The Deadly Corruption of Clinical Trials

http://www.motherjones.com/environment/2010/09/dan-markingson-drug-trial-astrazeneca

by kumarski • 13 years ago

111 59 13 years ago

Regulatory capture. Used to be a term we used to criticize corrupt developing countries. Nowadays it appears to be at the root of so many problems facing America.

cynicalkane• 13 years ago

What are you talking about? Regulatory capture assumes a working rule of law, and describes a process by which people subject to the rules of the game try to rewrite them. In corrupt developing countries the power owns the rules outright, and regulatory capture is no longer a process. I guess you could call it "trivial regulatory capture" but it won't get you any Nobel Prizes. The theory was developed by Chicago School economists who liked to criticize big government.

Regulatory capture has always been a bug of advanced, sufficiently honest societies. This is important because it tells us that well-intentioned rule of law is insufficient for running a democracy if you're not paying attention. I don't think anyone at the FDA wakes up and says, "I think I'll spend today getting in the way of good medical care and helping big pharma make money."

mschuster912• 13 years ago

The problem is the privatization of critical infrastructure, a cancer spreading all over the world.

To quote the last sentence of the article: "Many academic physicians are required by their universities to generate a substantial portion of their salaries by obtaining research grants."

Healthcare, basic transportation, energy and water distribution as well as basic telecommunication infrastructure must be in the hands of the state. The free market has massively failed in each of these areas, everywhere in the world. Just look at the US electricity grid, the Berlin water privatization disaster (less service, more expensive), the US (and partially, German too) broadband Internet service, the Berlin S-Bahn problems, the (recently HN frontpaged) problems in the US healthcare system...

Why can't we learn from the countless mistakes made all over the world?!

lemming• 13 years ago

For anyone interested in the generally fucked up state of clinical drug trials, I highly recommend Ben Goldacre's excellent book Bad Pharma. It'll make you angry.

Create3• 13 years ago

Bad Science also does the job.

The Emperor of All Maladies gives a candid and almost frank overview on how clinical drug trials are performed "from the inside".

Aspirin the Story of a Wonder Drug is also a nice account of "drug development" in practice.

mschuster912• 13 years ago

This makes me feel so sad.

I think the biggest lesson to learn from this case is that there has to be a independent entity in double-blind studies which knows the relation between test subject and the substance and has the power to interrupt or terminate a study when people's lifes are at risk.

Again, another fine example of how capitalism does not take unforeseen issues into account - to refuse to drop out Dan of the study program is "manslaughter by neglicence" for me (not by malice, as the responsible actors could not have definite knowledge that Dan would kill himself).

protomyth6• 13 years ago

You site capitalism, when all rules governing a clinical trial are strictly enforced by the FDA which is a government entity. The FDA has study rules and the entity conducting the study is not the drug company. It is an independent entity (for example http://clinicaltrials.mayo.edu) paid by the drug company under the supervision of the FDA (in the US).

logjam• 13 years ago

That is some really utterly fascinating spin right there.

Next up: criminals are not really responsible for crime, since "government entities" like municipal police forces supposedly "strictly enforce" criminal conduct.

I'm going to take a wild guess here and assume you've never had anything whatsoever to do with a clinical trial, because the modern FDA has almost nothing to do with the way a clinical trial is conducted.

A pharma company (the capitalists) writes the protocol the way they want, and hires a university/researcher to run the study. They file a simple two page form with the FDA to apply for an "indication" to study a treatment. The university/researcher gets the university IRB to sign off on the study protocol. Sometimes, if the IRB is any good at all, they'll tweak the protocol to try to protect the research subjects. But guess who picks the IRB members? The university/researcher making money off the capitalist pharma company. Guess how that influences an IRB.

The FDA has no input. There is no supervision in any meaningful sense by the FDA. The capitalists - researcher and sponsor - run every aspect of the study. Each decision is theirs.

The decisions spelled out in the article were made by the capitalists - the pharma company and researchers.

Barring some complete misstep like all the active arm study patients suddenly dropping dead, the FDA will eventually (read: years after the study is completed) examine the data (typically highly massaged by the capitalists) from the study to make a determination on whether to approve a medication for sale.

But pretending that the FDA (especially after years of being stripped of regulatory powers by, you guessed it, capitalists who buy and sell legislators) has much of anything to do with the way a clinical trial is conducted is complete bullshit.

capnrefsmmat3• 13 years ago

Most clinical trials are now conducted outside of the US.

ScottBurson2• 13 years ago

I would go farther than that. I think the FDA has to interpose itself between the drug developer and the entity doing the study. The developer should provide a sufficient amount of the drug to do the study, and write a large check. Everything else should be handled between the FDA and the university or whoever it is doing the study; the latter should have no contact with the developer. Ideally, the drug developer and the university should not even know each other's identities. The university would be answerable to the FDA for the quality of its work and to no one else.

