customers happy, so they can get repeat business),

This is an unfair generalization. I'd like some examples, please. Also, I think you misunderstood what I meant by "they don't have financial stakes" in the matter -- that means they aren't supposed to be shareholders for the company they are doing trials for, and they are not supposed to be independently employed by the company. Despite your claim that researchers want "repeat business", it's the drug companies who need researchers, not the other way around. Researchers frequently work off of grant money, anyway.

although the use of double-blind conditions is specifically targetted at eliminating this problem, along with others. The double-blind study is a powerful tool, but it can be (and has been) manipulated. You can't prevent that, all you can do is make sure that there are consequences for those who do (manipulate, that is).

Even without a double-blind study, scientific researchers will eventually be exposed if they fudge their results. Science is not based on lies, and no respectable scientist (again, one who is at least not working for the corporation) will lie about results.

|3.8.06 @ 1:44PM|#

"but why not rely on products liability law instead of all-or-none regulation?"

How 'bout c., none of the above? At least the products liability law that currently exists. The 3rd restatement is a huge step in the right direction, but hasn't been widely adopted.

I'd rather just leave it in the general ballpark of Article 2, with perhaps a few pharmaceutical specific rules.

|3.8.06 @ 1:46PM|#

My understanding is that the evidence is decidely...um, mixed, in regard to HIV and polio. But I admit that I haven't followed the issue carefully, and that's really the sort of thing where patent attorneys have more expertise than professional scientists. So I'll just let Dave W. accuse me of being blinded and bribed by Big Pharma or whatever, in regard to that one.

As far as wasting pigs, yeah, God forbid that people taking a biology class learn anything about anatomy by dissecting a vertebrate. Real waste of resources there.

|3.8.06 @ 1:46PM|#

There are some diseases/injuried/etc where I wouldn't care if there were "side effects". If I'm in severe pain and my life is a living hell, do I really care if the drug that makes my life somewhat bearable will kill me in a few years or make me addicted to it? No.

And RC nails it: all these regulations do not help the little guy, they help make sure the little guy can't compete. I mentioned it on another thread, but most regulations created when our country was still young were to stiffle competition.

As for drug testing: I do believe that the testing is a worthwhile thing. We don't need snake-oil salesmen selling poison and calling it medicine. But do we really need a government agency for this purpose? As with most things, I'm thinking not.

|3.8.06 @ 1:50PM|#

Joke, T. It was a joke. The non-joke answer is that they should do patient and animal trials the same way they do now. My plan leaves the liability picture there undisturbed. Sheesh.

|3.8.06 @ 1:51PM|#

Dave W., it would be easier to tell the jokes from the serious statements if the serious statements weren't so bizarre.

|3.8.06 @ 1:56PM|#

In order to prevent the working of an estoppel, I must repeat that I do not know this Dave W. I am a ghost who was cast out of Heaven for bizarre statements and loitering. In life I ran a profitable insulin factory. Now I wander the Earth in search or truth, justice and consumer choice.

|3.8.06 @ 1:58PM|#

Hey, I'm all for wastin' pigs!


oh. You meant...biology. research.

|3.8.06 @ 2:03PM|#

Smacky,

It's not a generalization, it is a truth of the procedure, which says nothing about the character of the people involved. Can't give specifics without revealing more of my identity than I want to and possibly revealing info subject to non-disclosures I signed. However, I repeat that the clinical researchers (not the scientists/professors who work on the initial drug discovery) are VERY motivated to do clinical studies. It's essentially easy money, and a LOT of it. All I can do is tell you to watch your local 5 star hotel for study kick-off events. I've been to a few, and enjoyed the filet mignon, lobster tails, $500/bottle wines, etc. To call them extravagant would be to understate it.

So they do have an incentive to return positive results, and have in the past (which, BTW, is one of the big reasons why double blinds were mandated in the first place). You can search the FDA's registry for black listed physicians who have been barred from participation in future clinicals. We used to get a sheet (I think it was blue) that named all the people who the FDA had caught in fraudulent behavior, giving us notice not to employ them or contract with them.

