How the FDA Impedes Innovation

Mike Mandel  has a good piece, significantly published by the Progressive Policy Institute, on the FDA and innovation. MelaFind is a handheld computer vision system, database and expert system that helps dermatologists to identify which skin lesions should be biopsied for melanoma. The device works quite well but the FDA has deemed the device “not approvable” because (quoting Mandel):

  • The device did not do better than the experienced dermatologists in the study (“the FDA review team does not believe this is a clinically significant difference between MelaFind and the examining dermatologist”)
  • The device was tested on lesions identified by experienced dermatologists, not on the broader set of lesions that might be identified by “physicians less experienced than these dermatologists.”
  • The device did not find every melanoma in the sample (“Since the device is not 100% sensitive, if use based on the device’s diagnostic performance reduces the number of biopsies taken, harm could ensue in the form of missed melanomas.”)
  • The device was not demonstrated to make inexperienced physicians the equal of experienced dermatologists (“Currently, formal training is offered to physicians to become board certified dermatologist and thus be able to diagnose clinically atypical lesions. The FDA review team would have to compare this board certification training to that offered by the sponsor to those physicians operating MelaFind to determine if it is found adequate.”

Some of these complaints are legitimate, others not so much. But even if MelaFind is not perfect today nor appropriate in all circumstances it’s exactly the type of innovation that we should encourage. Devices such as MelaFind could not only improve medical care they can reduce costs and make good quality medical care more widely available in developing countries, for example, where experienced dermatologists are in short supply.

Most importantly, innovations get better over time. But if you impede the first generation the second generation may never come into existence and, as Mandel notes, no first-generation device could satisfy the FDA’s conditions. It’s like refusing to give the Wright Brothers a license to fly because their first airplane only flew for 59 seconds.

The signal that is being sent by the FDA impedes all medical innovation.

For more on the FDA see


People who boldly claim that "no one can halt the march of technological progress" don't seem to realize what a good job we're doing of halting it...

We're not halting effective care. Note that the device has to be held by a dermatologist. So approving this only means that your care costs more - because every dermatologist will have to have one. Why? Because it's another test they can charge money for. That's why costs are skyrocketing and why HCAs will do nothing to help.

MRIs are getting ordered every time possible, second tests, second opinions, extra surgery rooms, extra days in the hospital. The 'market' is structured so that the people who can afford care have to have the most thorough expensive care possible, while other people suffer and die. And every year, the care gets more expensive so fewer people can benefit. That's why our system has worse results with more cost. That's why America's economy is so inefficient and it's a drag on everything else that we do.

Such a device would save costs by simply being available to general practicioners and clinics who might then see if a specialist visit is really necessary for a suspect skin thingie. Savings of time and $$ for patient and better use of specialist time where really needed.

Intelligent tools will be needed to commoditize (erode the cost of) the expensive walled-garden health systems and arbitrary specialties we have evolved.

Are we still living in America?!!! All of you have missed the entire point of the article. I DON'T WANT THE FDA TELLING ME WHICH TREATMENTS I CAN CHOOSE TO HAVE, OR WHICH EXPERIMENTAL DRUGS I CAN PUT IN MY BODY! I AM NOT THE PROPERTY OF THE STATE! The FDA has done nothing but restrict freedom and progress, and is responsible for the death of millions of people. Noah, it sounds like you work for the FDA? Let me ask you, what drug or treatment have you ever invented?

Hey, if men were meant to fly, God would have given us wings.

Seriously though, this strikes me as being more to protect the dermatologist cartel than anything else.

That's the crux. A lot of healthcare policy is aimed at protecting interests other than the patient.

Yep, and this is true across more than just healthcare.

Unfortunately, rent-seeking has often been a wise investment, a lot better than that risky new technology that might not pan out.

I want my melanoma's to go undetected.

From the post: "The device did not find every melanoma in the sample (“Since the device is not 100% sensitive, if use based on the device’s diagnostic performance reduces the number of biopsies taken, harm could ensue in the form of missed melanomas.”)"

I want my family physician to use this equipment, even though the equipment maker has not tested whether the equipment can detect atypical lesions.

What was the FDA thinking?


Would you like me to comment?

"...reduces the number of biopsies taken..."

Well, at least the doctors aren't trying to create more billable procedures. Haha.

If the melanoma cannot be detected other than biopsy, perhaps that is what you want to choose.

