" If you’re at very high risk of dying in the very near future from a fatal disease, then you probably aren’t as worried about the downsides of experimental treatments."
There's little real downside to me being harmed or even killed by experimental treatments: recurrent/metastatic squamous cell carcinoma is already virtually always fatal. The FDA is "protecting" me from harm, and the result is going to be me dying from disease. This is nonsensical and yet standard practice at the FDA.
As it is, I may have to go to Mexico to get intratumoral immunotherapy injections that aren't formally approved in the U.S., but should be, given that the immunotherapies themselves are: https://williamscancerinstitute.com/.
Thank you! It's been and continues to be a challenge. I wish I weren't in line to be the face of the otherwise invisible graveyard, but that's the way things are going.
I think this analysis is good on the merits, but misses key details. My qualifications: I am a PhD in Cancer Biology, currently still in academia but have done an internship in a biotech company and have taken coursework related to the economics of biotech/pharma.
Matt's analysis focuses on "safety", but that is not the key driver of cost in clinical trials. Safety concerns are mostly addressed in Phase I and/or II trials with relatively small patient cohort sizes. The larger driver of cost are Phase III efficacy studies, where larger patient cohorts are required in order to have sufficient statistical power to distinguish response rate. Critically, the FDA requires these large, expensive studies to prove that the trial drug is *better* than standard of care. Merely matching, or very slightly underperforming standard of care is a failure of the trial. These Phase III trials cost tens or hundreds of millions of dollars, take years of time, and are (to my knowledge) the main driver of drug development to the point of dwarfing all other costs. And, as MY briefly touches on, a pharma company has to price in the cost of each failed trial into a successful drug. (A new drug doesn't just have to pay for its own development costs; it also has to pay the costs of all the failed drugs the company has not gotten past the FDA. Otherwise the economics of developing new drugs doesn't work out.)
While it is possible that a drug fails in Phase III due to previously-undiscovered safety concerns (patients end up advancing in disease, or having toxicities not seen in the Phase I/II), the more common failure is due to lack of sufficient efficacy. (Which is what you'd expect from an efficacy trial). There's obviously enormous benefit to society to have drugs on market that are equally effective to standard, or perhaps even slightly less effective. No clinical trial is a perfect representation of the population, so even a "less" efficacious drug might have use in certain patient populations. The pharma company could even price the "worse" drug at a small price. Patients could decide themselves to take a risk on a cheaper drug, and the company would still lose money, but far less than they'd lose if they had to scrap the whole program. Win-win. The merits of an "equally" beneficial drug are even more obvious. More market competition! Even relaxing the rules would allow more drugs to come to market (and thus change the cost/benefit equation of drug development) with very limited risk to society, in my opinion.
None of this even touches on the biotech-vs-pharma roles, how VC feeds into the ecosystem, etc. Biotech is heavily dependent on venture capital, which has dried up in recent years due to increased interest rates. Nearly every drug is brought to market by pharma, but nearly every drug begins its lifecycle in biotech, and IP is then acquired by pharma (for reasons to lengthy to get into here). Biotech has been in a recession since early/mid 2022, and adding increased pressure right now with price controls in the IRA has really worried the industry. There's so much more on the "development" side that MY either doesn't know or didn't have space to touch on. The US really is the envy of the world though when it comes to medical innovation, and that really is driven by market economics in a way that could be greatly disrupted by pricing controls. Further changes to price controls need to be paired with supply-side reforms (like FDA regulations, etc.) if we want to continue to see robust drug development.
Thanks for tuning in here! Doing research for this article, it was hard to find good reforms for the FDA approval process. Are you aware of any that we missed?
Nope! I think everything in the article was accurate as written; I just don't think that "safety" concerns are the leading driver of trial cost. Obviously it would help the economics to relax safety concerns, especially for terminally ill patients that have no better options. This has been done in the past with HIV/AIDs meds, and I believe with breast cancer as well. Those efforts are well documented and resulted from populist pushes from patient communities (and associated activists) that wanted change.
Most of my education and experience has centered on understanding *why* costs are so high, and why high prices are necessary to meet the cost of development. It's successfully converted me from a stance of "pharma companies are bleeding the taxpayer dry" to "actually while some (ie Martin Shkreli) are leeches, many drugs that get bad press are actually priced to the market". Unfortunately, I have never been part of any activist or political communities on the subject, so I don't have any experience with "ok how do we change this?". I don't have any well-researched stances on reform. I took a course taught by someone who had founded a successful biotech company, sold it, and then used the profits to found a VC firm focusing on funding the biotech sector. He stressed the "superiority" in efficacy part of drug development, and he claimed (logically) that an ever increasing burden of superiority would lead to an ever increasing cost of development, as all the "low hanging fruits" got picked. That's why I focused on that topic in my comment, but I should have clarified that this is coming secondhand from an industry expert who likely didn't do any truly rigorous policy analysis on the subject.
What I failed to appreciate before taking that course was just how incredibly labyrinthian drug development is. The relationships between pharma, VC, and biotech are very complex, the risk is far higher than most industries would tolerate, and VC/pharma needs to be leveraged to the tune of hundreds of millions, if not billions of dollars, for many years before even sniffing the possibility of an ROI. The IP clock starts from patent filing, not from drug approval (as I understand it), so the length of the drug approval process plays a huge role in subsequent drug pricing. I was taught that a drug typically has a 5-10 year window of profitability before generics take over the market, in which time the company needs to recoup all of its losses, plus make enough money to keep new R&D alive. It really feels like the alluring prospect of multi-billion dollar/year blockbuster drugs is the only thing that keeps the process moving.
Thanks for this explanation of how VC money has infiltrated R&D. This happens in pretty much every industry. Large companies used to have to fund R&D themselves, but have effectively outsourced it to VC/PE over the last 20 or so years due to quantititive easing creating. Why absorb the risk when you can pass it on to the finance guys to fund these biotech companies with great board seats for the semi-retired corporate executives to sit on? Everybody was winning until the money dried up when interest rates hit 5%. It's not a bad model, but what I would like to see the government get into the game as a VC entity to support R&D efforts that are in the national interest. I think the Obama administration was heading that way.
Certainly they could do a *lot* more, but there is at least the SBIR/STTR programs that are congressionally mandated to provide money to early-stage startups. It's only up to ~$1m, which is just seed-round-level funding though.
I haven't seen this elsewhere, but I'd like to see a law passed requiring the FDA to use a consistent efficacy/risk calculation across therapeutic areas. I'm not saying that a new cancer drug should be held to the same risk level of a new rash treatment cream, but that the balancing process should be standardized.
The medical experts at FDA come from the therapeutic areas they are juding. Medical oncologists for oncology drugs, ophthamologists for eye drugs. The culture of cost/benefit is very different across these fields. Then add in the non-FDA advisory groups who exacerbate those biases.
The result is that oncology drugs get approved with risk/reward ratios that are much less favorable than drugs for untreated diseases that get denied.
Having the FDA do cost benefit analysis (on its own process) is one thing I've seen repeatedly. Glad Matt mentioned cost-benefit and hope to see more in a future post some day.
Reducing the cost of clinical trials is often mentioned, and it seems like the government could help reduce per patient costs by investigating what costs (recruiting, administration, etc) and then building capacity or changing regs.
Dedicate an NIH lab to replicate promising academic results to save time/money pursuing findings that do not replicate. Right now companies do that and do not publish it, so a bunch of companies do it.
There is some good work where different companies share a placebo group. So one placebo group for two drugs instead of two if the trials were separate. Saves a bit.
Ethically fraught, but you could consider letting academic hospitals give terminal patients tool compounds to test new targets/strategies for their disease. Maybe there is some leeway to do this now, but for whatever reason I see very little data from it.
Often mentioned, but get regulatory requirements from around the world harmonized better.
I was just about to raise this point myself. I'm involved in the design and conduct of phase III trials, and at least for my therapeutic area, efficacy concerns determine sample size - not safety concerns.
That said, there are some aspects of phase III study design which are geared towards addressing safety concerns - such as the prevalence of open label extensions.
What about the price per patient? I would have guessed that average P3 study sizes are going down just becuase more drugs are targeting on orphan diseases where smaller studies are needed for a variety of reasons. But the cost per patient has risen dramatically. What do you see in your world on both front?
Why cost per patient could be rising -- some of which could be improved by the FDA
- lengthier study periods to demonstrate efficiacy. Sometimes this is due to the condition being treated, sometimes due to FDA having a higher bar to how efficacy is measured
- Requirements for more diverse study populations. I think this is a good thing, but it really drives up costs
- More follow-on studies are needed. A lengthy Phase 4 study can add a lot of cost and really should be considered as part of the Phase 3 registration cost. This is a bit complex since pharma will do extension studies both becuase regulators require them and for commercial reasons
- More byzantine documentation requirements. Probably a clear opportunity for FDA
I’m not very familiar with the price per participant side of things.
One thing I’ve noticed is just how comprehensive some studies are. Even “confirmatory” phase III studies don’t just have primary and key secondary efficacy and safety endpoints - they have lots of exploratory endpoints as well. I feel like the ease with which companies can collect, store, and analyze huge amounts of data makes it far more financially beneficial on average to over-collect, rather than potentially under-collect, data.
it's hard to disentangle which exploratory endpoints are there for commercial versus regualtory reasons. Often they are there at the reuqest of the marketing team wanting to maximize what they can promote with and to set the stage for future clinical research not done by the pharma company
Interesting! You probably know way, way more on me about the subject, then. Do you agree with me that relaxing efficacy benchmarking to “non-inferiority” would significantly impact the economics of therapy development?
Firstly, it's not always the case that an active comparator arm is required; there are some drugs for which superiority to placebo is sufficient for approval. Obviously for those drugs it wouldn't make a difference.
Secondly, although I'm fairly removed from the marketing and finance side of things, my understanding is that "not worse than the standard of care" is not a very compelling angle from the sales perspective. I think you'd still see studies designed to show superiority to SOC in some aspect - fewer side effects, better effectiveness in a particular subpopulation, etc.
I agree. Clinicians lean on the long-term, real-world safety data of standard of care and are reluctant to switch to something new based on non-inferiority, unless, the standard of care is significantly safer (non-inferiority is a measure strictly of effectiveness). Take warfarin: decades of data, but known issues. If the next gen anticoagulants can beat it in non-inferiority, they will win, but this is an unusual case. For the most part, as a pharma company, investing in the marketing and sales of a drug that has only proven non-inferiority is a losing proposition.
Does the FDA actually itself require a new med to beat standard of care, or is that just an artifact of IRB ethical guidelines? My understanding was the latter, which is why you have a lot of “me too” drugs for non lethal ailments. eg: A new hypnotic doesn’t have to beat Lunesta to be approved for insomnia.
Good question! I am not sure to be honest. I have a strong memory of discussing the superiority requirement as an FDA regulation, but you’re also right that there’s a whole class of “Me Too” drugs that just hop on a winning horse. I suspect that if a given MoA has been proven to the FDA, then the regulatory burden changes. For example, if you make a new molecule that inhibits the same target, and can prove that it acts similarly, it would still be a new IP (a novel compound) but the same mechanism. I imagine that could be a different approval route through the FDA than a totally new treatment modality. I also believe that once the patent has expired people can make generics without significant FDA approval processes.
it really depends on the drug. IRBs and FDA are using the same general framework, although it plays out very differently. Both are weighing the clinical benefit versus risk of the drug for a study or for broad market approval.
The higher the risk, the higher the efficacy hurdle. If a drug could have worse side effects than standard of care, it will need to promise better efficacy to move forward -- non-inferiority won't be enough. If a drug has better safety than standard of care, it might be possible to do a study that doesn't show non-inferiority.
An investigational drug almost always has higher risks than an approved drug solely becuase it hasn't been studied as much. That's why you seldom see them held to a lower hurdle than the standard of care
Also, FDA essentially is the IRB for clincial studies that will be used to support approval. IRBs are only the sole judge when it's a post approval study or, rarely, a pre-approval study not used for registration.
Your understanding of safety is constrained. It’s not just finding terrible toxicities, it’s also establishing a profile of adverse reactions that are not so serious as to prevent approval of a drug, but are necessary for clinical judgment as to when to prescribe a drug or recognize when adverse events occur that are due to that drug
This comports with my understanding of the situation. Let the FDA sort out the safety issue and let the market sort out efficacy, which is a much more complex informational problem.
