My take on this is counter-consensus. I don't believe that screening drugs based on effectiveness is a proper role for any government agency. There is a role for them to guard against fraud and to protect rights, but not to prevent people from making risky choices. Adults should not be protected from the decision to take any drug, however harmful, even if instantly fatal, if fraud or force is not involved.
These are extraordinarily complex choices that seldom have a right answer for all possible patients. Therefore they should not be made centrally for everyone. We will learn more and faster from decentralized decision makers, that is, from massively parallel computation from a mass population of decision makers.
Better would be for the only role of the FDA in decisions like this to be determining if the manufacturer is making fraudulent claims. Even better (because of greater computational parallelism) would be for this to be legally decided only by individual states. Best would be to treat adults as adults, and let them and their medical providers make such choices.
What a terrible idea. Nobody has the knowledge base to make a real informed decision on the efficacy of a drug. Not filtering for effectiveness is to endorse fraud and blame it on consumers when it happens.
The whole reason the FDA is set up this way was due to the rampant fraud that was occurring before there were standards for what could be sold to consumers as medicine.
> Nobody has the knowledge base to make a real informed decision on the efficacy of a drug.
This is probably true for most patients, but not doctors. Normally, a patient consults a doctor on what drugs to use for their illness and the risk of taking it.
This is the idealized spherical cow theory of medicine. Doctors have more knowledge but it’s not any more reasonable to expect them to follow the scientific literature for hundreds of different topics than it is to expect the average HN reader to follow the standards and development process for every technology they use. This is especially true given the amount of money at stake: a company has enormous windfalls if they can convince doctors to prescribe it and there’s plenty of evidence showing that they’ll try everything to do that. The amount of money going into sales to doctors & advertising is similar to their R&D spend and there’s no reason to expect the steady stream of ethical complaints would go down with less regulatory oversight.
Agreed, it is not to be expected that a doctor is able to follow the scientific literature for hundreds of different topics. However, I would expect a specialist to know their field well enough to assist a patient/guardian in making a somewhat informed decision.
I take your point that companies can falsely advertise their products. I think there is a place for a government body to make sure claims about healthcare products are factual. In this particular case we are talking about a drug that appears to not cause any harm. So I don’t see any reason to prevent an adult to make a decision with the help of a physician to administer said drug.
> I take your point that companies can falsely advertise their products. I think there is a place for a government body to make sure claims about healthcare products are factual.
The problem is that deciding whether something is factual is a lot easier for simple things. For something like this, there are vaguer connections between facts like plaque removal and more nebulous things like meaningful outcomes for patients.
> In this particular case we are talking about a drug that appears to not cause any harm.
Other than brain swelling or bleeding, headache, falling, diarrhea, and confusion, delirium or disorientation, you mean?
Doctors are doctors not research scientists. They are much more akin to engineers than scientists. There is also a massive amount of specialization in medicine (but not always in prescribing of medicine). Doctors absolutely rely on the work of scientists to help guide them. Doctors are not really trained in how to gauge the efficacy of medicines.
Isn't it more or less fact that doctors are (for obvious reasons) the target of marketing campaigns by the pharma industry to encourage prescribing certain products? Yes, doctors presumably have a higher degree of informedness/resistance to bullshit, but I think you are fooling yourself if you believe that everything is fine with doctors as the thing standing between you and the obvious financial incentives of the pharma world.
The only two options are not either that the FDA screens and approves drugs or it being down to individuals (or even these individuals' doctors). We could have private screening and trials companies that the drug producers themselves would pay. We already see this in other industries where there are both public and private inspections done. I've heard in those cases the private inspectors are actually stricter as well, as they actually have more incentive to be correct.
>Nobody has the knowledge base to make a real informed decision on the efficacy of a drug.
Informed decisions are not risk-free decisions. If you're saying that literally no one can provide informed consent to take a drug until years and years of trials have been performed, and their inability to provide informed consent invariably subjects them to functional degradation or death, then this is a perverse system, to say the least.
Note that many people made the same arguments against fast-tracking COVID vaccines for public distribution, which undoubtedly saved thousands if not millions of lives.
I personally agree that the FDA isn't always perfect (they should never have approved SSRI's, for instance) but claiming that there's no tradeoff here is just incorrect.
The fact the people who make a policy have a specific thing they want to achieve with that policy does not automatically make it a good policy. That remains true even if the thing they want to prevent is bad and could kill people.
I met one of the thalidomide babies once, he was a PI for a research project I worked on. He was a relatively tall man, and had tiny arms that would have looked fine on a 2 year old.
Stopping stuff like that seems like a reasonable thing for the FDA to do, even at the cost of it taking longer to approve drugs.