"Even without a double-blind study, scientific researchers will eventually be exposed if they fudge their results"

True. The double blind just makes it harder to manipulate, among other benefits.

|3.8.06 @ 2:11PM|#

Another problem is that the mechanism that causes a reaction could require long-term exposure, or could take a long time to display itself. So a study could show a 1/1000 risk, based on the time periods reviewed, even if life-long use would lead to near-certain death. Anyway we are all facing certain death and the MS folks facing it a lot faster than most.

The FDA could help everyone out a lot more if they focused on how drug studies are done and how the results are communicated and left the decision to use a drug or not to doctors.

|3.8.06 @ 2:19PM|#

Update: Promising, risky MS drug clears hurdle

|3.8.06 @ 2:19PM|#

I've noticed that Dave W. likes to toss around accusations of people here being corporate shills.

I'd just like to ask this:

Hey, Dave, who can I talk to about becoming a corporate shill? I hear its quite lucrative. Know where I can send a resum�?

|3.8.06 @ 2:35PM|#

How about just showing that the incidence of the side effect has turned out to be 10X more prevalent than advertised? I say this in itself can meet the burden (assuming it is not credibly rebutted). Agree or disagree? Not clear from your last post, Timothy.

Ghost i think your understanding of how studies are done and how chance enters into them is keeping you from understanding Timothy's point.

If I understand you correctly, you believe that if the rate of negative side effects is 10 times greater than the rate given by the drug company that this automatically means that the drug company is lying. Thus you seem to be arguing that a lower burden of proof should apply in cases where the defendant is a drug company.

The problem with this is that even though the rate of incidence of negative side effects may be 10 times greater than believed, under some circumstances this can be due to a number of issues with the studies which are not the result of negligence or deceit by the manufacturer.

First off, it could easily be that chance skewed the results of the initial study. For drugs used to treat rare diseases it will often be difficult to create a large sample size for the initial studies. In studies with a small sample size it is easier for chance to play a significant role in the outcome of a study.

Hypothetical example:

Imagine drug X is going through its first stages of human testing. Drug X treats an uncommon condition, so the experimental group in the study is only 3000 people. Suppose the rate of negative side effects of drug X is 1 in 1000. We should expect 3 people in the study to show negative side effects. What if, however, only one person in the sample has the characteristic that causes the negative reaction? The observed rate of negative side effects would in that case be 1 in 3000, creating a 3 times difference between what the manufacturer will state when bringing the drug to market (based on the study which was skewed by chance) and the results we will see when the drug is used by the wider public. The question which we must answer is how significant is the difference of 2 people out of a sample of 3000. My knowledge of statistics is not advanced enough to properly quanitify that deviation from what is expected. That is why Timothy mentions the need for statistics experts to testify in the case. Most people don't understand statistics well enough to give a definitive answer.

Another reason the initial study might give different expectations of risk is a selection bias in the chosing of the initial sample. In many cases is could be an issue of self selection; those who are willing and eager to participate in the study may have different characteristics than those who prefer not to.

Another hypothetical example:

Recent research shows that there are genetic differences between "Black" and "White" populations which affect the way these groups respond to drugs for heart disease.

Now consider what this means for heart disease treatments tested before this was widely known. The standard understanding was that the biological differences between "racial groups" was negligable. Given that, a pharmecutical company may have reasonably not made an effort to match their experimental group's racial demographics to that of the population.

Now imagine this scenario. A medical journal publishes information about the upcoming start of human trials for a new heart disease treatment. Many people contact the company to take part in the initial trials. Given that black adults are significantly less literate than white adults ( 76% compared to 93% according to the NCES) the group of people who read the medical journal and apply to be part of the initial trials will tend to include fewer blacks than the population as a whole. If the drug being tested happens to present a higher risk to blacks than whites, then the study will tend to underestimate the risk for the population as a whole. When the drug is released the incidence of negative side effects will be much higher than expected (especially as blacks have a higher rate of heart disease). In this case should the drug company be assumed to be dishonest simply because the rate of negative side effects they gave was lower than the one observed when the drug became widely available?