I think if I have cancer I'd prefer not to be part of the market that "sorts out efficacy" but would rather like to know if this drug is going to help me.
That's why company-funded trials would still be necessary! No one will take a drug that is totally unproven. And likely the FDA will always need to play some role in ensuring efficacy. However, the approval process being simplified or streamlined could be better for everyone: pharma companies, patients, the taxpayer, etc.
You might be surprised that, in my experience, cancer patients tend to disagree with you. Critically ill patients with no alternatives often feel that the FDA is standing in the way of them obtaining new experimental treatments. (See the top comment on this article from jseliger.)
I understand that desperately ill patients are willing to try anything with any chance of working. That's why the FDA allows humanitarian exceptions. I would just not leave the whole efficacy testing phase to the market. As we obviously agree.
“ You might be surprised that, in my experience, cancer patients tend to disagree with you. Critically ill patients with no alternatives often feel that the FDA is standing in the way of them obtaining new experimental treatments. (See the top comment on this article from jseliger.)”
Yes, that’s how you get snake oil sold to desperate people. You could easily be giving toxic meds with no hope of working, so there has to be some process for these exceptions.
The issue is that most things are not cancer with months to years of life expectancy. They are "I have a disease and I need a medication to prevent me from having a heart attack or stroke, or ending up on dialysis, or needing something operated on or transplanted in 10-20 years." It's not even as straightforward as "does this normalize your blood pressure, heart rate, glucose, or cholesterol". Most things are even more complex. How do you have the market sort that out?
My prior is that markets are better than planning and that more better cures for cancer would be found under a market regime than not. A more risk-averse person could simply not partake until doctors were positive something worked. A less risk-averse person would make a different choice.
This is absolutely not the case in pharma based on the incentives of two central actors:
1. Pharma wants to minimize clinical study risk while maximizing commerical returns. In practice, drugs with better efficacy and higher risk could get quashed in your scenaio.
2. Insurors are dis-incentivized for breakthough treatments to come out. They do best in a market with lots of me-too drugs. Look at the Ozempic, etc. weight loss drugs. These have the potential to generate unbelievable health and economic value, but payors are dragging their heels because of the short-term P&L hit. Payors will resist anything with less than a 2-year payback, sometimes even shorter.
Re 2, my layman's understanding is that this is overblown and that companies will frequently produce drugs that put other drugs out of market for short term gain. The entire point is that companies are in a prisoner's dilemma with respect to themselves. They can't just coordinate to hold off on releasing better drugs that actually solve an issue, because each actor is incentivized to generate profit where they can.
One of the biggest frustration within pharma is the behavior of payors, especially how indifferent they are to rewarding innovations that improve health.
There may be a bias in the market-generated information on the drug's benefits, because those who benefit will send signals that, hey, this is a great drug, and the ones who die will be, well, dead.
Yea people die all the time the question is what system saves more lives than not. The idea is that if we feel like there is a market failure in information generation on the part of safety, then we can have the government regulate that. There is not a clear public interest in regulating efficacy.
Fundamentally, my prior is that markets are better than planning. Doctors will likely not prescribe drugs with a bad reputation, and patients will hold out for drugs that are more proven. Some people will choose the riskier option.
My prior is that this will produce higher aggregate welfare, since markets are better than planning. That means they produce better results, and in this case better treatment. Regardless, safety could be regulated if people think there is an information problem there. I'm not an economist so I don't have a strong opinion on that.
You mention biotech -- would it help if the federal government invests directly in biotech companies? Either with subsidies and tax incentives, or (more appealing to my gut), with something like a sovereign wealth fund acting as an investor, and standing to profit from biotech companies that succeed in business?
It would help if Biden wasn't proposing a rule that makes it less likely for biotech to partner with the government to use science based on federal funding. (basically threatening to take intellectual property)
Also, just more funding for research, and funding to help pay for promising trials would help
>basically threatening to take intellectual property
That sounds vaguely familiar, but I'd appreciate a link or details anyway. I've heard of proposals for something like that for COVID vaccines -- preventing infectious diseases has particularly large "global public good" effects, so you can make a case for wealthier people to have to subsidize them, but there are perverse incentives with the subsidies coming specifically out of reducing the value of developing future medical treatments.
In my opinion it's not a legal attempt as the law has never been used in this manner, and the authors of the law state that it's not supposed to be.
But if successful it would mean that biotech companies would no longer use government research. Of course the whole purpose of the law was to get companies to start using government research so it didn't go to waste.
I don't think the government could do this well - it's hard enough for VCs to know what to invest in - and would probably result in a backlash. Everybody remembers that the Obama administration loaned money to Solyndra and it went bust; nobody remembers that they loaned money to Tesla.
I'd rather see massive government investment in foundational work that makes clinical trials less expensive. Instead of saying "foreign data are bad," build models that can help pharma recruit non-US patients who are collectively representative of the US population.
I know governments also invest in basic biology, which contributes eventually to future applications. Maybe expanding this investment would be a good way to boost drug development 10, 20, 30 years down the road? Though that's a long time horizon, too long to pay off for today's incumbents. And it may be bottlenecked by the availability of good researchers.
I like your idea of models to recruit non-US patients. Making clinical trials cheaper will pay off sooner than advancing basic biology.
I imagine that better knowledge of biology might have more dramatic effects in the long run -- a better understanding of the mechanisms of diseases, the molecular biology and cellular targets that you can go after, etc. might not just suggest new drug candidates, but maybe also let companies start out with better candidates for early clinical trials, improving the success rate, so you don't have to pay for so many candidates that fail in phase 2 or 3 because they just don't work in humans like they do in mice.
In principle, of course, the government can do both -- it's a question of what kind of resources they will want to put into future drug development.
You have a good point about Solyndra. The government is notoriously risk-averse, but that's where all its incentives point.
More money is always nice for an industry. The government already heavily invests in basic research though. Academia is essentially fully reliant on government grants. Govt funding probably wouldn’t move the needle on the problem IMO, and just change the nature of biotech into something closer to private academia.
I’d be curious to know how far from saturated basic research is with respect to medical advances. I’ve expressed before how many conditions are far from being well-managed or cured. Like why couldn’t or shouldn’t basic research get 10x the funding it currently does?
As someone making a career in basic research, I would love for it to get 10x the funding, and I think it would have enormous benefit for humankind and the US more specifically. Funding for cancer research, at least, seems to do slightly better than the average discretionary program, but it's still subject to all the same politics. Biden is a huge proponent of cancer research funding (for obvious reasons), and has championed multiple big bills increasing funding, during his presidency and Vice Presidency. However, the recent budget reconciliation that passed the House has funding flat or cut in certain places, because Republicans want to cut public funding! The exact same politics are at play in federal research funding as for any other public good. Elect more Democrats and the budgets will probably go up, etc.
Before the really unfortunate politicization of vaccines took hold I thought Operation Warp Speed might inspire a whole new appreciation for the power of medical research. It’s quite strange really how we don’t make the connection between all of the medical issues we and our loved ones could face and the possibility of addressing them at a faster pace. Some weird combination of missing political agency and medical denialism.
Not that I’m immune to this. It only occurred to me in the last few years. I never heard anyone make the case. Or it never stuck.
The government should just nationalize all pharma/biotech obviously play stupid games win stupid prizes that’s what they get for all their patent manipulation
The FDA's standard for equivalency really doesn't impact the cost of trials after you consider study design and market dynamics. Consider three situations:
1. My new drug is expected to be significantly better than standard of care. Becuase I expect a large difference in efficacy, my study doesn't have to be very large. And, because of the efficacy difference I can expect to get a price and/or share advantage in the market.
2. My new drug is expected to be roughly equivalent to standard of care. Because the efficacy is similar, I need a huge trial. Now, when I go to market, I've spent a ton of money on the trial and probably have to take a price discount. If I do that and start taking share, I have to consider that the incumbent will lower prices to block me. Project like this tend not to have a positive ROI and dont' get funded.
3. I think my drug will be equivalent or inferior to standard of care and I do a smaller study comparing it to placebo (kind of your scenario). Neither payors nor physicians will support something that doesn't meet the standard of care. Most importantly, physician malpractice liability is highest when physicians deviate from standard of care.
In #2 and #3, I also have to spend a ton more money on sales and marketing to convince people to use my drug, so the economics get even worse.
Great points! Do you think it's fair to summarize thusly: "while the FDA has high standards for approval, if a drug wouldn't meet them, it probably doesn't have enough market appeal to succeed anyway"?
At a high level yes. But it's more complex by the detail. The rest of a re-post of my response to Ben's question about reforms
I haven't seen this elsewhere, but I'd like to see a law passed requiring the FDA to use a consistent efficacy/risk calculation across therapeutic areas. I'm not saying that a new cancer drug should be held to the same risk level of a new rash treatment cream, but that the balancing process should be standardized.
The medical experts at FDA come from the therapeutic areas they are juding. Medical oncologists for oncology drugs, ophthamologists for eye drugs. The culture of cost/benefit is very different across these fields. Then add in the non-FDA advisory groups who exacerbate those biases.
The result is that oncology drugs get approved with risk/reward ratios that are much less favorable than drugs for untreated diseases that get denied.
Thank you for your input! I'm wondering what you think about the idea of separating the efficacy test from the safety test. What would you think of a separate third category of drugs that are basically approved for safety but not efficacy?
I don't know enough about the situation, so this is speculation. I think it would be interesting to have conditional approval after phase I/II trials, but it's unclear if anyone would take them! Healthcare is a weird industry, because the customer is not actually the payer. In most cases, the patient and doctor are making purchasing decisions, but insurance/Medicare are deciding what things will be covered. In a standard market, you could imagine a case where a patient would rather pay 1k for an experimental drug vs 100k for a tried and true drug. However, the patient is choosing to pay a 1k co-pay for an approved drug (while insurance takes the 99k) vs 1k out of pocket for an uninsured experimental treatment. Typical market economics don't apply! I think that efficacy approval is going to be necessary for drugs to have any chance on the market. I don't see any future where the FDA gets out of phase III, efficacy testing, etc. It's interesting to speculate though, and relieving regulatory burden in drug approval is one area where I believe that society can make life cheaper without really impacting the quality of newly approved drugs.
The patent system wasn't bestowed on humanity by a deity. Governments around the world created the patent system in order to foster innovation. It strikes me as absurd that governments chose to create a system that increases costs for the governments, and then the governments try to find ways of reducing those costs as though it was nothing to do with their own system.
If the patent system currently strikes the perfect balance between fostering innovation and cost, then there is no advantage in attempting to subvert it. If it fosters rent seeking and bloat (and it does!), then modify the patent system directly.
You are forgetting that the IRA price negotiation mechanism is structured as a tax, which can pass through reconciliation with a simple majority in the Senate. Revising patent laws would go through regular order and be subject to the filibuster.
Has any party just tried to appoint a cronyist parliamentarian willing to rubber-stamp anything and everything as subject to reconciliation? This seems like an obvious tactic.
Didn't the Republicans fire their parliamentarian in the early Dubya era for ruling against them on something? I'm too lazy to look up the details now.
I read a lengthy article by an attorney one time about how random & capricious the rulings from the parliamentarian are, about what will increase the deficit (and so won't qualify for reconciliation) and what won't. Doesn't seem like a great system to invest that much power in 1 single person
I don't have any specific knowledge about the details of the legal mechanism being a tax, but my immediate presumption is that, if the reasoning is as you state, that's another good reason to oppose the whole thing. It's not only a stupid way to do it, it's also doing violence to the constitution in the pattern of "A mandate isn't a mandate if we call it a tax." If the only way to pass a law is to lie about what it is to hide it in the reconciliation process and then hope the courts fail to notice then it's not a real law.
The relevant roadblock here is not the courts or the constitution, it's because of archaic Senate rules that require 60 votes for most stuff.
Your suggestion of just not doing stuff if it requires massaging and torturing its meaning is a fine-sounding idea until you realize that following this logic would totally paralyze the American state and prevent the legislature from functioning.
The best solution would be to end the filibuster and then just pass the actual law, but unfortunately huge swathes of people have decided they are committed to this institution for no reason other than that it gives them good vibes and feels Moderate.
How much stuff does or doesn't get done because of the filibuster is a distinct question about calibration that people are welcome to argue about. Not liking the current equilibrium does not excuse assaulting the process instead.
Again please explain how Congress, which is given the power to tax, enacting a 95% tax if companies do not come to the negotiating table, is assuaging the process. Seems to me like a totally valid exercise of the taxation power.