Go for it.

I want my family physician to use this equipment, even though the equipment maker has not tested whether the equipment can detect atypical lesions.

Bill, what metric does your family physician use to decide whether or not to recommend you to a trained dermatologist?

My doctor uses the SIIK scale, which stands for "Heck if I know"

If the family physician cannot detect, and the equipment cannot detect, I am sure you cannot detect, unless you are a dermatologist.



You got it bass ackwards. People do detect them, then they ask their doctor to look at it.

There are many melanomas in places where 'the sun don't shine' so to speak, so waiting for a dermatologist to happen across that means you better be married to one.

Andrew': You're imaginary facts upon which you based the answer are inconsistent with this statement in the post: "■The device was not demonstrated to make inexperienced physicians the equal of experienced dermatologists."


General practitioners ALREADY refer people to dermatologists. They are using a false control.

Andrew', You seem to be shifting your position. First it is that melanomas where the sun don't shine can only be found by a dermatologist. Then you say that GPs refer persons to the dermatologist, so the machine may be unnecessary, but that was preceded by the post that said all melanomas were not found and no evidence was offered that the machine performed better than either the physician or dermatologist. The burden of proof is on the person seeking to get the equipment to displace something better.

Bill, the device isn't as good as an expert dermatologist. But it doesn't have to be, it only has to be better than the family physician.

Right now, the family physician, not an expert, looks at a lesion and decides to refer to a dermatologist or not. "It's probably nothing," he tells the patient. The patient may make an appointment in a few weeks, or not.

With this device, it's likely that more lesions would be looked at, not less. Family physicians would still be able to refer people to experienced dermatologists if they wanted a second opinion, but it would be quicker and cheaper to get a good first pass on things.

It wouldn't be too hard to tweak the device to get extra false positives instead of negatives, if you wanted to optimize it for that. You generally can trade off between Type I and Type II errors.

It's interesting that by setting the goal to be "replace dermatologist" the FDA basically ensures that that ultimate goal cannot happen.

John, You might want to look at the responses below to answer your argument on quality control grounds. Sanjay, Joe, etc. which have, in my mind, probably not yours, addressed them, and who am I to judge. But, as you seem to be interested in cost control, a new tool in the hands of a gp that 1) has many false positives 2) gives revenue per click or has its own payment has its own dynamics.

I do not know, and question whether you could know, whether your statement : "It wouldn’t be too hard to tweak the device to get extra false positives instead of negatives, if you wanted to optimize it for that." is true.

Do you have some inside information on whether this device is easy to tweak as between type 1 and type 2 errors, as you said?

I have my issues with the FDA but this criticism seems oddly misguided. I have used FDA-nonapproved technioques all the time. What the FDA is doing with approval is talking about how the thing is going to be used in patient care: a doc want to feed patient biopsies into this thing, then use the results to make diagnoses and prescribe a course of action. Based on the criticisms above it sound like the intended use is to bypass the pathologist. They looked at it, thought, nuh-uh, and the reasons sound good.

What the FDA nonapproval does not do, is prevent you giving the thing patient samples for research purposes and further developing it. It does turn off a potential spigot for funding. But if you want to see stuff like this developed, you make research dollars available though NIH or the inventors make the case to investors that they can improve the device to overcome all of the FDA's objections -- in some sense the FDA has helped that kind of development by giving the maker a clear benchmark.

Medical innovation is not really the main function of FDA. Validating it is, Sure, any validation is going to screen stuff out.

The issue is that the FDA is screening ENTIRELY based on effectiveness. And the FDA isn't even saying it's ineffective - it's just not as effective as the best possible alternative. But it's a lot cheaper. Why can't the FDA let us exchange reduced quality for reduced costs?

"Why can’t the FDA let us exchange reduced quality for reduced costs?"

It's worse than that. Cheaper and more often IS better. Noone is compelled to do anything based on information A that they know is not as good as information B. You can just take A and then go get B. But it's even worse than THAT! If a technology is not allowed to cash flow itself through the weening stage, it never has the chance to improve to be superior to the previous alternative.

That's the question, what is the box that the FDA is approving for? Could they just approve the device under the title "somewhat less effective, but more automated diagnosis procedure"?

Why does it have to compete head to head with something that it is not?

Because that would be RATIONING! Dum Dum Dum!