I don't see any solution short of doing this. A study can't be considered truly blind unless the study design itself is done by researchers who do not know the "desired" outcome.

xijuan• 13 years ago

This article has left me speechless.. It reminds me of what my stats prof said about this kind of research..Researchers doing this kind of research have to believe that they will find the results the pharmaceutical company wants before they conduct the study. In other words, they are not testing if the drug works; they are just showing that the drug works.

tomjen3• 13 years ago

Thats bad enough, but they essentially forced him (at the threat of imprisonment) to participate in a trial.

refurb7• 13 years ago

Most of these clinical trials are triple-blinded. The doctor, patient and person analyzing the data don't know who took the drug and who took a placebo.

How do they bias the results in that case?

DanBC4• 13 years ago

Not all studies are effectively blinded, nor effectively randomised.

Merck-funded researchers trivialised the problems of Vioxx. Those researchers were not evil or stupid. Subtle influences are hard to spot in yourself.

When you look at the data there's a remarkable gap between independently funded studies and company funded studies. It's very clear that studies funded by companies more often show benefits and less often show harm. (There are several published studies on this.)

And even if the blinding and randomising is effective there's no requirement to publish. Thus, a company can run ten trials and only publish the one that's positive.

ChristinaM• 13 years ago

By asking the right questions. In this case by aiming to find no significant difference between the 3 drugs under study then making sure the sample size was too small to find a difference no matter what.

ScottBurson2• 13 years ago

In some cases, according to the article, by manipulating the dosages. The drugs they don't want to do well are given at either too low a dose, so as to be ineffective, or too high, so as to have an excess of side effects.

capnrefsmmat3• 13 years ago

There are many choices to make when statistically analyzing the data, and it is not terribly hard to play around with choices until you get a favorable result. For example, one common problem is truth inflation: clinical trials are conducted with too-small sample sizes, so they can't detect the result they're looking for. If you run enough of these, the only statistically significant results you will get will significantly exaggerate the true effects of the drug.

Some explanation and citations here, along with many other statistical tricks:

http://www.refsmmat.com/articles/statistics.html#truth-infla...

There's also a well-known problem called "vibration of effects" where you can play with controlling different variables, using different analyses methods, and so on, resulting in an estimate of a drug's effects that oscillates by a factor of two. Look for John Ioannidis' paper "Why Most Discovered True Associations Are Inflated".

nodata• 13 years ago

Ben Goldacre has been pushing to fix this.

See his campaign site: http://www.alltrials.net/

His blog/news site: http://www.badscience.net/

His Ted talk: http://www.ted.com/talks/ben_goldacre_what_doctors_don_t_kno...

pygy_• 13 years ago

The problem goes deeper than that. "Healthcare industry" is an oxymoron. The incentives of pharmaceutical firms and patients are not aligned, and will not for the foreseeable future (not that I have a better model to offer :-/).

danpalmer• 13 years ago

One suggested solution is to remove all patenting on drugs. That way, competition will force prices down, and result in two things. In countries with national healthcare such as the UK, the NHS bill could be drastically reduced, and the government could give grants to drug companies for R&D purposes. In countries with no healthcare system like the US, the general population will have much more money, and so taxes could be increased (hopefully not all the way back up) and grants could be given again.

For those who think that drug companies need high prices for recouping R&D costs, their marketing budgets are far greater than R&D budgets, so this isn't really true. In the UK they aren't allowed to market to end-users, so the advertising costs are generally much lower here.

refurb7• 13 years ago

I highly suggest reading some of the critical reviews of Goldacre's work. He gets a lot of things wrong and some of his ideas just simple won't work with the current system.

He's writing a book about an industry that everyone loves to hate. Few people question a book when it agrees with their preconceived views.

DanBC4• 13 years ago

Could you recommend some of the critical reviews?

I've seen some critical reviews that are just nonsense. I don't want to spend time wading through nonsense, but I realise that maybe he's made mistakes, so I'd welcome curated suggestions.

jdietrich• 13 years ago

The critical reviews are all irrelevant nit-picking. The vast bulk of Bad Pharma is an attempt to explain to the layperson the implications of a very dry but very important problem. He makes it clear that most of the book consists of illustrative anecdotes.

His core argument is extremely simple: If you conduct any sort of medical research, you have a moral duty to publish all anonymised data associated with that research, promptly and in full. There is absolutely no reason why that should not also be a legal duty.

tokenadult2• 13 years ago

He gets a lot of things wrong

I'd like to see specific examples of issues that Goldacre gets wrong, as I see his work recommended by a lot of astute doctors and medical researchers, and I'd especially appreciate you making the effort, for the community here, to link to the "critical reviews of Goldacre's work" that have public-visible links that you most recommend. I'm willing to hear you out, but please give those of us reading along in this discussion more to follow up on.

wpietri4• 13 years ago

You know, I am a big fan of capitalism. Having started 5 companies, I'd have to be.