That’s a silly argument Dave. A 95% tax (see John from FL’s comment above) is still a tax, and Congress has the power to tax. If you don’t like how they exercised that power, vote for a different Congress.
See also the ubiquitous discussion in media outlets about how if a judge finds the FDA incorrectly approved an abortion or birth control-related drug that results in a "ban" of the drug, which completely ignores that (1) the only reason such a ruling would "ban" the drug is because of other already existing federal statutes and regulations that presumptively prohibit the sale of any pharmaceuticals in the absence of FDA approval, and (2) Congress has the power to amend those statutes and render the court decision moot.
But we could have a better discussion! One way to helpfully conceptualize for the Slow Boring crowd might be to imagine the government provided housing for all American citizens over 65. Should the government employ price controls in its housing vouchers to ensure greater housing is available?
Arguably we do *something like this* with Section 8, aiming at Fair Market Rents (FMRs) which HUD publishes regularly for a given area. Should HUD bargain those rates down further? If we assume we're expanding this to seniors (and we're assuming a world where seniors take up disproportionate housing use like they do with prescription drugs), would it be productive to set price controls this broadly across the industry? Would we get more housing over the long term?
I think this is more helpful of an analogy than bulk discounting for federal employee hotel use (which I assume is not the majority of hotel consumption.) In that spirit, here are two pieces by Manhattan Institute's senior fellow Chris Pope on the ideal drug R&D industry situation and the practical harms Congress could impose on it in the future.
For a starter, granting monopolies is a completely archaic practice that belongs straight to the 17th century. The modern alternative (which exists under various forms since the 19th century!) is compulsory licensing, whereby the patent holder is guaranteed a ROI in the 10-20%pa range on their R&D expenditures, but production and sales still follow market competition.
Anything a single government (including the US) can do is try to *emulate the effect of compulsory licensing* by controlling prices based on production and R&D costs, and shorten patent length.
Large-buyer price negotiation is consistent with the right for a patent-holder to charge monopoly rents, though, as long as the patentholder has the right to take its ball and go home.
I don't know that the patent system alone is the problem, though there is certainly a case for better calibrating it. The problem in this particular case is more that it operates, at absolute best, in a strictly agnostic manner with all of the many other moving parts in the larger healthcare system, rather than towards a coherent set of goals and priorities. The entire US healthcare system is full of these kinds of issues.
I agree with Matt's analysis of how the FDA should improve its regulatory approach.
Matt is underselling the benefits of drug negotiation, though. The IRA has shown the way, and we should replicate the drug negotiation process across the rest of the federal budget. The relevant agency will determine what it is willing to pay for a product or service, and if Company A refuses to sell for that price, then any revenue from that product sold to others can be taxed up to 95%. Or Company A can refuse to sell anything to the relevant agency. Easy-peasy.
If Matt thinks the current negotiations over hotel rooms are beneficial, just wait until we apply this new process to hotels! The government can tell Marriott that federal employees can stay at a Courtyard hotel for $23 / night. I mean, the room is already there, the hotel is already built. It doesn't cost much for a little bit of water and to run a vacuum through it the next day. $23 seems fair enough, and that price isn't subject to the APA or judicial review anyway. If Marriott doesn't agree to it, we will just take 95% of their Courtyard revenue as a tax. Now Marriott can avoid this outcome if it agrees not to rent a room to any federal employee across any of its hotel brands. If Marriott does this, then we just have a similar negotiation with Hilton or Holiday Inn. We'll give them an offer they can't refuse.
Somehow, price negotiation has existed in most other OECD countries for decades, that produce 50%+ of major Pharma r&d, with also much much better health and pricing results for their citizens.
But I guess the fringe case 1984 dystopian scenario is a much more entertaining frame of reference for this convo!
I don't like the way costs are distributed either, but it's an incredibly capital intensive industry, and currently the US is basically dragging the rest of the world with it in terms of development costs. From anecdotal experience in the field, we didn't even project European revenues when considering which candidates to bring to clinical because a drug that wouldn't be approved in the US could never break even its research cost, no matter how many countries outside the US approved it.
One counterargument to "works in other countries" is if the U.S. is paying "full freight" and others are negotiating, then we are subsidizing their R&D.
Now, that's not exactly great either (when we're talking about subsidizing England, France, Germany) - I can certainly believe even if the doomsayers are correct that it would be better for those countries to pay _more_ than they currently do and for us to pay _less_ at the same time, but that's not the same statement as "we could just drop our prices down to Western Europe's prices" and that would be fine.
Or maybe the US drops prices and EU’s rises a bit, leveling the playing field a bit, which I think would be ideal. If US wasn’t such a cash cow, pharma companies would have to negotiate harder in the EU.
As it is they don’t have to negotiate in the US, so they don’t need to worry as much as EU
One possibility (I want to say something of the sort was debated in Congress in recent years, but I'm not sure) is for the US to institute price controls tying medication prices to other peer countries.
For instance, for all medications that have been approved in multiple OECD countries for a certain time, calculate an average price (maybe adjusted for purchasing power or GDP per capita or median personal income, and weighted by population). Then, by law, the US price for that medication can be no more than 120% of this average price, for "peer" countries excluding the US. US prices will still be on the high side, but not as big an outlier as they are now. By explicitly tying it to other countries' prices, pharma companies are also incentivized to drive harder bargains in other countries in order to increase the price they can get in the all-important US market. Thus, American patients will only provide a small subsidy to their foreign peers, if any, while the US market will still be very lucrative.
European pharma companies still make most of their profit from the US market, though -- which affects the benefit side of their cost-benefit analysis. If a drug is worth, say, $20 billion in US revenue, $5 billion in EU revenue, and $5 billion in "rest of world" revenue if it succeeds -- it matters little which of those is considered your "home" market, just that the total is $30 billion. E.g. Novo Nordisk is making a fortune from the US market despite being based in Denmark, and moving their HQ won't obviously change the picture.
I'm all for lower prices, but I'm not convinced that the success of European pharma companies proves that high prices in the USA are not driving their investment in new products.
That extra margin is specifically *not* going to R&D - it’s going to shareholders.
If the extra cost of pharma drugs in the US were going directly to R&D, it wouldn’t allow them to net profit at multiples of other S&P500s. This implies rent seeking, which IMO would be reduced if they had to negotiate with another large entity.
No need for doomsday scenarios, all these impacts can be measured
"Agree to the government's price or don't sell any of your stuff to or via the government" actually seems like a credible form of negotiation. If anything the tax (rather than complete market denial) alternative is something of a sop to the producers here.
I'm not taking a position on the merits here, but I think this is credibly a negotiation.
That's not what the law says. Instead, it says "Agree to the government's price or don't sell any of your stuff to or via *anyone in the government's jurisdiction*". And then, yes, you get 5% back.
I am absolutely not an expert--nor even really an amateur--regarding the particulars of the IRA's price-negotiation, but I'm referring to the following from the link John posted:
"If an agreement on the maximum fair price is not reached by August 1, 2024, manufacturers may be subject to an excise tax, which will be administered by the IRS, as specified in the Inflation Reduction Act. In the revised guidance, CMS outlined an expedited process manufacturers can follow if they choose to not participate in the negotiation program, which would enable them to withdraw their drugs from coverage under Medicare and Medicaid to avoid paying the excise tax"
I find it difficult to read that in a way that doesn't suggest that manufacturers can still sell to or through private insurers or pharmacies outside of Medicare/Medicaid notwithstanding their being subject to the government's jurisdiction (as US corporations or entities).
I'm not an expert either, and you're probably right about interpreting that quote. If it's accurate, then that does change my opinion on the IRA quite substantially.
I made a comment in reference to yours and I now realize that you were criticizing this method of negotiation, not praising it. Ok, yes I agree with you. I'm 3 hours back.
I think this reasoning implies that the healthcare regimes of countries explains the variance in health outcomes. That doesn't make much sense to me. Health outcomes are likely determined by the variance in pre-healthcare factors given that those vary more than healthcare regimes.
I had an experience that shows how the government can do this much better and also how perceptions of this cannot be separated from political affialiation.
For patients who require continuous home oxygen, most of whom are on Medicare, the government historically set a reimbursement rate by fiat. In 2011 it was a bit over $200 per month per patient. Then they switched to an auction system. Seven years later it was ~$85 per month. It drove massive consolidation in the Durable Medical Equipment (DME) industry that served those patients, typically by undifferentiated mom and pop providers. The data on how it impacted care are murky. It probably diminished it slightly. But the savings were significant.
I'll never forget the (very representative) DME owner who spent 30 minutes ranting to me about how Obama was killing his business with his socialist policies, despite this being a shift from a top-down to a market-based pricing mechanism and the mechanism being created by George W Bush in the Part D legislation.
I agree that the IRA mechanism is terrible and would have preferred a market mechanism. it would have been much better to focus on areas where multiple drugs with therapeutically similar drugs existed and get manufacturers to competete for volume with price.
It's a starting point. Once implemented, it can be applied in an increasingly expansive manner until we successfully stamp out any profit-making enterprise.
It makes sense for the federal government as a very large consumer of say hotels, hire cares or flights, to use its size to negotiate for a better deal. In fact it would be negligent not to.
It makes sense for a government healthcare system as a very large consumer of drugs to use its size to negotiate a better deal. To point to the utility it believes the drug provides, to refuse to pay more than that, and to threaten not to buy the drug if a fair price can’t be reached. This is what many governments already do (say Australia).
The historic ban on Medicare negotiating prices with drug companies is almost a criminal misuse of taxpayer money.
I agree pharmaceutical development is important - that doesn’t mean we should just throw money at incumbent pharmaceutical companies. Lots of things are important, no one suggests overpaying for electronics to spur more innovation in Silicon Valley.
The whole no-negotiating drug prices prohibition was inserted to circumvent litigation and a hold up problem. Like many institutional structures it’s purpose has disappeared yet it persists as a drag on the system.
I think it's because drugs are overwhelmingly provided by for-profit corporations whereas most other healthcare services are provided by non-profits. The public discourse is much less sympathetic to for profit providers. This is probably a mistake though as the non-profits are just as culpable for the overpricing they do and as you point out, are responsible for a much bigger share of the overall problem.
Agree. Taken as a whole, this could be one of the more cost-efficient dimensions of the American healthcare system. We spend so much more on things that provide no net benefit to health outcomes. I confess I was very disappointed with this post given the exciting title. Boosting innovation in the creation of highly effective drug treatments for deadly and debilitating diseases seems like one of the most important things humanity can pursue. It seems unlikely to me that fiddling the requirements for demonstrating effectiveness, thereby improving the risk / reward profile for a handful of drug manufacturing and marketing companies, is going to do much to "boost innovation".
This actually is where much of the leverage is. Drug developers spend an average of a billion dollars for each new approved drug. If that could be reduced by 10x (and it was 10x smaller just a few decades ago) we'd have much more drug innovation.
One important note: this doesn't just effect a handful of drug developers, it affects every single person who is developing a drug.
Indeed, the only other leverage points I know of are (1) improve academic research funding. That's also a great value, but all the academic research is stuck behind the bottleneck in clinical trials and approvals. (2) focus specifically on reducing the cost of a clinical trial.
That's fair, but isn't it more about the “hit rate” for truly effective drugs than the cost of trials per se? It seems far more likely to me that the “lower the gate” approach results in more ineffective but highly marketable drugs going into the stream of commerce, so the prescription drugs industry becomes like the supplements industry, but with much bigger advertising budgets. Opening a pathway for a large number of attractive-sounding but ineffective drugs that can make back their trial costs more easily does not seem like the best way to improve health outcomes. It should be much cheaper for Pharma to weed out the losers before they have to spend hundreds of millions on expansive trials, but that seems more like a technical problem that will be solved with better biochemistry and maybe better AI, which is where I thought the post was going…
Yes, I think those are both good points. The first is what I meant by calling it a libertarian solution. Mind you, the customer in this case is partly the doctor, who is very sophisticated, so this might work well, but there are many details and questions. Companies using AI to make better decisions about what to pursue is definitely a big angle within the industry now. I don't know how successful that will be, but the good news there is that those companies should be highly motivated to do that weeding out, since it's their money on the line. We'll see how it goes.
(I think you're mostly right on the merits, but good luck convincing enough people. Also, the devil is in the details: how "heroic" is "heroic," and who decides? We use some kind of "$$ per QALY" metric like the British NHS, presumably?)
