[ramanujan]

The issue is not really whether these treatments work or not. Every new treatment is by necessity unproven and more risky than an established one. Some people are early adopters or terminal patients with high risk-tolerances; others would rather suffer for a while than chance a drug that might make things worse.

Both of those are fine as preferences. The trouble arises because every US citizen is forced to have the global minimum of risk-tolerances across the population. It would be as if you could not try a development version till it became user-friendly enough for your grandparents. Specifically, in the US, you can't be an early adopter: the FDA does not allow citizens to opt-out unless they leave the borders of the USA. Indeed, in Cowan vs. US (1998) it successfully sued in federal court to prevent a dying AIDS patient from trying an experimental drug:

http://www.leagle.com/xmlResult.aspx?page=4&xmldoc=19981...

  Plaintiff requests that Dr. Davis be authorized to inject 
  Plaintiff with the with the experimental goat neutralizing 
  antibody drug [1] and that the FDA be enjoined from 
  interfering with Dr. Davis' treatment of Plaintiff. ...

  The Court is sympathetic to Plaintiff's situation. 
  However, the law is very clear, and under the current 
  statutes and regulations, Plaintiff's physician may not 
  administer the goat neutralizing antibody drug absent 
  prior approval of the FDA. In Court, Plaintiff argued that 
  he should have the right to take whatever treatment he 
  wishes due to his terminal condition regardless of whether 
  the FDA approves the treatment as effective or safe, and 
  that to prohibit him from taking the treatment he wishes 
  violates his rights under the US Constitution. ...

  This Court is in no way criticizing the intentions of 
  Plaintiff and his physician or the potential effectiveness 
  of the proposed treatment. Plaintiff's physician should 
  pursue approval of his Investigational New Drug 
  application as quickly as possible. Plaintiff's doctor 
  must obtain appropriate approval through the proper 
  regulatory authorities. As much as this Court may 
  empathize with Plaintiff, the authority to provide some 
  type of exemptions for individuals such as Plaintiff rests 
  with Congress and not with this Court.

Plaintiff was denied. Plaintiff died. Who knows whether the drug would have saved him, but he wasn't given the chance to try. Occasionally, if you have tremendous political connections, you can get a waiver, like Fred Baron:

http://blogs.wsj.com/health/2008/10/17/lance-armstrong-and-b...

  “We did a safety review, consulted with experts on PML, 
  and worked closely with FDA to come up with a risk 
  management program that allowed us to bring it back on the 
  market in a way that limited its use,” a Biogen 
  spokeswoman told the Law Blog. The plan prohibits giving 
  Tysabri for unapproved uses.

  Biogen Idec is running an early-stage trial of the drug in 
  multiple myeloma, but Baron doesn’t meet the criteria to 
  participate.

  Baron’s a prominent donor to the Democratic party, and 
  many of his powerful friends, including Lance Armstrong 
  and Bill Clinton, made appeals on his behalf. And the 
  family agreed not to sue if anything goes wrong.

  Ultimately, his doctors at the Mayo Clinic worked directly 
  with the FDA to find a “legal basis” for giving Baron 
  Tysabri. The deal was announced on Baron’s son’s blog late 
  yesterday.

So, if you are wealthy you can travel outside the US to opt-out (though US-based companies will usually not administer treatment for fear of getting on the FDA's bad side). And if you are politically connected you can sometimes get an experimental treatment, like Baron.

But this is not really optimal. If you are an academic, you accept the concept of QC/quality checking, but you aren't stuck with just one journal to submit to. You can revise and resubmit somewhere else. And if you are an end-user, you don't have to take a reviewer's opinion into account when choosing between movies, books, bikes, or virtually any other physical good with star ratings on Amazon.com. Except for drugs. Then you, as the end-user, cannot opt-out and take the FDA's opinions with a grain of salt. In part this is because the FDA will sue you directly. In part it is because companies that even think of trying this route will get slapped by the FDA for trying to game the system, and subsequently find their approvals slowed or (nowadays) outright denied.

We need to carve out a jurisdiction where patients and entrepreneurs alike are free to take informed risks, recognizing up front that sick people do die in medicine, and also recognizing that society already allows people to take incredible risks in other contexts (joining the military, bungee jumping, walking tightropes). Whether that new jurisdiction is Singapore, or Estonia, or a seastead, or a medical cruise ship, or something else is to be determined. But that has to be the goal.

[1] In case a "goat antibody" sounds weird to you, it's a common thing in molecular biology. Google it, or see for example Thermo's web page: http://www.pierce-antibodies.com/custom-antibodies/goat-anti...

[ataggart]

Counterfactuals are hard, but I wonder if anyone has studied whether or not the FDA is actually a net saver of lives. It seems to me that every time the FDA approves a new drug that will save X lives a year, it has necessarily cost X lives every year that the drug was delayed in the mandatory approval pipeline.