I know this was not a political post, but the subtext of these items is usually that big bad government interference from the liberals is spoiling the machinations of the free market. Of course, when the Democratic Party attempts to make an agency function smoother, especially in the healthcare arena, it comes under attack from the Republicans. The idea seems that be that if a government agency exists, then it must be made to function as wastefully and poorly as possible.

I think your statement that cheaper but less effective alternatives should be available in healthcare is now the default liberal position, and staunchly opposed by conservatives.

I think your statement that cheaper but less effective alternatives should be available in healthcare is now the default liberal position, and staunchly opposed by conservatives.

No. The default liberal position is that cheaper but just as effective alternatives should and do exist. The President's "blue pill/red pill" hypothetical was about the idea that two alternatives are just as effective, but one is cheaper. Almost nobody publicly supports the idea of cheaper but less effective alternative, because that idea is distasteful to too many people/voters.

Consider the Democratic Party response to the idea of using vouchers for Medicare, and the idea that people might choose cheaper but less effective alternatives. Also consider that liberals in Congress are also banding together to ensure that the IPAB doesn't actually force people to use cheaper but less effective alternatives.

Supporting cheaper but less effective alternatives is the province of healthcare policy wonks, but not most people and not most politicians. Politicians will always promise that it's really "cheaper but just as good."

Not sure you've got it straight there. Single payer countries have cheaper and better care. Most importantly, the health care costs are skyrocketing because of more expensive tests and procedures that are deemed more effective, but cost-effective is never considered.

Not sure you’ve got it straight there. Single payer countries have cheaper and better care.

mjw149, thank you for confirming my point, and disagreeing with bulldog. As I said, the liberal mantra is that we can have cheaper and better. Bulldog argued that the default liberal position was cheaper but less effective.

As we see, the actual default liberal position (and conservative position) is that we can have our cake and eat it too, that with the right change we can have both cheaper and better care.

No one argues for cheaper but even slightly less effective, outside of wonks.

"when the Democratic Party attempts to make an agency function smoother, especially in the healthcare arena, it comes under attack from the Republicans."

Not enough data points to confirm.

"The idea seems that be that if a government agency exists, then it must be made to function as wastefully and poorly as possible."

Now, here we do have enough data to confirm.

IPAB, comparative effectiveness, Medicare advantage, eliminating subsidies for employer sponsored insurance, etc. Plenty of data points. Conservatives scream "it sucks," liberals counter "we'll fix it," conservatives respond with "rationing!"

eliminating subsidies for employer sponsored insurance

You mean the thing that McCain suggested in the 2008 election, and Senator Obama responded with ads saying "McCain wants to tax employer healthcare FOR THE FIRST TIME EVER?"

I didn't realize that Obama was the conservative and McCain the liberal.

McCain suggested eliminating it and replacing it with Health Savings Accounts, which weren't sufficient. And liberals certainly have been shifted to the right by the past two years of Republican blockade. The Democratic Party's stance on healthcare, in particular, is basically center right, especially when compared to where it was during the campaign.

But that's beside the point and you know it. This isn't about scoring cheap political points. It's about pointing out that the conservative position in the US today is either a government function should cease to exist, or it should function poorly.

This isn’t about scoring cheap political points. It’s about pointing out that the conservative position in the US today is either a government function should cease to exist, or it should function poorly.

I'm sorry, but I find your second sentence entirely "about scoring cheap political points."

McCain suggested eliminating it and replacing it with Health Savings Accounts, which weren’t sufficient.

McCain suggested replacing it with a refundable tax credit that would have been worth more than the deduction for low income individuals, but less than the deduction for high income individuals. (Since deductions are worth more for people in higher brackets.) It's bordering on the dishonest to say that he suggesting replacing it only with HSAs if you don't mention the refundable tax credit.

I have to say that you seem intent on scoring cheap political points.

I think the Mandel piece is rather misleading.

First, he doesn't list all the problems the FDA had with Melafind - he cherry picks a few. Worse, he states those few as if the FDA demanded that each of those concerns was a sufficient ground for disqualification in and of itself.

The point of the FDA's review is not to determine that the device is perfect, but to determine whether MelaFind improves the current standard of care. In other words, if you give this to a competent doctor, is he going to be more accurate in deciding whether to take (expensive, invasive, potentially disfiguring) biopsies?

The FDA had a five page bullet point list of concerns with MelaFind. I'll just cover a few of the main concerns. (link at p. 44).