The issue here is that for many reasons it simply is not possible to know the exact risk posed by a new drug until is becomes available to the public. Even in cases where the actual risk is many times higher than the initially reported risk, there may not be a legitimate case to seek damages against the drug companies. That is why we need a legal procedure to bring to light and consider all the relevant facts in the case at hand. That is also why the plaintiff must provide evidence of deception by the company.

Consider the alternative that you provide: If the risk observed after approval is "x" times higher than the estimate, it is assumed that the company is engaged in deceptive practices. Lowering the bar this far on damages will make is financially costly for companies to develop drugs for rare diseases, as they will not be able to create studies large enough to rule out the possibility of chance corrupting their results. That means people with rare diseases never get drugs to alleviate their suffering.

There are many considerations coming into play here, and i think that "just showing that the incidence of the side effect has turned out to be 10X more prevalent than advertised" is definitely not a high enough bar to set.

|3.8.06 @ 3:11PM|#

Hey, [mystery person], who can I talk to about becoming a corporate shill? I hear its quite lucrative. Know where I can send a resum�?

Upon reflection, I think Joe nailed it last time the issue came up. It is not merely disclosures about your stock portfolio (tho this is always a good sign and should probably be considered de rigeur). It comes down to your intellectual honesty. Joe pointed to Sullum as someone who makes his sponsorships clear, but also deals with his client-related and investment-related issues with some intellectual honesty.

I don't have a bright line test for intellectual honesty. Its totality of the circumstances, know it when I see it, etc., etc. I believe with Sullum he will often rhetorically adopt the positions of his opponents, but show how their positions are inconsistent, even if we take their concerns as potentially true. Even if I believed that 2d hand smoke causes cancer, I don't see how . . . That is intellectually honest.

What would have been intellectually honest is for Bailey to say, "yeah, I can imagine a drugmaker underestimating the risk and here is how I wish that situation gets handled . . ." Instead he responded by saying (if I understand him correctly): (1) only qualitatively missing a risk is bad, but not underestimating a known risk; and (2) the risk in this case has not been underestimated. Number (1) maybe addresses the issue, but what a stinky answer. Number (2) is just plain evasive. And this is in keeping with his other threads. Shill.

Compare that to my approach. I am supposedly anticorporate, but there I am proposing that we GET RID OF THE FREAKING FDA!!! No predictable pattern. A real and comprehensive framework for protecting consumers without gov't regulation. An accountability scheme where drugmakers can make or sell any drug, tested or not, WITH NO LIABILITY WHATSOEVER so long as they stand by their risk assessments. Balanced analysis. Sullumonic. Unshill.

Final note: the stuf in caps is that way because these are the parts where I am going against my perceived biases. Key!

|3.8.06 @ 3:16PM|#

If I understand you correctly, you believe that if the rate of negative side effects is 10 times greater than the rate given by the drug company that this automatically means that the drug company is lying. Thus you seem to be arguing that a lower burden of proof should apply in cases where the defendant is a drug company.

No. I am saying that drug companies should build these uncertainties into their published risk numbers such that actual risks are highly unlikely to exceed those actually observed. How? However they want so long as they realize that they are gonna be the ones paying the price if they (greedily) underestimate.

If the maker really has no probabilistic handle on the long term effects of a drug, then they should say 100% because they have no basis for saying it is any lower. If they get a basis and lower the number then money goes where mouth is. That is my plan.

|3.8.06 @ 3:29PM|#

If the maker really has no probabilistic handle on the long term effects of a drug, then they should say 100% because they have no basis for saying it is any lower.