If FDA approval was optional (perhaps requiring a big "NOT FDA APPROVED" label), would the number of people who die from taking unapproved drugs (and who otherwise would have lived) be more than the number who die from being denied a drug (and who otherwise would have lived)?

As I write this, it reminds me a bit of the IP issue, and whether the patent system really is a net benefit or not.

[ramanujan]

You might be interested in these studies. The second link is to a discussion of Andy Grove's editorial in Science calling for the FDA to only certify safety (rather than safety + efficacy + comparative effectiveness). But it's the third that gets at your question:

http://www.cato.org/pubs/regulation/regv27n2/v27n2-8.pdf

http://marginalrevolution.com/marginalrevolution/2011/10/and...

http://fdareview.org/harm.shtml

  The delay and large reduction in the total number of new 
  drugs has had terrible consequences. It is difficult to 
  estimate how many lives the post-1962 FDA controls have 
  cost, but the number is likely to be substantial; 
  Gieringer (1985) estimates the loss of life from delay 
  alone to be in the hundreds of thousands (not to mention 
  millions of patients who endured unnecessary morbidity). ... 

  If the U.S. system resulted in appreciably safer drugs, we 
  would expect to see far fewer postmarket safety 
  withdrawals in the United States than in other countries. 
  Bakke et al. (1995) compared safety withdrawals in the 
  United States with those in Great Britain and Spain, each 
  of which approved more drugs than the United States during 
  the same time period. Yet, approximately 3 percent of all 
  drug approvals were withdrawn for safety reasons in the 
  United States, approximately 3 percent in Spain, and 
  approximately 4 percent in Great Britain. There is no 
  evidence that the U.S. drug lag brings greater safety.

Ultimately this boils down to a classification problem. There will be type I and type II errors associated with any kind of centralized approval process. And when studied in its totality, there is quite a bit of evidence that the type II errors are predominating: good drugs being slowed or denied.

The only way to prove this definitively is a side-by-side experiment with a new jurisdiction in which patients and entrepreneurs alike are free to choose and the FDA has no power.

[Gormo]

> Counterfactuals are hard, but I wonder if anyone has studied whether or not the FDA is actually a net saver of lives.

When did utilitarianism become an acceptable justification for usurping the right of individuals to make their own risk-reward tradeoffs?

[gus_massa]

Even with careful studies, there have been some awful messes. One famous example is Thalidomide: http://en.wikipedia.org/wiki/Thalidomide (If you are not impressionable you can use Google images.)

Some drugs investigations are discontinued because they cause complications or deaths in the small test groups. Another important number is how many additional dead could have been if everyone could take unapproved treatments.

And most people would sign whatever the doctor tells them to sign, for example a "NOT FDA APPROVED"-waiver. People thrust whoever has a white coat and promises a cure. There has been problems with peopled enrolled in official experimental drug test, that didn't understand the details, but had signed just another "usual" form handled by the doctors.

[wissler]

It's impossible to compute the number of treatments that weren't even invented because the inventor saw the FDA barrier as too much trouble for him to manage.

The real question is the unpopular one: by what right does the FDA insert itself between doctor and patient, between drug manufacturer and doctor?

Ah, dogma. It's killing us.

[jlgreco]

It is very hard to say that any patient should be allowed to opt-out of the FDA's regulation because almost without exception every patient that buys into snake-oil thinks it is a good idea that is worth a shot.

How do you distinguish between "this patient is making an informed decision to opt out and use this unapproved drug" and "this patient is an idiot who thinks that throwing money at a "doctor" to mainline bleach will cure his disease"?

[ramanujan]

Three points.

First, many (most?) medical advances seem crazy and kooky because they haven't been tried before. So it's not usually as obvious as "this is a proven scam" vs. "this really works". It's much more frequently "this is unproven". And like Barry Marshall's famous self-experiment with H. pylori, someone has to be first for it to ever get proven.

Second, when the government gets it wrong, it gets it catastrophically wrong. The USDA Food Pyramid recommending "6-11 servings of grain" is still being slavishly followed, and in the fullness of time we might well find it partially responsible for the epidemic of obesity and type II diabetes. The FDA is still doing Phase I/II/III clinical trials despite all the evidence in favor of adaptive trials. And tens of millions of people were irradiated by TSA x-ray scanners fast-tracked through the FDA approval process, scanners criticized by UCSF scientists, scanners which have now (finally) been withdrawn. These errors are magnified in impact because no one can opt-out, because a .gov has a bully pulpit, and because strong political incentives exist to silence criticisms.

Third, if people have the right to euthanasia, or the right to walk near bridges, I do believe they have the right to try what treatments they want. Frankly I don't consider someone else's medical affairs my business, anymore than I'd ask why they had an abortion. You can argue that vaccinations present a public health issue, and I might agree with you there. But otherwise this strikes me as a right to privacy and right to bodily integrity issue.