(1) The studies compared MelaFind's recommendation to the expert dermatologist's recommendation. MelaFind did not provide instructions on how the dermatologist should use MelaFind's results, nor did it provide any evidence of how dermatologists would use Melafind's results to provide a baseline. Thus, neither the FDA nor dermatologists using MelaFind if it were approved can determine whether the risks outweigh the benefits in actual clinical usage (the FDA is especially concerned about false negatives from the machine influencing doctors not to order a biopsy, but the flip side is also a problem). This is especially true since MelaFind only produces "positive" or "negative" results, nothing in between.

(2) The studies were all done on lesions which board certified dermatologists selected for biopsy or close scrutiny because of suspicion of melanoma. If you want MelaFind to be used by physicians rather than dermatologists, you need to show evidence that physicians can pick out the appropriate lesions, or at least show us that MelaFind is accurate if physicians use it on inappropriate lesions.

(3) The tests were all done on atypical lesions which a dermatologist screened as suspicious for Melanoma. MelaFind's indication that it can be used to screen "all atypical lesions" is too broad.

(4) The claim that the machine helps find early melanoma isn't supported by the evidence, as it has only been tested on lesions dermatologists already picked out as melanoma risks.

"The point of the FDA’s review is not to determine that the device is perfect, but to determine whether MelaFind improves the current standard of care. In other words, if you give this to a competent doctor, is he going to be more accurate in deciding whether to take (expensive, invasive, potentially disfiguring) biopsies? "

And the point everyone else is making is that that's a ridiculous standard. If it were 99/100 as effective but 1/10 the price, they'd reject it!

Medical innovation is not really the main function of FDA. Validating it is,

Except that the FDA does a lot more than validate. Sure, doctors prescribe drugs for known good off-label uses all the time-- but the drug still has to be FDA approved for something first.

Yes, you can take the route of the supplement market and refuse to mention any scientific studies or real claims. That's fine if all you have is fluff. But as soon as you try to actually marshal real statistics or science to back your claims up (instead of weaseling out of them in the fine print), the FDA will crack down.

"if you give this to a competent doctor, is he going to be more accurate in deciding whether to take (expensive, invasive, potentially disfiguring) biopsies?"

That's actually not the question. They compared the device against a dermatologist. If that is the "standard of care" (even granting that exactly matching that should be the goal) that is a false "standard of care" in the real world where most people don't actually go see dermatologists, at least until it is too late.

I was imprecise, it was a dermatologist.

The reason that the FDA did not consider physician screening is that the machine was only tested on lesions that expert dermatologists thought were potentially melanoma. The FDA wanted evidence either that physicians would (or could be trained to) select largely the same lesions the dermatologists did, or that the machine would not be tricked by lesions that the dermatologists excluded. Otherwise their testing would have had a significant sampling bias.

I actually think this is like refusing to allow the Wright Brothers a license to carry passengers because the first flight was only 59 seconds - which is actually pretty reasonable.

No, it's like refusing to allow carrying a pilot.

The high costs ensured by such policies are part of the rising cost death grip by which we can pretend that we can fill out a best-practices database without actually having any people fill in the blanks.

I suspect even the "off-label" loophole by which ALL the risk is transferred to the guinea pigs will soon be closed by our 'policy innovations.'

It's not a policy! It's one decision made by medical experts cherry-picked by an economist! There's no pattern here.

The rising health care costs are due to TOO MANY technologies getting approved, built then obviously the equipment HAS TO BE USED because doctors NEED TO MAKE MONEY to cover their costs. There's your crisis, and it's all due to a market mechanism. Which is why single payer systems have better outcomes for less cost. It's fundamentally a broken market.

In economics, a "need" to make money to cover costs is just a "want" to make money to cover costs.

Unless both blades of the scissors conspire so that the want becomes reality, it doesn't happen.

Just an FYI on thinking like an economist.

You might think the passengers would be able to consent all on their own...

For decades, the FDA has responded to criticism from congress and the media, to the exclusion of nearly all other factors. With the brief exception of the AIDS activist groups in the '90s, FDA never gets yelled at for withholding advances in medicine or medical devices. They do get yelled at if some rare side effect becomes visible post-approval. Always safer to withhold approval.

The actual alternative for comparison is not a doctor, it is the people who don't go to the doctor.