If the maker really has no probabilistic handle on the long term effects of a drug, then they should say 0% because they have no basis for saying it is any higher.

|3.8.06 @ 3:43PM|#

Fair enuf, but that sounds like a hard sell politically to folks who are not yet Reason readers. How many votes have you Reasonoids? If I am going to trash the FDA, I don't want to be out on a limb without public support. I still remember what happened last time I dabbled in healthcare reform.

nmg|3.8.06 @ 3:45PM|#

Good grief how dense do you have to be not to understand the issues behind free choice vs. liability?

If a drug company says we have found a 1 in 1000 rate of adverse affects, and we have studied 1000 patients, that's not a very good study. We can assume this drug is new, understudied, and risky. You can take it if you think it's worth the risk, perhapd because you suffer horribly from a debilitating case of MS and it's your only hope, or you can wait until they have more data.

If they *lied* about what went into their study, then that is fraud and you have a good case. If they present the facts openly, you have no case and no moral position either. They offered you a free and informed choice to take a risk in the hope that your life would improve.

You get to make the call.

It's really not that hard to understand? What's wrong with these people who can't get it?

nmg

R C Dean|3.8.06 @ 3:48PM|#

I may not always agree with R C, but I have never, ever gotten even the faintest hint of shill from him.

In my dreams, t. Poppa's got a mortgage to pay, and corporate green folds like any other.

FWIW, my investments are mostly in boring broad market funds, with a sprinkling of insanely volatile energy, financial, and retail sector micro-caps. That were kicking ass until the middle of last week.

He probably force-feeds corn syrup to kittens, too.

Of course not. Kittens are too chewy. Now, a nice corn-syrup fed puppy sizzling on the grill - heaven!

nmg|3.8.06 @ 3:49PM|#

To put it more clearly.

Drug makers *cannot* underestimate the risk. They can only publish the results of the trials. It is up to others to interpret those results. If the drug makers lie or falsify data from the trials then yes, they are liable.

If you read the stats and like your odds, you have no cause for complaint unless the stats were lies.

nmg

|3.8.06 @ 3:50PM|#

If a drug company says we have found a 1 in 1000 rate of adverse affects, and we have studied 1000 patients, that's not a very good study.


That's not the case in the Tysabri studies. There have been thousands of patients involved in natalizumab trials. The 1 in 1000 risk factor is only a calculation based on all of the gathered data from various, multi-site trials that involved many more patients.

|3.8.06 @ 4:03PM|#

Actually, about 7000 patients in total were exposed to Tysabri, including clinical trials AND the 3 months it was on the market. The 2 patients who died were part of long term clinical trials, BUT WERE ALREADY SERIOUSLY IMMUNOCOMPROMISED by earlier therapies. Stats are stats, and if they are presented in good faith after extensive due diligence, the manufacturer should not be liable.

|3.8.06 @ 4:04PM|#

Drug makers *cannot* underestimate the risk. They can only publish the results of the trials. It is up to others to interpret those results.

Well, somebody has to interpret those results so that we can do the kind of risk evaluation we all think patients should be doing (as nicely outlined by Bailey). So who does the interpretation? I see three choices:

1) consumers

2) manufacturers; or

3) the government.

Right now the government is doing it. They suck at it for reasons I don't think I need to explain here.

The consumers did it up until the invention of the FDA. The historical experience was that consumers sucked at it and they weren't improving over time either. There is a special madness here at HnR that the consumers would suck less now at this interpretve process, but enuf ppl have enuf commonsense such that we will never go back to this way of doing things so it isn't even worth discussing.

That leaves the manufacturers. They can calculate confidence levels. If they do a study of 1000 patients, then they have a pretty good idea of what that study tells us (if anything). There is a whole professional science. F stats and p stats and intervals. A science that people employed by the companies in Bailey's portfolio already know thru their work with the evil regulators. The choice is clear.

|3.8.06 @ 4:06PM|#

On the subject of physician compensation for clinical trials (sorry if it's already been said above and I missed it)

When a particular trial is complete (and the patent ducks are in a row) the results of said trial are published. The bigger the trial, and the bigger the results, the more prestigious the journal the study is published. For instance, the New England Journal of Medicine.