Yes, Congress has passed multiple laws that has lead to programs seeking ways to reduce healthcare costs, such as value based care initiatives from CMS, Medicaid Health Homes, ACO's, etc., creating large datasets for research like NIH's All of Us program, the VA/DOD are looking at ways of increasing physician efficiency to retain good doctors for less money, and on and on.
With drugs, it's a combination of utilization and cost. The cost side is much easier to attack by just not spending a ridiculous amount, it is the low hanging fruit. Just reducing that from 9% to like 7.8% (all else staying the same) would be huge.
At my previous job, my insurer did a thing to encourage regular annual checkups. If you didn’t have a “wellness visit” before June 30, then your premium went up by $30 a month for the following year.
Unfortunately, if I came to the doctor for a routine checkup to re-approve a prescription, the doctor wasn’t able to write that up as a “wellness visit”, and a separate visit needed to be arranged.
In implementation, a lot of these good ideas often turn stupid, and I worry that “discounts for being fit” would do that. (I’ve often seen questionnaires about fitness activity, and they specifically usually ask how many minutes of moderate heart rate exercise you get that isn’t related to your job. I’m sure there’s some research reason that it makes sense to separate out people who get exertion through work and people who do exercise recreationally, but it could easily turn perverse if it became an incentive. I think my bike commutes often didn’t get counted by doctors.)
You can speak to your employer to see if they could get a program like that going, but that's likely beyond the levers available to the government to make that happen via provider/payer incentives.
I suppose Congress could enact something like the Presidential Fitness Test for adults, and the government could cut checks to health people or something, but, as Kenny Easwaran suggests above, that would absolutely become silly for a million reasons.
> "If you’re at very high risk of dying in the very near future from a fatal disease, then you probably aren’t as worried about the downsides of experimental treatments."
"A pandemic reverses the usual asymmetry of risk. Now it is the status quo that is immensely dangerous, and a typical sort of medical intervention (such as an experimental drug or vaccine) is comparatively less so. As a result, we should expect to find many cases in which the potential benefits of innovation outweigh the potential risks. Doing nothing new, and allowing the pandemic to continue unabated, should be recognized as a far riskier prospect—for many individuals, and especially for society at large—than trying experimental or otherwise uncertain solutions for which the risks are orders of magnitude lower than the risk otherwise posed by the virus."
(Of course, it's also importantly true of many individuals in non-pandemic contexts too!)
The FDA process is pretty damn good. One of the issues they have to suss out and why it takes so long is the FDA doesn't do the testing. The companies producing the drug do. There are MANY statistical manipulations the companies are allowed to use in regards to adverse events(aka side effects). One of the primary ones is if a participant is using another medication, has an unrelated health issue, contracts an unrelated disease, the company can remove this adverse event data from the test population. This makes sense on the first read as all these things are not necessarily under the drug companies control when they are running the trial. However, the company doesn't really have to justify it either, meaning they are under no real obligation to prove it wasn't their new treatment target that caused the adverse event. This is why the FDA is so cautious. Basically a drug/treatment usually has to show efficacy above or equal to current treatment technology to be market viable, side effects are huge part of that efficacy math. So companies are highly incentivized to dump adverse effect data under any pre text. As Matt already stated even with 10000 participants you won't catch all the side effects(adverse events), especially longitudinal ones. The FDA process as restrictive as it seems isn't flawless either. Look no further then Fen-Phen, the Opioid Crisis, or problems with host of other drugs that end up making a media splash.
I do think negotiating prices is the right call. Medicare/Medicaid are huge buyer. They should be able to work with pharmaceutical companies for better pricing, especially for life saving small market drugs that often are the target for crazy price increases.
Would it make sense to change that? E.g. for phase 3 trials, you have a government agency coordinating the trials and healthcare providers carrying it out; the manufacturer of the drug candidate is not directly involved, and the FDA will not have to worry that a business with billions of dollars at stake is trying to fudge the results. The studies can be paid for by application fees from the manufacturer, enough for a reasonable Phase 3 trial (e.g. 1000 patients or enough power that the manufacturer confidently expects p<0.01; several visits during the trial, in which you measure outcomes, ask about various side effects, ask if there were any unexpected effects, etc.) The agency can get additional public funding for further studies if it finds some sufficiently interesting unexpected results.
The thought is that this way, the efficacy study will not be designed and managed by people strongly motivated to want a specific result. It might save time and effort that would otherwise go into trying to get away with manipulating the results, and trying to catch any manipulations. And financing it by application fees would mean that the cost of bringing a drug to market, from the company's perspective, should not change dramatically.
There might be additional benefits to theoretical efficiency (and logistical difficulties) if the drug assessment agency were an international consortium. It could test once for the world; recruit 1000+ patient samples each to be representative of the US, UK, EU, China, and the world, and do subgroup and pooled analysis, all to be included in the final report; and the patients can be recruited preferentially in countries where good medical research can be done at a relatively modest cost. The FDA and EMA would then be able to request (or perhaps conduct) follow-up studies if they think more information is necessary before approval.
The national or international medication testing agency could, in principle, also help to coordinate and fund post-marketing, Phase 4 trials to assess things like rare or long-term adverse effects, head-to-head comparisons of different drugs approved for the same indication, maybe studies to look at correlates of why some patients respond better to one medication over another (genetic, metabolic, etc).
I'm not an expert in medical research or government agencies, though I know a little about both, but separating the testing from the interested parties sounds like a good idea to me.
I'm just not sure we want to pay for it at that level but it might be a good idea. Right now the system actually functions pretty well was my point. It's not nearly as restrictive as it's portrayed. It's highly regulated sure, just like financial markets but that doesn't mean it's stifling innovation or costing us more per drug target than the risk mitigation is worth.
It would be interesting to what the cost of your suggestion would be though just gross dollar wise. I suspect that would well I'm excess of any political will to establish it.
Much of the cost would theoretically be borne by pharma companies -- they would pay a huge fee, close to the cost of the clinical trial, for the agency to conduct it, instead of directly working with healthcare providers to sponsor the study.
Yeah similar to the way User Fees are already treated. The problem is that would likely raise costs(in the short term) and put a number of current testing groups out of business or in tighter margins. It's why I said I don't think the political will exists. It's not just about finding the money, it's also about the political will to piss off the current structure of the market.
If you’re saying that a drug company doesn’t need to report an adverse reaction in a study subject because it believes that it was due to another drug or some other cause, that’s false. They can make that argument to FDA but they still have to report it. FDA is usually agnostic about causation in controlled studies because such unrelated events are equally likely to occur in the control group so they cancel out in comparison.
It's not false. You nicely point it out right the end of your comment. The FDA is agnostic to causation in controlled studies because unrelated events are equally likely to occur in the control group.
A participant reports an adverse event. The group conducting the study then determines if that event is cause for removal of the subject from the study. Then it is determined if the data can be retained and used or not. It's very possible that the data just gets dropped from the final submission. Are they supposed to do this? No. Do they do it? Yes. Scientific American had an article about it a few years back. In some cases it just reported as drop, in some cases it's not reported at all.
This may have been the case 20 or 30 years ago, but procedures have tightened considerably. Investigators are required to keep enrollment records. It's illegal to remove or drop records. And FDA routinely conducts investigations.
It was still true as of 2017. Especially as more trials we're done outside of the US. Again, I'm not really faulting the FDA. I think they do a great job. I'm just saying the economic incentives highly inform the process and can lead to massaged, manipulated or outright manufactured data. Especially for a drugs that are in tight market or late stage trials. They represent a huge investment and cost especially for the smaller companies.
I don’t know where you’re getting your information; maybe the Scientific American article was from 2017 but references incidences from years earlier. I have been working with this data at FDA for 20 years and never seen anything like this or even how it would be possible.
This is the chicken and egg of trial cost verus trial quality.
Given pharma's incentive to frame data in a favorable way, the more transpareny the better. The more endpoints the better. But all of this drives study costs up.
There’s no additional cost to reporting everything. In fact, spending time deciding and explaining whether or not an adverse event is due to the study drug is actually cost and generally a waste of time and money for the reason that I explained.
As far as endpoints go, FDA requires endpoints to be specified in advance to prevent biased framing.
The cost is in the data collection, especially when collecting the data requires more patient encounters, additional tests, and makes recruitment and study completion harder. Agree that the data analysis is trivial.
As far as consistency goes anyone who argues that women have a right to bodily autonomy and hence that the government can't prevent doctors from giving them drugs/treatments to enable abortion should also believe that same justification protects the right of patients to try experimental or untested medications.
Sure, maybe you can justify restricting genuine snake oil consistent with that moral position but the underlying principle is that when deciding what kind of medical treatments you are allowed to get the government doesn't get to tell you how you should weigh values/risks be they the moral status of the fetus or the chance you paid for a medicine that just harms you.
I mean if your bodily autonomy doesn't let the government stop you from legally accessing drugs to remove a fetus why does it let them stop you from drugs designed to remove a tumor? What if it's an effective drug but just hasn't passed the regulatory hurdles here? What if abortion drugs only worked 50% of the time, 85%, is there some threshold of probability of success at which it becomes moral for the government to substitute their judgement re: risks/benefits to your body over yours?
--
To be clear one can, as I do, support abortion rights for good old fashioned consequentialist grounds but it's the attempt to claim a special moral status that's dubious.
I 100% agree with your first paragraph, which is why I also roll my eyes at all the lefties who claim that the U.S. Constitution mandates anarcho-capitalism for a handful of specific medical procedures/products.
(Seriously, I would love if the Constitution contained an express "right to privacy," but it doesn't -- I say that as someone who has co-authored a book on privacy law -- and if you're going to interpret the "penumbras and emanations" of the Constitution to create such a right, then it should at least be applied in a logically consistent manner. If there's bodily autonomy for abortions, birth control, and puberty blockers, then there should certainly be bodily autonomy for taking any other sort of medication you want to take, at least so long as there's full disclosure about the experimental/untested status of it.)
I don't per se disagree but I'd argue, a la Balkin's originalism, that the OPM of the constitution included an understanding of the courts as engaged in a common law practice of substantively explicating what the law means -- even to the extent of finding new rights (I mean the rule of lenity, suppression, and virtually all 1a law aren't written in the constitution and the later was clearly not OPM either).
But what matters is coherence. What makes the court not a legislature is the need to follow or overturn precedent. So I'm fine with the court extending Griswald to Roe but it's got to take the principal seriously and apply it to drugs generally etc.
I think it's wrong for the government to forcibly extract someone's kidney and give it to someone else to save their life. But I still think the government can ban medical treatments that don't work. (It's easy to say "patients should be allowed to make their own decisions" but that doesn't really address the institutional problems you create when companies can make profits by pushing treatments with weak or no evidence.)
Exactly the same thing goes for abortion. A person's body is not a public resource that can be forcibly exploited for the benefit of others. But I don't have a problem with regulating abortion methods that don't work or are particularly unsafe, or imposing malpractice liability on doctors who botch abortions, etc. The trouble in that sphere is that most of the safety concerns turn out to be pretextual trumped-up stuff driven by motivated reasoning.
But that's not what the government is literally doing when they ban abortion. If you break it down they are banning the use of certain medications and surgical treatments because they judge the harms exceed the benefits. Sure, making it literally murder is a different level but just banning abortions requires no more than saying those drugs/surgeries are not approved for that purpose and no off label.
Sure, when they judge those harms they are substituting their values about the fetus's life for your values. But how is that different than when they come in and say: nah if we think the evidence says it's only a 5% chance of a cure but a 20% chance of a nasty side effect and a 95% chance of losing your money you can't legally take the med any different?
You're looking at it at too high a level of abstraction. The intent of an abortion ban is to prevent someone from terminating a pregnancy. The intent of medical regulation is not to prevent someone from extending their life. It's literally true of course that banning a medical treatment infringes on "autonomy" in that you can't choose to get the treatment, but the substantial autonomy interest at stake in abortion is not the specific procedure but the outcome. The same thing is true of life-extending drugs (the autonomy interest is in being able to choose to do treatment to live longer) and precisely for that reason regulation aimed at effectiveness is different in kind from regulation that rejects the goal as such.
Ok, suppose I want to be in the state of being really fucking high -- not just drunk but feeling the intense state of reward provided by many hard drugs. Now you can't argue that the government isn't systematically trying to block all the treatments that put my brain into that state. Fuck, it's specifically legal to use many of the drugs that produce this effect as long as that isn't the purpose of the use.Sure, maybe it's legal to be happy but that's analagous to it being legal to have a miscarriage.