I agree with Sanjay, actually; however its meant to be used, you know that a machine like this would turn into a magic, infallible computer at the family doctor's office, and a lot of lesions would go undiagnosed.

A lot of lesions go undiagnosed anyway. To me, you're conceding the entire argument by talking about the family doctor's office as the place where it would be used.

So the device isn't as good as a trained dermatologist. So what? Most people don't go to a trained dermatologist without going to a family doctor first. If the device makes the family doctor better than nothing, it's of huge value, even if it doesn't make the family doctor as a good as a board certified dermatologist.

People are assuming that tons of people go to dermatologists.

Show of hands, who here has taking one of their bumps to a dermatologist for an EARLY detection?

Level set: The only reason false negatives are a problem ("This spot is NOT cancer") is because a visit to a dermatologist is so rare.

I am a physician entrepreneur and I have put on hold development of all projects that have to go through the FDA, as there is no winning path. Case in point: I sat for 5 years on a joint NIH-FDA panel trying to 'solve' the problem of inadequate pediatric drug availability (over 80% of drugs used in pediatrics have not been tested in children). There are at least 4 distinct groups in pediatrics (infants, toddlers, pre-pubescent children, and peri-pubertal children; each group metabolizes and uses drugs differently), thus the market for drugs is fragmented and one or two orders of magnitude smaller than adult markets for the same disease. Since the FDA is even more risk averse with children than with adults (and effectively assigns the risk of the status quo at zero), it sets up insurmountable barriers to pediatric drug development.

A development pathway that costs $1 billion per approved drug will never develop drugs that have potential markets of $20 million per year. Now the FDA is applying the same approach to iPhone apps and vitamins. Not only that, but they are pulling drugs off the market that have had decades of safe use because 50 years ago they did not follow the current development pathway. I am waiting for the FDA to outlaw table salt because it is 'clearly' a toxic product that causes high blood pressure (in around 10% of the population).

You should read the MSDS of table salt sometime.


It is easy to denigrate lack of perfection, but let's look at how it could actually be used:

Current practice--100 patients, 90 not screened (doctors are busy, only screen skin lesions in a complete physical or when patient brings it up during the visit). Of the 10 screened, 5 go to the dermatologist (maybe 10% of those from the complete physical, and 100% of those who bring it up--most primary care docs are not comfortable with skin lesions, and for liability reasons will just refer if there is an issue). Assume the dermatologist is 100% correct ;-) and primary docs 95%; end result is about 10% correct diagnoses, less than 1% incorrect, and 90% unknown.

With new machine--100 patients, 90+ are screened (cheap, easy, the nurse can do it on check in, etc.). Say the machine is only 90% accurate and primary docs never send questionable lesions to the dermatologist (unlikely, since the increased screening frequency is likely to increase awareness and dermatology referrals); end result is 81% correct diagnoses, 9% incorrect, and <10% unknown.

Which scenario is better is not a function of the machine versus dermatologist, it is dependent on the characteristics of the disease. If the disease is common and penalties for delayed diagnosis are high (e.g., melanoma), it is more important to get people out of the 'unknown' category, even at the risk of more in the 'incorrect' category.

I'm not going to comment on your specific example, other than to point out that while both the determatologist and the test are very good at finding cancer when it exists, they both produce a ton of false positive biopsies. I believe it was 1300 out of 1600 or so for the machine.

The FDA did consider your issue, though the company raised it in the context of the physician referring for a biopsy, not deciding whether to refer to a dermatologist. They basically said that the company didn't give the FDA enough facts. The device was only tested on lesions that were picked out as potentially malignant by an expert dermatologist. They basically said that if you want this to be a screening tool for physician use, fine, but you need to show us that either (1) regular physicians are just as good as expert dermatologists in picking which lesions are potentially malignant, or (2) that the machine still does a good job when it isn't given carefully preened data.

I do some freelance medical research for clients and I've dealt with the FDA numerous times. The FDA is annoying, but there's a long list of drugs that were rushed through approval thanks to pressure to not hold it off the market(usually for bad reasons like a drug company got a politician to lean on the FDA), and it led to avoidable health problems in the country.

Yes, the process should be faster. The slow wheels of the FDA leads to advances and drugs getting delayed from reaching the people who could use it, but I'd rather that be slow then see people suffer or die because a drug or device wasn't properly tested.