> even at the cost of it taking longer to approve drugs.
"Taking longer to approve drugs" can be restated as "forbidding the use of potential treatments for longer". To the extent that those treatments work better than the next-best option, delaying approval is effectively forcing the worse treatment on everyone who would have switched for the duration of the delay. If there is no alternative treatment, delaying approval is effectively equivalent to forcing the people affected by a disease to just live with that disease for the duration of the delay.
Picking pretty much at random from stuff recently approved by the FDA, we find that Ocrelizumab, a drug to treat multiple sclerosis (MS), was approved in 2017. Patients treated with Ocrelizumab have just over half the chance of developing symptoms bad enough to require a walking aid (4.2% vs 7.3%) as compared to the next-best treatment (interferon beta-1a) over 96 weeks.
Phase III clinical trials of Ocrelizumab started in March of 2011, and completed in July of 2015. In October 2015, Genentech presented the results of the phase III trials to the FDA. The FDA recognized how important this treatment was, and gave it breakthrough therapy designation, fast track designation, and priority review. It was approved for use in March of 2017.
That is uncommonly fast for the FDA, but even so it took them a year and a half to "move fast" on this treatment. Let's be generous and say that determining that the phase III results were not literally fraudulent takes fully 6 months -- that means that in the case of this particular drug, the FDA only delayed approval for a single year. About 150,000 people take Ocrelizumab worldwide, most of which are probably in the US.
As such, the FDA's (much shorter than usual) delay for this particular drug "only" cost about 2,500 people the ability to walk unaided.
How many delays like this does it take before we decide that the cure (delaying effective treatments for nasty diseases) is worse than the disease (occasionally letting through a medication with extremely nasty side effects)?
This isn't a rhetorical question by the way. My personal answer is probably somewhere in the ballpark of "aim for a 1:1 ratio of QALYs lost to drugs that were approved and shouldn't have been : QALYs lost to drugs that should have been approved and weren't". Maybe 1:5, if we think harm from inaction isn't quite as bad as harm from action. But my best ballpark guess for the actual ratio the FDA is achieving is probably more like 1:100 or 1:1000, and I don't find that even remotely reasonable.
(Note that this entirely ignores the question of whether the current FDA does even have the benefit of delaying / not approving stuff that doesn't work, which is the subject of another rather active discussion on HN today: https://news.ycombinator.com/item?id=27472107).
The problem is that it costs $56k/year. Most people don't have that kind of money, so for the most part we're talking about whether that cost should be socialized, either via insurance or the government. That's an awful lot of money that could be used to treat a lot of other sick people. So it's not primarily a question of whether people should be allowed to take risks, it's more a question of how best to allocate finite resources.
I am a smart person, and I am not remotely qualified to make decisions about the risks associated with an untested drug. I’d like to know when I get a prescription that someone else went through the process of being a willing volunteer in a clinical trial.
Frankly, I think most doctors aren't qualified to make such a decision by virtue of lacking PhDs: their expertise is in clinical care, not in evaluating research.
Sure, but the doctor doesn't have to do much research in this case. This is front-page news in the NY Times. And about "every single medication": this is the only approved medication to delay clinical decline due to Alzheimer's. And there are only a few medications to treat symptoms.
I know this isn't a perfect world, but it doesn't seem to be asking a lot for a doctor to know this. (Especially for a drug that currently costs $56K/year and presumably will cost more next year.)
> I don't believe that screening drugs based on effectiveness is a proper role for any government agency...Better would be for the only role of the FDA in decisions like this to be determining if the manufacturer is making fraudulent claims.
I think you come in with wrong assumptions. The FDA has little to no authority when it comes to off-label uses — doctors have a lot of leeway to prescribe drugs for whatever non-approved use they judge to be effective.
So the FDA's role already is intertwined with marketing. A doctor can today prescribe Viagra as a treatment to Alzheimers, but it would be illegal for Pfizer to market Viagra for treating Alzheimers until it gets FDA approval for that specific use.
> Better would be for the only role of the FDA in decisions like this to be determining if the manufacturer is making fraudulent claims.
Let us recall some of the key facts about the drug in question here. There were two trials. One of them showed such little benefit that it was halted midtrial on the basis that it would be unethical to continue giving people a drug that didn't work. The second showed marginal benefit.
Biogen argued that there is a rational explanation that explains the two trials that points to the drug working. The expert committee at the FDA had 10 people vote "that's bunk", 1 person vote "unsure", and nobody vote "I agree." That's a pretty strong endorsement for saying "Aduhelm treats Alzheimer's" is a fraudulent claim, and yet the FDA decided that the claim is merely disputed.