The whole idea of the abortion autonomy debate is that the individual gets to say: I want to change my body in a way that wouldn't happen without intervention and you don't get to tell me I can't.
--
But even if we leave that issue aside I think you have another problem. Your changing the argument from a right to bodily autonomy to a right to be free from interference with a certain kind of intent.
I think it's going to be very hard to formulate any such principle in an attractive way. Exactly what is the kind of intent that triggers this right and why is that an attractive principle.
The appeal of the autonomy argument was always the simple: but I get to decide what happens with my body. Once you start stacking on extra qualifications it's no more persuasive than just saying I think this particular thing should be allowed.
--
However, if you can formulate a compelling principle that rules out the examples I raise that would be the kind of thing that could convince me. Like tell me what exactly it is that one has a right to.
I don't think it's true in reality that bans on hard drugs are about preventing intense highs--they're primarily about the other effects of those drugs. You can take some of them for medical reasons but that's because we think the dangers are worth it in those cases. But you're getting the point--a ban on drugs because they produce intense highs directly implicates an autonomy right to have an intense high. I'm not sure such a right exists though.
Intent does matter a lot to rights. An incidental burden is different from a direct burden. It's one thing for the government to say "you must pay $20 to go to the Legalize All Drugs rally" and another thing for the government to charge you the normal highway toll to get there.
It sounds to me that you are getting stuck on the difference between a slogan that you want to read to its furthest semantic potential ("my body my choice") and the actual views non-libertarian pro-choicers hold. But that's just to say that reality is always more complicated than slogans. The case for abortion rights has a lot to do with autonomy involving the body but it doesn't hinge on or require thinking that every choice involving the body in some way is equally protected. I'm sorry you don't find my argument appealing but the idea that there's a difference between safety regulations on medicine (or bans on hard drugs), on the one hand, and bans on terminating a pregnancy, on the other, is obviously very widely accepted.
Sure, they don't intend to stop you from getting high just like they don't intend to stop you from becoming non-pregnant -- just block the associated consequences (addiction/intentional death of a fetus).
And look I'm a huge fan of abortion rights on consequentialist grounds and "my body my choice" is a fine slogan but people frequently advance the bodily autonomy consideration as both a serious moral and legal argument (though my point would be valid regardless -- but only relevant to those who do accept this argument). For instance.
1) See Thompson's violin argument. A famous philosophical argument claiming to argue for abortion rights on something more than consequentialist grounds. This is covered frequently in college philosophy classes.
2) The legal argument for Roe has always turned on the notion of 'privacy' as articulated in Griswald which really means something very close to personal autonomy.
---
And sure, I'm happy to admit that you can understand the notion of autonomy in a more consequentialist fashion -- reproduction is especially salient and important to women's self-identity and ability to fully participate in careers etc like men do. And I believe all that.
The problem is that once you admit it's this muddy notion that turns on disputable value judgements (it's important that women can compete for jobs with men, it's really important that people get to control their reproduction etc) then you can no longer blast supreme court decisions or people who go the other way as acting illegitimately or similar invectives. You can strongly disagree and wish they brought your values to the table but it's no longer an illegitimate deciscion.
And while it's a bit off topic I think that distinction is hugely important. Trump trying to steal an election was illegitimate. The supreme court going the other way on Roe is a legitimate result of our country's system of government that is dealt with by winning enough elections for long enough to change the court (or just passing a federal law protecting abortion).
No, that's a perfectly cogent consequentialist argument for treating the positions differently. As I said, I accept such arguments.
My point is that you can't short circuit the whole abortion debate and push the anti-abortion position from the board without getting into the messy consequentialist details unless you are willing to do the same thing with drugs generally.
The existence of some other reasonable argument isn't at all helpful for the whole right to bodily autonomy position.
Should the same agency regulate safety and efficacy? I could imagine FDA regulating only safety and what efficacy claims could be made and Medicare using CBA to decide whether to pay for and how much to pay for new drugs.
I think that there's something to that. I don't think that you can totally get the FDA out of the efficacy business. You don't want FDA approved child leukemia drugs that do literally nothing. But the FDA approval doesn't do enough on relative efficacy of different treatments or sub-populations differing efficacy. And drug approval isn't the place to do that, but right now there doesn't seem to be enough incentive to do those kinds of efficacy tests post-approval, so a price negotiation CBA is an interesting approach.
Yes. I've hear that argument since at least 1972 and it is so Econ 101 that I doubt it was new then. And yes, there is the bureaucratic "embarrassment" effect (the same reason the Fed won't reduce the EFFR until they are SURE they will not have to raise it again in a month or two), but is that really the reason not to use cost benefit analysis, or rather to use cost benefit analysis in which bureaucratic embarrassment figures among the costs?
After all, FDA is not the only regulatory agency that does not use cost benefit analysis or we would have an XL pipeline more nuclear power plants and infill commercial and residential development. Did CDC use cost benefit analysis or provide information so that public decision makers could use CBA in deciding when and for how long and in what conditions to close schools, and restaurants and churches?
Now I'm not opposed to looking for agency specific solutions an OMB memo to all agencies saying "use CBA in making decisions" would not work, but at the same time we ought to think about systemic reasons for regulatory "misbehavior."
Oh that is good news. I assumed it was something much harder like bureaucratic culture, outdated attitudes or something like that. CBA is something that we ought to be able to get bipartisan agreement on.
I'm curious since you post a lot about these things. In what sense are you aware that cost-benefit analysis is used to inform regulatory policy at all? I see things like CBO estimates and that sounds like CB analysis I guess, but I'm not sure.
Ehhhhh... there are a lot of clinical trials run at least partially at foreign sites already. Any gains in price reduction there would be marginal.
And I'm not sure your hand waving of the innovation argument holds water. Drug discovery has a lot more upfront costs and failures that hotel chains do! Companies will not engage in a research project if the market isn't there (in fact, I've been on projects where the research has gotten pretty advanced and then the company has pulled the plug due to changes in the market).
"it's a strong argument for the FDA to NOT approve drugs without strong data since payers will be encouraged to take on exorbitant costs for no benefit."
No, it's not. Let people have the right to try. If necessary, one could devise a system in which cash payments for experimental drugs are fine, but insurance won't cover them.
One example of a drug that doesn't work, doesn't obviate the argument that those of us, like myself, with fatal diseases shouldn't be able to try.
I'll say up front that I am extremely sympathetic to you and anyone who suffers from a disease with no available treatment, or in the case of many cancers, inadequate treatment. It's literally the reason I do the job that I do (drug discovery research).
Having said that... the industry is littered with drug candidates that looked amazing in Phase II only to fall apart in Phase III. You need to run the trial to get the statistical power to determine both efficacy and safety. For cases of extreme unmet medical need, the FDA can approve on a conditional basis before Phase III is complete, and they've been doing so more regularly in response to patient advocacy groups.
Relyvrio is instructive because it shows the downside of that approach when the underlying data is questionable (and make no mistake, it was). This wasn't even a case of a company with years of expensive research seeing a payoff... it was literally taking two generics together and seeing an effect. Which would have been incredible if it had worked! But it didn't, and payers had to cough up a lot of money for the drug in the meantime, and those costs get spread to everyone in terms of higher premiums. You can argue that insurance should be decoupled from FDA approval, but then they'd need to have their own scientific review boards and that just adds time and expense to the whole process.
They didn't get the emergency use authorization because they screwed up the statistics on their original trial. I don't remember the exact timing but at some point the Pfizer and Moderna vaccines got full approval, meaning the EUA pathway was no longer available. And then they eventually withdrew their application.
I remember that there were significant issues with their original application. But I never understood why they never got full approval, the way that Novavax and perhaps some other latecomers eventually did.
"So basically whichever country in the world gets the best deal on the price of a drug, the United States has to also get." Your HIV drugs in Africa example is right as an example of why this is dumb. But would it be less dumb if a purchasing power parity formula was used to set a global price?
This is called price-refernencing and it aready exists, just not in the US. The result is that nearly all drugs are launched outside the US in the same sequence of countries. Every drug company has spreadsheets that solve this pricing problem. If the US did this, expect the sequence to change and US launches to get pushed back. Since the US is still the biggest market (higher usage, even if the EU, India, and China are bigger by population), the strategy would be to set the highest price in countries that the US references to maximize the US price.
Because the US contributes the greatest share of drug profits today, many development proejcts would be cancelled and innovation would suffer.
I generally think that Econ 101 arguments don't translate well to healthcare, but this is a case where they would.
I think global prices for drugs would be very good, but a global price would have to be higher than what Europe, Canada, Japan, etc. pay for drugs currently, so they are not going agree to such a global price regime without a lot of kicking and screaming.
This article is missing any discussion of whether the pharmaceutical industry is making outsize profits or margins. Certainly seems that way when Semaglutide blows up for Novo Nordisk, but what about all the failures. I’ve heard that pharmaceutical profits aren’t all that impressive compared to other industries, and that needs to be discussed before we engage in aggressive pricing controls.
Windfall profits will exist as long as pharmaceutical patents grant an unregulated monopoly. This JAMA article says pharmaceutical companies have average profit margins of 13.8%, which is 79% higher than the 7.7% average for other large corporations. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7054843.
Also ... why do I care - at all - about 13.8% profit margins? Seems shockingly cheap for the risks inherent in their business model. Facebook's net profit margin is up to 35%. I can find companies > 30% all day just due M&A consolidation and massive barriers to entry.
Other commenters said most of the cost seems to come from Phase 3 trials.
At least this limitation:
===============
First, this analysis focused on large, fully integrated pharmaceutical companies that generate revenue and profit primarily from drug sales. It did not consider small or midsized biopharmaceutical companies or biotechnology companies engaged in discovery research or early-stage development, which typically have little revenue and negative profits (losses).18 As such, the pharmaceutical data set was not representative of the broad biopharmaceutical industry and the results cannot be extrapolated to the industry as a whole.
=================
would be less relevant in that case, as I _assume_ the smaller companies in early-stage development are mostly doing Phase 1 & Phase 2 trials.
That's a fair question - the limitations section (the only section I looked at, after John from FL's comment) says:
==================
Third, this analysis did not consider whether companies in the pharmaceutical data set had excess profit. In accounting and finance, excess profit is defined as profit over and above a “normal” return on capital invested in the company—a return that is commonly associated with the risk of the investment. Future research can be directed at examining the relationship between investment risk, returns on capital investments, and reported profits.
===================
So I don't know the answer, and they didn't consider it.
An interesting political problem here is that a lot of voters/activists see pharma companies as profoundly unsympathetic - in many ways analogous to the “big developers” you see as housing boogeymen opposed by activists who ostensibly want a non-market solution. In practice, this leads to policies that make the cost/benefit for drug development worse without actually addressing any of the approval-related costs you mention. For example, the IRA recently reduced patent timelines for small molecule drugs, which, without any other reforms, just makes it less economical to develop them.
When you’re developing a drug, you’re sort of making a leveraged bet with a really long horizon whose success or failure likelihood can’t really be quantified - the signals appear way too slowly. A few years back, there was some interest in an X-Prize type system where you’d get a large up front payment for a successful approval (or reaching some milestones), then forfeit some of the longer-term monopoly rents. I think this has fallen out of favor, but I’m not really sure why.
For small biotechs, this is how a lot of the financing actually already works - you typically do deals with big pharma partners where you’d get payouts depending on whether or not predefined milestones are met, both pre-clinically and clinically. So I think a lot of the biotech ecosystem would be compatible with such a system, assuming it was sufficiently transparent.
First, the US has done a much better job over the past decade at supporting smaller trials for regulatory submission when there is unmet medical need and/or high burden with or without terminal risk. In oncology and rare disease, Phase 3s are often skipped altogether. The offset, however, is that it's harder to find patients and run the trials on a per-patient basis, and it can take multiple Phase 2s to get the trial design right in a disease that is not well understood (rare) or where the basic science is in rapid flux (oncology).
Regarding international trials: it's not just the patients and their genetics. It's the diseases they have (or don't), which are VERY different. It's the practice of medicine and the ability to actually execute a modern protocol with modern diagnostic practices, which sometimes involves specific medical training and sometimes also involves a great deal of technology and infrastructure, which can be vary greatly. There are many variables. The US and Europe are like siblings. Things are close enough. But the rest of the world is more like cousins (in the flexible sense of the term.. might be first cousins, or second cousins once removed, third cousins...).