Physicians working at research institutes and universities generally operate under the "publish or perish" doctrine. Thus a physician-scientist must not only resist direct financial "bribes" (in whatever form they are delivered), she must also resist the urge to skew results in order advance a career

nmg|3.8.06 @ 4:14PM|#

Ghost,

I think you forgot the best option: "Consumer Reports".

nmg

|3.8.06 @ 4:21PM|#

Clarification: "Consumer Reports" was subsumed in my treatment of the consumers option. Like I said, special madness.

nmg|3.8.06 @ 4:25PM|#

So when you say the "choice is clear", do you mean the manufacturers themselves? And you think that's a more reliable option than a disinterested third party with a profit motive to maintain a reputation for accuracy and integrity?

How?

nmg

nmg|3.8.06 @ 4:42PM|#

Also,

There is a special madness here at HnR that the consumers would suck less now at this interpretve process

Actually, it requires a special madness to think otherwise. Not because individuals are any smarter, but because mass communication, information dissemination, general consumer awareness are so much better. This is so patently obvious it hardly needs to be stated.

In 1906 there was no internet, no consumer reports, and everything was opaque to the consumer.

2006 is a fundamentally different landscape for producers and merchants. It's not even close.

nmg

|3.8.06 @ 5:02PM|#

What about the Underwriters Laboratories model?

|3.8.06 @ 5:27PM|#

any model is great (manufacturer liability, insurance, UL, Consumer Reports) so long as there are some teeth making sure that risks do get estimated in a form useful to the average patient. By "teeth" what I mean is a solvent party ready, willing and able to back the consumer-meaningful risk assessment with big payments under the contingency that the risk has been underestimated. I understand that the Big Pharma lawyer game is to make sure that either there is no liability, or, failing that, that the liability can be taken by some third party who can be isolated from their bottom lines and declare bankrutcy without having anybody dipping into the lucrative profits. My game is to make sure that "teeth" is teeth. Ergo, no Consumer Reports, no UL. They can do the work, but the liability ought to remain firmly lashed to the people making the profits. Anything else just ain't accountability.

nmg|3.8.06 @ 6:16PM|#

They can do the work, but the liability ought to remain firmly lashed to the people making the profits. Anything else just ain't accountability.

I guess I don't understand your position.

What holds a drug maker accountable if it's not a product liability suit or the FDA?

nmg

fyodor|3.8.06 @ 6:59PM|#

Ghost,

If you give drug companies the incentive to over estimate the risk, consumers will just take that into account and compensate by judging the risk to be less than what the companies say, and you'll be right back where you started! You can't legislate truth! Now, if you want to fund a study to compare how drugs do once they are used in the population compared to how they fared in published clinical studies, you will be adding to our body of knowledge. As it is, people probably expect risks to be understated for the same reasons YOU do. You're not so special, after all! But trying to force drug companies to err on the "safe" side will only force people to interpret what they are being told differently! Perhaps your flaw is when you give three choices for interpreting the data. Consumers ALWAYS interpret whatever information is at hand. Trying to game that information won't change that!

All that said, your proposal may be better than the status quo. But mainly only because it gets rid of the FDA! Then consumers will have a legal avenue of using their own interpretations!

|3.9.06 @ 8:30AM|#

Calliope,

But you're missing the fact that most of the physicians involved in clinical studies are PCPs, not university/professorial types.

If you want a specific (although this isn't an example that arises due to a desire to please the pharma sponsor) example, there are many published cases of doctors "inventing" patients to enroll in the study (to get more $$$). CRAs - a group of highly paid professionals whose ENTIRE JOB is to audit the records of physicians in clinical trials - can give you names on that issue.