You're not weighing the risks equally Chad S. You're also ignoring that the slow wheels of the FDA and the expense leads not just to delays, but to drugs never being developed at all. If it costs more and takes more time (and time is money) to develop a drug, then fewer drugs will be developed at the margin.

but I’d rather that be slow then see people suffer or die because a drug or device wasn’t properly tested.

It'd rather minimize suffering and dying. People suffer and die if something is rushed through, but people suffer and die because something is delayed, or because something is never developed at all. You can't just look at one kind of cost without looking at the other.

Go look at Alex's site, and you'll see that the weight of the literature and the evidence is firmly on his side. The FDA on net kills more than it saves.

More people will suffer and die if drugs are rushed out into the market without comprehensive testing. Especially since NIH, the FDA and all the testing looks for people with the conditions to test it on when it gets certain approvals.

No offense to anyone, but this discussion is absurd frankly. Its believed that nearly 30,000 Americans died as a result of Vioxx being pushed out without full testing(nevermind how many lives were permanently effected without costing them their lives). It wasn't pushed out to save people from life/death, but just because Merck wanted to cash in on it. That alone should end the debate on the utility of the FDA and why drugs shouldn't be rushed. I'm positive that it won't because people can't see past their philosophy.

I have patients who have been in pain ever since they lost the opportunity to buy Vioxx, nothing else provides the same relief. They would gladly make the trade-off of a slight increased risk of heart disease for decreased pain.

And since it turns out the problem is with the whole class of NSAIDs, including ibuprofen, which is still available OTC, there is no science behind the FDA--they should either ban the entire class, or return Vioxx.

Don't ignore the fact that regular physicians are already making the judgment to refer to a dermatologist.

And so, the failures of the FDA such as Vioxx are not really the fault of the FDA? It is actually possible that they slow down drug development while at the same time not really screening them properly.

I'm aware of the Hippisley-Cox study, but I've seen more recent studies aren't showing the same results(in fact, ibuprofen is one of the safest heartwise). Rofecoxib(which is Vioxx) is the least safe heart wise along with diclofenac(which also increases your stroke risk).

But, you know, believe whatever you want to.

"MelaFind is a handheld computer vision system, database and expert system"

Why is the FDA even in the business of approving or disapproving completely non-invasive technologies, things that don't even touch a patient? I could see them verifying research results to keep the company honest, but saying a technical diagnostic aid doesn't meet their standards seems to be a reach. What's next, approval for anatomic charts that doctors hang up on their walls?

The FDA has a long history of 'protecting' people from controlling their own health, e.g., limiting home health testing and smartphone apps.

Because its only "non-invasive" until it causes a doctor to order a biopsy.

Ah I see the problem.

(1) Computers don't order invasive procedures
(2) Orders aren't invasive procedures

"Some of these complaints are legitimate, others not so much."

Not so much legitimate?

I wonder if inexperienced dermatologists can outperform experienced dermatologists.

I'd also point out that if you're opposed to FDA regulation of medical innovation, then you have to be opposed to any laws which restrict the rights for citizens to sue drug makers et al if their products harm them(or family members). If you don't want the government to make sure that you'll be safe taking Drug X, then citizens must be allowed to hold the companies accountable for their decisions(or failures to act). And the potential negative effects of new drugs/equipment must be openly available for any citizen to check on their own without interference from the makers.

then you have to be opposed to any laws which restrict the rights for citizens to sue drug makers et al if their products harm them(or family members).

Sure. FDA regulation doesn't currently provide a safe harbor. The only real safe harbor is with vaccines, for the same reasons that cause the government to make vaccines mandatory; i.e., that there are positive externalities.

John Thacker,

You are wrong when you say that FDA regulation doesn't provide a safe harbor for devices, and only does for vaccines.

"Last spring [2009], the Supreme Court held that makers of medical devices that have gone through full-blown FDA clearance, if the products are made and labeled in accordance with FDA’s clearance, are not subject to state tort actions. The federal law, they ruled, “preempted” or overrode, any further state law requirements for those devices. They relied on a clause in the Federal Food, Drug and Cosmetic Act that explicitly said states couldn’t impose different or additional state requirements on medical devices."

Here is the link:

The FDA is disgusting. It should be abolished.

I have some products to sell you. My great grandfather used to sell them out the back of his covered wagon.

Work great on myasma and the willies. And, female complaints. And, you should see what it can do to a bald head.

Why would Mogden buy them? Are you assuming he is stupid?