It's a time value of money problem and a patent enforcement problem. Venture capital wants to make big investments in biotech today (yay!) in exchange for a big possible payout upon success (boo!) and unfortunately, that currently often relies on pricing in the US. The best way around this? Wait a few years until the patent exclusivity expires, and actually enforce it without the IP shenanigans that we so often see from big pharma who eventually acquires the successful IP.
" If you’re at very high risk of dying in the very near future from a fatal disease, then you probably aren’t as worried about the downsides of experimental treatments."
This describes me: https://jakeseliger.com/2023/07/22/i-am-dying-of-squamous-cell-carcinoma-and-the-treatments-that-might-save-me-are-just-out-of-reach, and it's why I've been beating on the FDA-reform drum since realizing that I've run out of approved treatments: https://jakeseliger.com/2024/01/29/the-dead-and-dying-at-the-gates-of-oncology-clinical-trials/
There's little real downside to me being harmed or even killed by experimental treatments: recurrent/metastatic squamous cell carcinoma is already virtually always fatal. The FDA is "protecting" me from harm, and the result is going to be me dying from disease. This is nonsensical and yet standard practice at the FDA.
As it is, I may have to go to Mexico to get intratumoral immunotherapy injections that aren't formally approved in the U.S., but should be, given that the immunotherapies themselves are: https://williamscancerinstitute.com/.
Best of luck to you and your family. God Bless.
This is so important, thank you for sharing.
Thank your for commenting and putting a human face to MY's argument. I wish you the very best of luck and applaud your advocacy. Godspeed.
Godspeed. This is so hard to see happening
Praying for you man
Thank you! It's been and continues to be a challenge. I wish I weren't in line to be the face of the otherwise invisible graveyard, but that's the way things are going.
Thanks for commenting. I wish you all the best in your treatment.
I think this analysis is good on the merits, but misses key details. My qualifications: I am a PhD in Cancer Biology, currently still in academia but have done an internship in a biotech company and have taken coursework related to the economics of biotech/pharma.
Matt's analysis focuses on "safety", but that is not the key driver of cost in clinical trials. Safety concerns are mostly addressed in Phase I and/or II trials with relatively small patient cohort sizes. The larger driver of cost are Phase III efficacy studies, where larger patient cohorts are required in order to have sufficient statistical power to distinguish response rate. Critically, the FDA requires these large, expensive studies to prove that the trial drug is *better* than standard of care. Merely matching, or very slightly underperforming standard of care is a failure of the trial. These Phase III trials cost tens or hundreds of millions of dollars, take years of time, and are (to my knowledge) the main driver of drug development to the point of dwarfing all other costs. And, as MY briefly touches on, a pharma company has to price in the cost of each failed trial into a successful drug. (A new drug doesn't just have to pay for its own development costs; it also has to pay the costs of all the failed drugs the company has not gotten past the FDA. Otherwise the economics of developing new drugs doesn't work out.)
While it is possible that a drug fails in Phase III due to previously-undiscovered safety concerns (patients end up advancing in disease, or having toxicities not seen in the Phase I/II), the more common failure is due to lack of sufficient efficacy. (Which is what you'd expect from an efficacy trial). There's obviously enormous benefit to society to have drugs on market that are equally effective to standard, or perhaps even slightly less effective. No clinical trial is a perfect representation of the population, so even a "less" efficacious drug might have use in certain patient populations. The pharma company could even price the "worse" drug at a small price. Patients could decide themselves to take a risk on a cheaper drug, and the company would still lose money, but far less than they'd lose if they had to scrap the whole program. Win-win. The merits of an "equally" beneficial drug are even more obvious. More market competition! Even relaxing the rules would allow more drugs to come to market (and thus change the cost/benefit equation of drug development) with very limited risk to society, in my opinion.
None of this even touches on the biotech-vs-pharma roles, how VC feeds into the ecosystem, etc. Biotech is heavily dependent on venture capital, which has dried up in recent years due to increased interest rates. Nearly every drug is brought to market by pharma, but nearly every drug begins its lifecycle in biotech, and IP is then acquired by pharma (for reasons to lengthy to get into here). Biotech has been in a recession since early/mid 2022, and adding increased pressure right now with price controls in the IRA has really worried the industry. There's so much more on the "development" side that MY either doesn't know or didn't have space to touch on. The US really is the envy of the world though when it comes to medical innovation, and that really is driven by market economics in a way that could be greatly disrupted by pricing controls. Further changes to price controls need to be paired with supply-side reforms (like FDA regulations, etc.) if we want to continue to see robust drug development.
Thanks for tuning in here! Doing research for this article, it was hard to find good reforms for the FDA approval process. Are you aware of any that we missed?
Nope! I think everything in the article was accurate as written; I just don't think that "safety" concerns are the leading driver of trial cost. Obviously it would help the economics to relax safety concerns, especially for terminally ill patients that have no better options. This has been done in the past with HIV/AIDs meds, and I believe with breast cancer as well. Those efforts are well documented and resulted from populist pushes from patient communities (and associated activists) that wanted change.
Most of my education and experience has centered on understanding *why* costs are so high, and why high prices are necessary to meet the cost of development. It's successfully converted me from a stance of "pharma companies are bleeding the taxpayer dry" to "actually while some (ie Martin Shkreli) are leeches, many drugs that get bad press are actually priced to the market". Unfortunately, I have never been part of any activist or political communities on the subject, so I don't have any experience with "ok how do we change this?". I don't have any well-researched stances on reform. I took a course taught by someone who had founded a successful biotech company, sold it, and then used the profits to found a VC firm focusing on funding the biotech sector. He stressed the "superiority" in efficacy part of drug development, and he claimed (logically) that an ever increasing burden of superiority would lead to an ever increasing cost of development, as all the "low hanging fruits" got picked. That's why I focused on that topic in my comment, but I should have clarified that this is coming secondhand from an industry expert who likely didn't do any truly rigorous policy analysis on the subject.
What I failed to appreciate before taking that course was just how incredibly labyrinthian drug development is. The relationships between pharma, VC, and biotech are very complex, the risk is far higher than most industries would tolerate, and VC/pharma needs to be leveraged to the tune of hundreds of millions, if not billions of dollars, for many years before even sniffing the possibility of an ROI. The IP clock starts from patent filing, not from drug approval (as I understand it), so the length of the drug approval process plays a huge role in subsequent drug pricing. I was taught that a drug typically has a 5-10 year window of profitability before generics take over the market, in which time the company needs to recoup all of its losses, plus make enough money to keep new R&D alive. It really feels like the alluring prospect of multi-billion dollar/year blockbuster drugs is the only thing that keeps the process moving.
Thanks for this explanation of how VC money has infiltrated R&D. This happens in pretty much every industry. Large companies used to have to fund R&D themselves, but have effectively outsourced it to VC/PE over the last 20 or so years due to quantititive easing creating. Why absorb the risk when you can pass it on to the finance guys to fund these biotech companies with great board seats for the semi-retired corporate executives to sit on? Everybody was winning until the money dried up when interest rates hit 5%. It's not a bad model, but what I would like to see the government get into the game as a VC entity to support R&D efforts that are in the national interest. I think the Obama administration was heading that way.
Certainly they could do a *lot* more, but there is at least the SBIR/STTR programs that are congressionally mandated to provide money to early-stage startups. It's only up to ~$1m, which is just seed-round-level funding though.
The biggest hole in the article is it doesn’t explain how the prices medicare pays are determined now
I haven't seen this elsewhere, but I'd like to see a law passed requiring the FDA to use a consistent efficacy/risk calculation across therapeutic areas. I'm not saying that a new cancer drug should be held to the same risk level of a new rash treatment cream, but that the balancing process should be standardized.
The medical experts at FDA come from the therapeutic areas they are juding. Medical oncologists for oncology drugs, ophthamologists for eye drugs. The culture of cost/benefit is very different across these fields. Then add in the non-FDA advisory groups who exacerbate those biases.
The result is that oncology drugs get approved with risk/reward ratios that are much less favorable than drugs for untreated diseases that get denied.
Having the FDA do cost benefit analysis (on its own process) is one thing I've seen repeatedly. Glad Matt mentioned cost-benefit and hope to see more in a future post some day.
Reducing the cost of clinical trials is often mentioned, and it seems like the government could help reduce per patient costs by investigating what costs (recruiting, administration, etc) and then building capacity or changing regs.
#1 thing: bring back low interest rates.
Dedicate an NIH lab to replicate promising academic results to save time/money pursuing findings that do not replicate. Right now companies do that and do not publish it, so a bunch of companies do it.
There is some good work where different companies share a placebo group. So one placebo group for two drugs instead of two if the trials were separate. Saves a bit.
Ethically fraught, but you could consider letting academic hospitals give terminal patients tool compounds to test new targets/strategies for their disease. Maybe there is some leeway to do this now, but for whatever reason I see very little data from it.
Often mentioned, but get regulatory requirements from around the world harmonized better.
I was just about to raise this point myself. I'm involved in the design and conduct of phase III trials, and at least for my therapeutic area, efficacy concerns determine sample size - not safety concerns.
That said, there are some aspects of phase III study design which are geared towards addressing safety concerns - such as the prevalence of open label extensions.
What about the price per patient? I would have guessed that average P3 study sizes are going down just becuase more drugs are targeting on orphan diseases where smaller studies are needed for a variety of reasons. But the cost per patient has risen dramatically. What do you see in your world on both front?
Why cost per patient could be rising -- some of which could be improved by the FDA
- lengthier study periods to demonstrate efficiacy. Sometimes this is due to the condition being treated, sometimes due to FDA having a higher bar to how efficacy is measured
- Requirements for more diverse study populations. I think this is a good thing, but it really drives up costs
- More follow-on studies are needed. A lengthy Phase 4 study can add a lot of cost and really should be considered as part of the Phase 3 registration cost. This is a bit complex since pharma will do extension studies both becuase regulators require them and for commercial reasons
- More byzantine documentation requirements. Probably a clear opportunity for FDA
- More secondary safety endpoints required.
I’m not very familiar with the price per participant side of things.
One thing I’ve noticed is just how comprehensive some studies are. Even “confirmatory” phase III studies don’t just have primary and key secondary efficacy and safety endpoints - they have lots of exploratory endpoints as well. I feel like the ease with which companies can collect, store, and analyze huge amounts of data makes it far more financially beneficial on average to over-collect, rather than potentially under-collect, data.
it's hard to disentangle which exploratory endpoints are there for commercial versus regualtory reasons. Often they are there at the reuqest of the marketing team wanting to maximize what they can promote with and to set the stage for future clinical research not done by the pharma company
Interesting! You probably know way, way more on me about the subject, then. Do you agree with me that relaxing efficacy benchmarking to “non-inferiority” would significantly impact the economics of therapy development?
I'm not sure.
Firstly, it's not always the case that an active comparator arm is required; there are some drugs for which superiority to placebo is sufficient for approval. Obviously for those drugs it wouldn't make a difference.
Secondly, although I'm fairly removed from the marketing and finance side of things, my understanding is that "not worse than the standard of care" is not a very compelling angle from the sales perspective. I think you'd still see studies designed to show superiority to SOC in some aspect - fewer side effects, better effectiveness in a particular subpopulation, etc.
I agree. Clinicians lean on the long-term, real-world safety data of standard of care and are reluctant to switch to something new based on non-inferiority, unless, the standard of care is significantly safer (non-inferiority is a measure strictly of effectiveness). Take warfarin: decades of data, but known issues. If the next gen anticoagulants can beat it in non-inferiority, they will win, but this is an unusual case. For the most part, as a pharma company, investing in the marketing and sales of a drug that has only proven non-inferiority is a losing proposition.
Prescribers take on significant malpractice liability risk when they prescribe something inferior to standard of care. It's a non-starter in the US
Does the FDA actually itself require a new med to beat standard of care, or is that just an artifact of IRB ethical guidelines? My understanding was the latter, which is why you have a lot of “me too” drugs for non lethal ailments. eg: A new hypnotic doesn’t have to beat Lunesta to be approved for insomnia.