But MBA-style capitalism is a technology for producing short-term wins and substantial profits next quarter. If you also care about long-term wins or negative externalities, it's an insufficient technology to apply on its own.

And of course, we hackers know that. Jeff Bezos has been telling the quarterly-profits people to go fuck themselves for 15 years. And he has been right to do so. AWS, Amazon Prime, the Kindle, and now their local delivery stuff: all impossible to achieve if you're more interested in beating quarterly profit predictions than reshaping the world.

Health care is another area where it doesn't work so well. Consider, for example, these graphs: http://www.washingtonpost.com/blogs/wonkblog/wp/2013/03/26/2...

I should note that allowing pharma companies to advertise to the public is a relatively recent change. 1997, I think. And the US is one of only two countries that allow it. We could change it back.

protomyth6• 13 years ago

Its not really the advertising that is a problem. That is the symptom of the short time frame most drugs need to pay for themselves. Quite a lot of them only have 3 - 5 years to make the money back.

So, either we figure out a way to make drugs cost less or we extend the timeframe for profit. Killing the advertising just means the number sold will be less and much higher for those buying it.

richardlblair3• 13 years ago

"Its not really the advertising that is a problem. "

Read this book: http://www.amazon.com/Selling-Sickness-Pharmaceutical-Compan...

It is an advertising problem. Big pharma makes massive profits, so you can't cite profits as a source of justification for advertising drugs to people who don't need them. Maybe you should have taken 30 seconds to google GSK's profits last year, or AstraZeneca who had a bad year last year but still make 2 BILLION dollars.

It goes beyond selling drugs to people who don't need them though, they over exaggerate and create new attractive names for illnesses. "Social Anxiety", for example, is not a thing. It's actually called "Social Phobia". A research study showed that marketing "Social Anxiety" yielded a more positive result, so big pharma called it just that. I'll leave you to figure out how much big pharma made off of that little stunt.

kumarski2• 13 years ago

How much money is the real question? We don't know the full costs of researching and bringing a drug to market? I've never seen a study that outlines these costs in detail.

wpietri4• 13 years ago

If they are doing studies to generate marketing claims, then advertising has to be part of the problem, in that it's a major vector for marketing claims. There could be other vectors, and I would be happy to eliminate those as well.

The total amount of money spent on pharmaceutical marketing is enormous. We would be much better served if doctors could pay for subscriptions to research companies that only take money from doctors.

tel2• 13 years ago

Note that this is already required by law for FDA regulated drugs and devices: http://clinicaltrials.gov

capnrefsmmat3• 13 years ago

These requirements are most often ignored. The law requires that trials be registered and their results published within a year of completion, but most registered trials never have published results:

http://www.bmj.com/content/344/bmj.d7373

The results are simply buried if they're not sufficiently favorable. There have been a number of meta-analyses which dig up unpublished results (which is very difficult) and demonstrate that medications no longer have statistically significant effects when you include the unpublished data. For example:

http://www.bmj.com/content/341/bmj.c4737

(Both these papers are open access, so you can read them if you'd like.)

tel2• 13 years ago

Cool, thanks for that research. I'm very interested in this problem today.

tokenadult2• 13 years ago

"September/October 2010 Issue"

The case reported in the Mother Jones article happened at my alma mater, a state university with many biomedical research studies going on at any given time. The article author later published a follow-up piece

http://www.thehastingscenter.org/Bioethicsforum/Post.aspx?id...

in another publication. Two articles by different authors in local publications last year

http://blogs.citypages.com/blotter/2012/11/dan_markingsons_s...

and just last month

http://www.minnpost.com/second-opinion/2013/03/controversy-o...

provide further updates about this particular local case.

A group blog post about medical research from 6 January 2012, "Why Do We Really Need Clinical Trials?"

http://www.sciencebasedmedicine.org/index.php/why-do-we-real...

examines some of the trade-offs involved in the still flawed, but still necessary, process of conducting clinical trials of proposed new treatments. Human beings who propose new treatments often have a bias to suppose that the new treaments are more safe and more effective than previous treatments--especially if they have a financial interest in the new treatment. People who have a financial interest in other treatments, or in not seeking effective treatment at all (e.g., faith healers) may have a bias that runs in the other direction. The way to check the safety and effectiveness of a treatment is to run an honest clinical trial with adequate justification from known science and sufficient statistical power. But keeping the trial honest involves human beings disciplining themselves not to jump to conclusions. As Richard Feynman said to graduates of Caltech: "The first principle is that you must not fool yourself--and you are the easiest person to fool."

The Deadly Corruption of Clinical Trials

Related Stories