Again, it is an entirely different issue when you're talking about drug discovery, or even phase 1 clinicals. There is very little incentive for fraud there. Phase 3 and some post market stuff? Massive incentives, which is why the FDA requires CRAs and watches them relatively closely. Another good example that shows how biased results can occur from these incentives are the post-market studies on the 2nd gen beta-blockers. The mfrs published a large study they claimed showed that 2nd gen was more efficacious than 1st gen. Treating doctors relied on that info for several years until an independent group re-sifted the data and showed that difference was statistically insignificant.

Like I said, you can't prevent this stuff from happening, just like you can't prevent crime. All you can do is make sure there are consequences for those who engage in such practices, and do your best to catch them at it.

|3.9.06 @ 10:31AM|#

Like I said, you can't prevent this stuff from happening, just like you can't prevent crime. All you can do is make sure there are consequences for those who engage in such practices, and do your best to catch them at it.

If you toss the responsibility to the drug consumers then you don't have to make any consequences at all for bad data. Sure, have a double blind study if you want to pay, but, really, who can afford that?!?!? Better to just have "caveat emptor" engraved on lots of headstones and wait for everybody to get what they deserve for being gullible. Send all the lawyers back to school, mortician college that is. That's HONEST work. That's how we did it when lifespans were lower.

|3.9.06 @ 11:21AM|#

Quasibill

I have to disagree with you regarding the types of physicians who carry out phase III and beyond trials. I invite you to visit:

http://content.nejm.org/

The first research article in this week's issue is a comparison between two FDA approved anti-virals for managing Hep B. Obviously there are enormous financial implications if the results result in widespread adoption of one drug over the other (only the abstract is avail. online) Check the authors, half of them are pharma (BMS) but the other half an independent scientists

If you look through past issues, you find the same thing, industry affiliated trials of drugs that have a mix of academic and industrial investigators

I am not suggesting at all that this particular study was biased towards anything except finding the best treatment for a fairly common disease. Although I will point out that the drug found to be superior is held by BMS

I think that by and large, the academics that carry out this work are completely impartial - fraud allegations are devastating to a career

However, academic physicisns are usually highly intelligent AND highly ambitious. When those two traits conflue (is that a word?) a potential exists to skew results to a more favorable end. this is no different than the upper echelons of finance and politics

Do you think if the results of the trial were: lamuivadine and entecavir are more or less equivalent in treatment of Hep B, the article still would have appeared in NEJM, or, would it be more likely in JAMA or even an infectious disease specialty journal?

fyodor|3.9.06 @ 11:22AM|#

If you toss the responsibility to the drug consumers then you don't have to make any consequences at all for bad data.

Not true. Intentionally trumpeting bad data is still fraud, at least if you can prove fraud in court. Now, that may be too high a bar for you, but that's what this is all about. You're willing to trash due process in return for a perceived benefit in return, and most of the rest of us are not.

|3.9.06 @ 11:29AM|#

In view of Fyodor's comment, change to:

--don't have to make any consequences for bad data unless a drug consumer can prove drug manufacturer intent in a court of law -- in other words no consequences for bad data as a practical matter, especially when intent is taken to mean intentional mistakes as distinct to reckless, negligent or easily concealed ones.--

|3.9.06 @ 11:37AM|#

Oh great, my editor doesn't know to from from.

|3.9.06 @ 11:42AM|#

I think that by and large, the academics that carry out this work are completely impartial - fraud allegations are devastating to a career

Yeah, just look what happened to that guy in South Korea once he made his fraud big enuf to detect. You can only get away with so much.

|3.9.06 @ 11:52AM|#

That was certainly a watershed instance of fraud - which I suspect is rare enough in drug trials to be non-existant. My suspicions are more towards actions like de-enrolling or post-excluding partcipants for various "reasons" and similar shades of gray activity.

Given the types of collusion that we see in other markets, I would be shocked to find that none of it has ever existed between a sponsor and executioner of a so-called "double blind" study

|3.9.06 @ 11:55AM|#

I think the South Korean researcher was a bad apple. Like Charles Graner in a way, but a bit less evil and a bit more powerful. An isolated case, not representative of a larger culture of anything bad.