No, I assume he, like everyone else, has a degree in biochemistry with a minor in organic chemistry.

Caveat emptor.

People buy plenty of products without understanding how they work or if they are safe. You can consult any number of auto magazines before buying a car. Before you buy a house you can hire someone to inspect it to make sure it's in good condition. Certain name brands for other products are associated with quality while others aren't.

I wouldn't buy any medical product being sold door to door and I doubt most people would either. I would only buy them from a reputable pharmacy, which in turn has an incentive to assess its effectiveness lest its name brand fall into disrepute for hawking bad products.

I can see the FDA's point. The scenario is: You go to your GP with a suspicious mole and your GP uses the device and based on its output tells you that it's nothing to worry about. Without the device you would have been referred to a specialist who decides whether the mole needs a biopsy. Effectively the device is an inferior substitute for referral to a specialist. I have heard it said that a GP's role is to know how to diagnose and treat the hundred or so commonest conditions and idf anything else comes up refer you to the appropriate specialist.

Many of the responses to my initial comment are ill-informed; it's helpful to see things at use in a hospital setting.

It's simply not true that the FDA will neg something because it's not as good as the gold standard thought cheaper (or faster, or what-have-you) -- I can think of literally hundreds of counterexamples, particularly in path-lab-type medical devices like this one. But there are also typically restrictions on who uses them and how (so for an example many people have seen a rapid strep test is nowherre _near_ as definitive as a culture but a nurse can use it bedside and if there's a treatment ambiguity the hospital has an SOP resolving that) (I'm a Ph.D., not an MD, and opnly rarely work in a patient setting (although it's worked out i have gained some certifications to do so) so this kind of thing really affects me -- what am I _allowed_ to do?) What should be happening from the review above is the thing is being negged for a specific application and idea of how it's going to be used, as many commenters have pointed out.

A fly in this ointment, pointed out by one commenter, is that many things are used "off-label." Well, sure, and I think that's a _terrible_ idea, let's use things in ways for which they've been validated! Like I said, I have lots of issues with the FDA.

Also note that you _can_ use FDA non-approved techniques in patient care (in fact I've been in a hospital where that was done for a particular prevalent infection) and there are special work-arounds for some tiypes of techniques too (e.g. commonly some molecular biological methods). To a first approximation this is about compliance with accrediting agency standards of care, and it's exactly analogous to why I as a Ph.D. am not allowed to do very veer many things that a far less knowledgeable RN maybe can for you: there's a lot of possibilities for what my training might be, less for hers, and so she's "certified" and that comes with some surety for consumenrs. The docs are certified for some techniques, the nurses and PAs are certified for some techniques -- and y'all are bitching and moaning that someone's trying to do that for technology? Really?

And the key thing is, Dr. Tabarrok's ultimate conclusion is just wrong. You _can_ improve the first gen device, you _can_ feed it real patient samples, people do. You just can't -- in most settings under most circumstances -- use it to dictate patient care. Nothing prevents them improving it, if they can convince the market that they can.

A fly in this ointment, pointed out by one commenter, is that many things are used “off-label.” Well, sure, and I think that’s a _terrible_ idea, let’s use things in ways for which they’ve been validated! Like I said, I have lots of issues with the FDA.

You scare me. You advocate policies that would massively increase death and suffering, and you do so with clear conviction. For me, the data and studies cited by Alex on his site are convincing enough.

I insure nascent medical device companies. A recent company had a device that after trials was found to be no more effective than current drug therapy, and therefore denied approval. It seems to me that the market should decide between two therapies, one a drug and the other a device, not an FDA czar.

Another company has a device that in past years would have been allowed to market in the US under 510K, but with the current activism of the FDA, was told they had to do a full PMA. Their response has been to forego any business activities in the US, and go into commericialization and sales in Europe and the Far East.

This has a far-reaching impact on US in important ways. First, the benefits of the innovations are not accruing to US citizens, and second, those that complain about American jobs being "exported" should take a closer look at this "protective" agency.

Hey Alex, I think people need a refresher on how expert systems work.

As in: "(Arbitrarily) Create a standard of care to be the state of the art practitioner. Do not let practitioner train expert system. Guess what happens next"

What bogus reasons! Behind this is a motivation to protect existing closely guarded and widely profitable monopoly. Medical experts consulting FDA protect other medics. Surprise!

Comments for this post are closed