The only way you can look on the FDA's decision favorably is if you know nothing about how drug development works, which is unfortunately true of almost the entire public. (Here's two facts that are important to keep in mind: most drug candidates don't work, and placebo effects are very real and surprisingly powerful).
Efficacy and safety are intertwined. Safety can sometimes only be evaluated in the context of efficacy. For example, you wouldn't amputate a healthy arm. But for a gangrenous one, it may be the only option.
For establishing efficacy and safety, trials are the only process that works. Nobody can predict the effects of drugs because the current understanding of biology is very incomplete.
Thalidomide [1] shows the wisdom of the current approach.
None of this vision would work for a variety of reasons, but the biggest one is that insurance companies are not going to agree to pay for medications that are not proven effective.
Your personal freedom to try whichever medications you want would be limited unless you are willing to pay out of pocket for everything.
I agree. The FDA should be a watch dog group that provides us with the knowledge to make our own decisions. How many people have died because a drug took too long or never made it to get approved?
I agree. And regardless of where you stand on the issue, the current regulations with regards to off-label usage are totally inconsistent. Doctors are perfectly free to prescribe drugs for conditions that are completely unrelated to the what they were approved for.
Drug X is proven safe but not proven effective. It's illegal to use to treat condition Q under any circumstances. Drug Y is proven safe, proven effective for condition M. It's perfectly legal to use for unrelated condition Q. How does this make any sense?
Maybe, but it’s not clear to me that the FDA is (or should be) experts in weighing economic incentives for good behavior. There’s plenty of bad economic incentives created by the current FDA process; just ask the Lovaza guys, charging a 10x markup for fish oil. In an ideal world, I don’t think the FDA should have any role in drug approvals other than preventing direct medical harm.
I believe the advisory panel's position is that it is not indicated for the treatment of Alzheimer's and that Biogen should not be allowed to say this.
The FDA could have said "this drug is safe and may be sold, but you may not describe it as an Alzheimer's treatment, and you must explicitly state that it has been found to be ineffective as an Alzheimer's treatment in clinical trials".
So how is Biogen going to get doctors to prescribe and people to pay $50K for a drug with no brand recognition nor eligibility for insurance, and no claims of actual effectiveness?
I suppose there is some fine ethical reasoning that would distinguish between selling something that you know not to work and selling something that you don't know won't work.
Of course, financially incentivizing ignorance is probably not the best precedent.
* Currently the US welfare agencies will provide ~any FDA-approved drug for someone with the illness for which the drug was approved
* In this case, they probably won't - in the worst case scenario it would double their spending on prescription drugs and seems unlikely to help anyone
* So this is going to sever the link between FDA approval and medicaid spending.
* Now you have an awkward situation where agencies that not really equipped to make judgements on dug effectiveness, and which are not insulated from politics, have to make those decisions.
This could all work out for the best in the long term but may also be very awkward in the short term.
Your argument misses some of the important reasons of the approval process.
Firstly, approval limits the legal liability for drugmakers. If we adopt a policy where approval does not exist. As long as the government takes part in the risk so it should have a say in the matter.
> treat adults as adults, and let them and their medical providers make such choices.
Signing out of medicare and all government assistance in case of becoming invalid, so that government is not in the hook for personal decisions would be a huge ask.
If 10 million individuals make the same decision independently, and it backfires, that has negative consequences for the rest of the country. One decision does not matter, millions of decisions become collective matter.
Safe drugs can be sold OTC with all the homeopathics. FDA approval here is about fraudulent claims of effectiveness.
You argue that unsafe drugs should be available for sale too, but that's not what happened here and IMO it ignores the societal risks of unsafe things floating around were they can harm non-consenting people, as well as the rights of people who want to be free from transient temptation to make irrevocable decisions.
So, do you think people should be allowed to take ivermectin and "that drug" (the one that rhymes with mydoxyborofin) for COVID-19 if they want to, given that both of those have been studied for years and are known to have very little side effects?
No, the patents on these have expired and nobody will make any money if we promote them. Also it would have been near impossible to get emergency use authorisation for these vaccines if we had at all listened to the science on these.
These are extraordinarily complex choices that seldom have a right answer for all possible patients. Therefore they should not be made centrally for everyone. We will learn more and faster from decentralized decision makers, that is, from massively parallel computation from a mass population of decision makers.
Better would be for the only role of the FDA in decisions like this to be determining if the manufacturer is making fraudulent claims. Even better (because of greater computational parallelism) would be for this to be legally decided only by individual states. Best would be to treat adults as adults, and let them and their medical providers make such choices.