Good question! I am not sure to be honest. I have a strong memory of discussing the superiority requirement as an FDA regulation, but you’re also right that there’s a whole class of “Me Too” drugs that just hop on a winning horse. I suspect that if a given MoA has been proven to the FDA, then the regulatory burden changes. For example, if you make a new molecule that inhibits the same target, and can prove that it acts similarly, it would still be a new IP (a novel compound) but the same mechanism. I imagine that could be a different approval route through the FDA than a totally new treatment modality. I also believe that once the patent has expired people can make generics without significant FDA approval processes.
it really depends on the drug. IRBs and FDA are using the same general framework, although it plays out very differently. Both are weighing the clinical benefit versus risk of the drug for a study or for broad market approval.
The higher the risk, the higher the efficacy hurdle. If a drug could have worse side effects than standard of care, it will need to promise better efficacy to move forward -- non-inferiority won't be enough. If a drug has better safety than standard of care, it might be possible to do a study that doesn't show non-inferiority.
An investigational drug almost always has higher risks than an approved drug solely becuase it hasn't been studied as much. That's why you seldom see them held to a lower hurdle than the standard of care
Also, FDA essentially is the IRB for clincial studies that will be used to support approval. IRBs are only the sole judge when it's a post approval study or, rarely, a pre-approval study not used for registration.
Your understanding of safety is constrained. It’s not just finding terrible toxicities, it’s also establishing a profile of adverse reactions that are not so serious as to prevent approval of a drug, but are necessary for clinical judgment as to when to prescribe a drug or recognize when adverse events occur that are due to that drug
This comports with my understanding of the situation. Let the FDA sort out the safety issue and let the market sort out efficacy, which is a much more complex informational problem.
I think if I have cancer I'd prefer not to be part of the market that "sorts out efficacy" but would rather like to know if this drug is going to help me.
That's why company-funded trials would still be necessary! No one will take a drug that is totally unproven. And likely the FDA will always need to play some role in ensuring efficacy. However, the approval process being simplified or streamlined could be better for everyone: pharma companies, patients, the taxpayer, etc.
You might be surprised that, in my experience, cancer patients tend to disagree with you. Critically ill patients with no alternatives often feel that the FDA is standing in the way of them obtaining new experimental treatments. (See the top comment on this article from jseliger.)
I understand that desperately ill patients are willing to try anything with any chance of working. That's why the FDA allows humanitarian exceptions. I would just not leave the whole efficacy testing phase to the market. As we obviously agree.
“ You might be surprised that, in my experience, cancer patients tend to disagree with you. Critically ill patients with no alternatives often feel that the FDA is standing in the way of them obtaining new experimental treatments. (See the top comment on this article from jseliger.)”
Yes, that’s how you get snake oil sold to desperate people. You could easily be giving toxic meds with no hope of working, so there has to be some process for these exceptions.
The issue is that most things are not cancer with months to years of life expectancy. They are "I have a disease and I need a medication to prevent me from having a heart attack or stroke, or ending up on dialysis, or needing something operated on or transplanted in 10-20 years." It's not even as straightforward as "does this normalize your blood pressure, heart rate, glucose, or cholesterol". Most things are even more complex. How do you have the market sort that out?
My prior is that markets are better than planning and that more better cures for cancer would be found under a market regime than not. A more risk-averse person could simply not partake until doctors were positive something worked. A less risk-averse person would make a different choice.
This is absolutely not the case in pharma based on the incentives of two central actors:
1. Pharma wants to minimize clinical study risk while maximizing commerical returns. In practice, drugs with better efficacy and higher risk could get quashed in your scenaio.
2. Insurors are dis-incentivized for breakthough treatments to come out. They do best in a market with lots of me-too drugs. Look at the Ozempic, etc. weight loss drugs. These have the potential to generate unbelievable health and economic value, but payors are dragging their heels because of the short-term P&L hit. Payors will resist anything with less than a 2-year payback, sometimes even shorter.
Re 2, my layman's understanding is that this is overblown and that companies will frequently produce drugs that put other drugs out of market for short term gain. The entire point is that companies are in a prisoner's dilemma with respect to themselves. They can't just coordinate to hold off on releasing better drugs that actually solve an issue, because each actor is incentivized to generate profit where they can.
That just isn't remotely the case.
One of the biggest frustration within pharma is the behavior of payors, especially how indifferent they are to rewarding innovations that improve health.
There may be a bias in the market-generated information on the drug's benefits, because those who benefit will send signals that, hey, this is a great drug, and the ones who die will be, well, dead.
Shout out to all those ancestors who figured out what was poisonous.
"Welp, Thog's dead. I guess don't eat *those* mushrooms. Let's see what these red berries do....."
Yea people die all the time the question is what system saves more lives than not. The idea is that if we feel like there is a market failure in information generation on the part of safety, then we can have the government regulate that. There is not a clear public interest in regulating efficacy.
You seem to be discounting the chance of active harm. Cancer medications are often not fun.
Fundamentally, my prior is that markets are better than planning. Doctors will likely not prescribe drugs with a bad reputation, and patients will hold out for drugs that are more proven. Some people will choose the riskier option.
My prior is that this will produce higher aggregate welfare, since markets are better than planning. That means they produce better results, and in this case better treatment. Regardless, safety could be regulated if people think there is an information problem there. I'm not an economist so I don't have a strong opinion on that.
Doctors get their info based on how big the tits of the sales rep are; it’s important to have this space regulated.
You mention biotech -- would it help if the federal government invests directly in biotech companies? Either with subsidies and tax incentives, or (more appealing to my gut), with something like a sovereign wealth fund acting as an investor, and standing to profit from biotech companies that succeed in business?
It would help if Biden wasn't proposing a rule that makes it less likely for biotech to partner with the government to use science based on federal funding. (basically threatening to take intellectual property)
Also, just more funding for research, and funding to help pay for promising trials would help
>basically threatening to take intellectual property
That sounds vaguely familiar, but I'd appreciate a link or details anyway. I've heard of proposals for something like that for COVID vaccines -- preventing infectious diseases has particularly large "global public good" effects, so you can make a case for wealthier people to have to subsidize them, but there are perverse incentives with the subsidies coming specifically out of reducing the value of developing future medical treatments.
https://www.latimes.com/business/story/2023-12-13/a-new-initiative-by-biden-that-might-finally-bring-drug-prices-down
In my opinion it's not a legal attempt as the law has never been used in this manner, and the authors of the law state that it's not supposed to be.
But if successful it would mean that biotech companies would no longer use government research. Of course the whole purpose of the law was to get companies to start using government research so it didn't go to waste.
I don't think the government could do this well - it's hard enough for VCs to know what to invest in - and would probably result in a backlash. Everybody remembers that the Obama administration loaned money to Solyndra and it went bust; nobody remembers that they loaned money to Tesla.
I'd rather see massive government investment in foundational work that makes clinical trials less expensive. Instead of saying "foreign data are bad," build models that can help pharma recruit non-US patients who are collectively representative of the US population.
I know governments also invest in basic biology, which contributes eventually to future applications. Maybe expanding this investment would be a good way to boost drug development 10, 20, 30 years down the road? Though that's a long time horizon, too long to pay off for today's incumbents. And it may be bottlenecked by the availability of good researchers.
I like your idea of models to recruit non-US patients. Making clinical trials cheaper will pay off sooner than advancing basic biology.
I imagine that better knowledge of biology might have more dramatic effects in the long run -- a better understanding of the mechanisms of diseases, the molecular biology and cellular targets that you can go after, etc. might not just suggest new drug candidates, but maybe also let companies start out with better candidates for early clinical trials, improving the success rate, so you don't have to pay for so many candidates that fail in phase 2 or 3 because they just don't work in humans like they do in mice.
In principle, of course, the government can do both -- it's a question of what kind of resources they will want to put into future drug development.
You have a good point about Solyndra. The government is notoriously risk-averse, but that's where all its incentives point.
More money is always nice for an industry. The government already heavily invests in basic research though. Academia is essentially fully reliant on government grants. Govt funding probably wouldn’t move the needle on the problem IMO, and just change the nature of biotech into something closer to private academia.
I’d be curious to know how far from saturated basic research is with respect to medical advances. I’ve expressed before how many conditions are far from being well-managed or cured. Like why couldn’t or shouldn’t basic research get 10x the funding it currently does?
As someone making a career in basic research, I would love for it to get 10x the funding, and I think it would have enormous benefit for humankind and the US more specifically. Funding for cancer research, at least, seems to do slightly better than the average discretionary program, but it's still subject to all the same politics. Biden is a huge proponent of cancer research funding (for obvious reasons), and has championed multiple big bills increasing funding, during his presidency and Vice Presidency. However, the recent budget reconciliation that passed the House has funding flat or cut in certain places, because Republicans want to cut public funding! The exact same politics are at play in federal research funding as for any other public good. Elect more Democrats and the budgets will probably go up, etc.
Before the really unfortunate politicization of vaccines took hold I thought Operation Warp Speed might inspire a whole new appreciation for the power of medical research. It’s quite strange really how we don’t make the connection between all of the medical issues we and our loved ones could face and the possibility of addressing them at a faster pace. Some weird combination of missing political agency and medical denialism.
Not that I’m immune to this. It only occurred to me in the last few years. I never heard anyone make the case. Or it never stuck.
ARPA-H, a new agency spun up by the Biden administration, aims to do just that: https://arpa-h.gov/
> Funding for cancer research, at least, seems to do slightly better than the average discretionary program,
Sure, but the paylines for NCI are some of the worst.
That surprises me. I got a F31 funded in the 2020/21 cycle and I seem to recall it having one of the best pay lines of all the NIH agencies.
I believe it's held that a dollar invested in medical research yields tens of dollars of benefit, and that much more would be better. Surely 2x.
The government should just nationalize all pharma/biotech obviously play stupid games win stupid prizes that’s what they get for all their patent manipulation
The FDA's standard for equivalency really doesn't impact the cost of trials after you consider study design and market dynamics. Consider three situations:
1. My new drug is expected to be significantly better than standard of care. Becuase I expect a large difference in efficacy, my study doesn't have to be very large. And, because of the efficacy difference I can expect to get a price and/or share advantage in the market.
2. My new drug is expected to be roughly equivalent to standard of care. Because the efficacy is similar, I need a huge trial. Now, when I go to market, I've spent a ton of money on the trial and probably have to take a price discount. If I do that and start taking share, I have to consider that the incumbent will lower prices to block me. Project like this tend not to have a positive ROI and dont' get funded.
3. I think my drug will be equivalent or inferior to standard of care and I do a smaller study comparing it to placebo (kind of your scenario). Neither payors nor physicians will support something that doesn't meet the standard of care. Most importantly, physician malpractice liability is highest when physicians deviate from standard of care.
In #2 and #3, I also have to spend a ton more money on sales and marketing to convince people to use my drug, so the economics get even worse.
Great points! Do you think it's fair to summarize thusly: "while the FDA has high standards for approval, if a drug wouldn't meet them, it probably doesn't have enough market appeal to succeed anyway"?
At a high level yes. But it's more complex by the detail. The rest of a re-post of my response to Ben's question about reforms
I haven't seen this elsewhere, but I'd like to see a law passed requiring the FDA to use a consistent efficacy/risk calculation across therapeutic areas. I'm not saying that a new cancer drug should be held to the same risk level of a new rash treatment cream, but that the balancing process should be standardized.
The medical experts at FDA come from the therapeutic areas they are juding. Medical oncologists for oncology drugs, ophthamologists for eye drugs. The culture of cost/benefit is very different across these fields. Then add in the non-FDA advisory groups who exacerbate those biases.
The result is that oncology drugs get approved with risk/reward ratios that are much less favorable than drugs for untreated diseases that get denied.
Thank you for your input! I'm wondering what you think about the idea of separating the efficacy test from the safety test. What would you think of a separate third category of drugs that are basically approved for safety but not efficacy?
Many people think this is a good idea. Obviously it's more than a bit libertarian.
Well yeah, I lean that way.
I don't know enough about the situation, so this is speculation. I think it would be interesting to have conditional approval after phase I/II trials, but it's unclear if anyone would take them! Healthcare is a weird industry, because the customer is not actually the payer. In most cases, the patient and doctor are making purchasing decisions, but insurance/Medicare are deciding what things will be covered. In a standard market, you could imagine a case where a patient would rather pay 1k for an experimental drug vs 100k for a tried and true drug. However, the patient is choosing to pay a 1k co-pay for an approved drug (while insurance takes the 99k) vs 1k out of pocket for an uninsured experimental treatment. Typical market economics don't apply! I think that efficacy approval is going to be necessary for drugs to have any chance on the market. I don't see any future where the FDA gets out of phase III, efficacy testing, etc. It's interesting to speculate though, and relieving regulatory burden in drug approval is one area where I believe that society can make life cheaper without really impacting the quality of newly approved drugs.
"for reasons *too lengthy to get into here", ugh
The patent system wasn't bestowed on humanity by a deity. Governments around the world created the patent system in order to foster innovation. It strikes me as absurd that governments chose to create a system that increases costs for the governments, and then the governments try to find ways of reducing those costs as though it was nothing to do with their own system.
If the patent system currently strikes the perfect balance between fostering innovation and cost, then there is no advantage in attempting to subvert it. If it fosters rent seeking and bloat (and it does!), then modify the patent system directly.
This seems so incredibly obvious to me I find the whole discussion pretty baffling.
You are forgetting that the IRA price negotiation mechanism is structured as a tax, which can pass through reconciliation with a simple majority in the Senate. Revising patent laws would go through regular order and be subject to the filibuster.
I thought it qualified for reconciliation because the CBO estimated it to be a deficit reducing measure.
Ultimately what does and doesn’t qualify for reconciliation is up to the Senate parliamentarian—see $15 minimum wage.
Right, but the CBO score informs the parliamentarian's decision.
Has any party just tried to appoint a cronyist parliamentarian willing to rubber-stamp anything and everything as subject to reconciliation? This seems like an obvious tactic.
The current one is a holdover from when McConnell was majority leader. I wish they would’ve sacked her at the start of the session.
Didn't the Republicans fire their parliamentarian in the early Dubya era for ruling against them on something? I'm too lazy to look up the details now.
I read a lengthy article by an attorney one time about how random & capricious the rulings from the parliamentarian are, about what will increase the deficit (and so won't qualify for reconciliation) and what won't. Doesn't seem like a great system to invest that much power in 1 single person
I don't have any specific knowledge about the details of the legal mechanism being a tax, but my immediate presumption is that, if the reasoning is as you state, that's another good reason to oppose the whole thing. It's not only a stupid way to do it, it's also doing violence to the constitution in the pattern of "A mandate isn't a mandate if we call it a tax." If the only way to pass a law is to lie about what it is to hide it in the reconciliation process and then hope the courts fail to notice then it's not a real law.
The relevant roadblock here is not the courts or the constitution, it's because of archaic Senate rules that require 60 votes for most stuff.
Your suggestion of just not doing stuff if it requires massaging and torturing its meaning is a fine-sounding idea until you realize that following this logic would totally paralyze the American state and prevent the legislature from functioning.
The best solution would be to end the filibuster and then just pass the actual law, but unfortunately huge swathes of people have decided they are committed to this institution for no reason other than that it gives them good vibes and feels Moderate.
How much stuff does or doesn't get done because of the filibuster is a distinct question about calibration that people are welcome to argue about. Not liking the current equilibrium does not excuse assaulting the process instead.
Again please explain how Congress, which is given the power to tax, enacting a 95% tax if companies do not come to the negotiating table, is assuaging the process. Seems to me like a totally valid exercise of the taxation power.
That's not true. There were a number of large bi-partisan pieces of legislation that got through the filibuster. But it requires actual compromise
Big changes should require bi-partisan buy in and compromise. I fully support the filibuster
That’s a silly argument Dave. A 95% tax (see John from FL’s comment above) is still a tax, and Congress has the power to tax. If you don’t like how they exercised that power, vote for a different Congress.
Exhibit #2,893 on why the filibuster is bad.
And it's clear that since nothing has gotten past the filibuster in years, it's not worth trying to craft legislation that would get 60 votes.
See also the ubiquitous discussion in media outlets about how if a judge finds the FDA incorrectly approved an abortion or birth control-related drug that results in a "ban" of the drug, which completely ignores that (1) the only reason such a ruling would "ban" the drug is because of other already existing federal statutes and regulations that presumptively prohibit the sale of any pharmaceuticals in the absence of FDA approval, and (2) Congress has the power to amend those statutes and render the court decision moot.
But we could have a better discussion! One way to helpfully conceptualize for the Slow Boring crowd might be to imagine the government provided housing for all American citizens over 65. Should the government employ price controls in its housing vouchers to ensure greater housing is available?
Arguably we do *something like this* with Section 8, aiming at Fair Market Rents (FMRs) which HUD publishes regularly for a given area. Should HUD bargain those rates down further? If we assume we're expanding this to seniors (and we're assuming a world where seniors take up disproportionate housing use like they do with prescription drugs), would it be productive to set price controls this broadly across the industry? Would we get more housing over the long term?
I think this is more helpful of an analogy than bulk discounting for federal employee hotel use (which I assume is not the majority of hotel consumption.) In that spirit, here are two pieces by Manhattan Institute's senior fellow Chris Pope on the ideal drug R&D industry situation and the practical harms Congress could impose on it in the future.
1) https://www.city-journal.org/article/a-price-worth-paying
2) https://www.city-journal.org/article/congress-innovation-killing-drug-habit
Price controls make me happy
For a starter, granting monopolies is a completely archaic practice that belongs straight to the 17th century. The modern alternative (which exists under various forms since the 19th century!) is compulsory licensing, whereby the patent holder is guaranteed a ROI in the 10-20%pa range on their R&D expenditures, but production and sales still follow market competition.
But these archaic patent monopolies benefit from ironclad protection by the WTO. Not even Covid vaccines were made available for compulsory licensing! https://en.m.wikipedia.org/wiki/TRIPS_Agreement_waiver
Anything a single government (including the US) can do is try to *emulate the effect of compulsory licensing* by controlling prices based on production and R&D costs, and shorten patent length.
Are there other alternatives to monopoly besides compulsory licensing?
Another is granting prizes for inventions, which is roughly equivalent to mandatorily buying out patents at a pre-set price.
Large-buyer price negotiation is consistent with the right for a patent-holder to charge monopoly rents, though, as long as the patentholder has the right to take its ball and go home.
Except with the new Medicare negotiations the patentholder doesn't really have that right in the normal sense.
"If it fosters rent seeking and bloat (and it does!)," examples please, and proposed remedies
What do you think should be changed in the patent systems of various countries?
I don't know that the patent system alone is the problem, though there is certainly a case for better calibrating it. The problem in this particular case is more that it operates, at absolute best, in a strictly agnostic manner with all of the many other moving parts in the larger healthcare system, rather than towards a coherent set of goals and priorities. The entire US healthcare system is full of these kinds of issues.
Say more? What do you think are some of the important bad interactions?
I agree with Matt's analysis of how the FDA should improve its regulatory approach.
Matt is underselling the benefits of drug negotiation, though. The IRA has shown the way, and we should replicate the drug negotiation process across the rest of the federal budget. The relevant agency will determine what it is willing to pay for a product or service, and if Company A refuses to sell for that price, then any revenue from that product sold to others can be taxed up to 95%. Or Company A can refuse to sell anything to the relevant agency. Easy-peasy.
If Matt thinks the current negotiations over hotel rooms are beneficial, just wait until we apply this new process to hotels! The government can tell Marriott that federal employees can stay at a Courtyard hotel for $23 / night. I mean, the room is already there, the hotel is already built. It doesn't cost much for a little bit of water and to run a vacuum through it the next day. $23 seems fair enough, and that price isn't subject to the APA or judicial review anyway. If Marriott doesn't agree to it, we will just take 95% of their Courtyard revenue as a tax. Now Marriott can avoid this outcome if it agrees not to rent a room to any federal employee across any of its hotel brands. If Marriott does this, then we just have a similar negotiation with Hilton or Holiday Inn. We'll give them an offer they can't refuse.
For more on how the negotiation works: https://www.kff.org/medicare/issue-brief/faqs-about-the-inflation-reduction-acts-medicare-drug-price-negotiation-program/
Somehow, price negotiation has existed in most other OECD countries for decades, that produce 50%+ of major Pharma r&d, with also much much better health and pricing results for their citizens.
But I guess the fringe case 1984 dystopian scenario is a much more entertaining frame of reference for this convo!
For one thing, the US accounts for roughly 2/3 of R&D spending in the OECD countries (https://www.oecd-ilibrary.org/sites/6e38c622-en/index.html?itemId=/content/component/6e38c622-en) but profits are what funds that spending anyway- and the US accounts for at least 2/3, possibly as much as 4/5 of global pharma profits alone (https://healthpolicy.usc.edu/wp-content/uploads/2018/01/01.2018_Global20Burden20of20Medical20Innovation.pdf).
I don't like the way costs are distributed either, but it's an incredibly capital intensive industry, and currently the US is basically dragging the rest of the world with it in terms of development costs. From anecdotal experience in the field, we didn't even project European revenues when considering which candidates to bring to clinical because a drug that wouldn't be approved in the US could never break even its research cost, no matter how many countries outside the US approved it.
Understand and agree that’s all very relevant - my point was that there is a *huge* amount of wiggle space between
a) current broken system
b) gulag government destroys all profit making entities, as outlined above
Reactionary takes like this kinda ruin any chance at iterative adjustments to policy that would preserve the benefits of both elements.
We’ll never find out what works if anytime we try to implement new policy, people just start screaming ‘communists!’
It’s worked in other countries, therefore it’s worth testing and analyzing here.
One counterargument to "works in other countries" is if the U.S. is paying "full freight" and others are negotiating, then we are subsidizing their R&D.
Now, that's not exactly great either (when we're talking about subsidizing England, France, Germany) - I can certainly believe even if the doomsayers are correct that it would be better for those countries to pay _more_ than they currently do and for us to pay _less_ at the same time, but that's not the same statement as "we could just drop our prices down to Western Europe's prices" and that would be fine.
Yeah that’s definitely possible.
Or maybe the US drops prices and EU’s rises a bit, leveling the playing field a bit, which I think would be ideal. If US wasn’t such a cash cow, pharma companies would have to negotiate harder in the EU.
As it is they don’t have to negotiate in the US, so they don’t need to worry as much as EU
One possibility (I want to say something of the sort was debated in Congress in recent years, but I'm not sure) is for the US to institute price controls tying medication prices to other peer countries.
For instance, for all medications that have been approved in multiple OECD countries for a certain time, calculate an average price (maybe adjusted for purchasing power or GDP per capita or median personal income, and weighted by population). Then, by law, the US price for that medication can be no more than 120% of this average price, for "peer" countries excluding the US. US prices will still be on the high side, but not as big an outlier as they are now. By explicitly tying it to other countries' prices, pharma companies are also incentivized to drive harder bargains in other countries in order to increase the price they can get in the all-important US market. Thus, American patients will only provide a small subsidy to their foreign peers, if any, while the US market will still be very lucrative.
European pharma companies still make most of their profit from the US market, though -- which affects the benefit side of their cost-benefit analysis. If a drug is worth, say, $20 billion in US revenue, $5 billion in EU revenue, and $5 billion in "rest of world" revenue if it succeeds -- it matters little which of those is considered your "home" market, just that the total is $30 billion. E.g. Novo Nordisk is making a fortune from the US market despite being based in Denmark, and moving their HQ won't obviously change the picture.
I'm all for lower prices, but I'm not convinced that the success of European pharma companies proves that high prices in the USA are not driving their investment in new products.
Yes the US will always be a huge market for all pharma companies, we’re rich.
Again, it’s not about destroying pharma profits, it’s about bending the curve.
Currently they net about double your average large public business.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7054843/
That extra margin is specifically *not* going to R&D - it’s going to shareholders.
If the extra cost of pharma drugs in the US were going directly to R&D, it wouldn’t allow them to net profit at multiples of other S&P500s. This implies rent seeking, which IMO would be reduced if they had to negotiate with another large entity.
No need for doomsday scenarios, all these impacts can be measured
I'm all for a price negotiation. But what is being implemented is not a negotiation in any sense of that word.
"Agree to the government's price or don't sell any of your stuff to or via the government" actually seems like a credible form of negotiation. If anything the tax (rather than complete market denial) alternative is something of a sop to the producers here.
I'm not taking a position on the merits here, but I think this is credibly a negotiation.
That's not what the law says. Instead, it says "Agree to the government's price or don't sell any of your stuff to or via *anyone in the government's jurisdiction*". And then, yes, you get 5% back.
I am absolutely not an expert--nor even really an amateur--regarding the particulars of the IRA's price-negotiation, but I'm referring to the following from the link John posted:
"If an agreement on the maximum fair price is not reached by August 1, 2024, manufacturers may be subject to an excise tax, which will be administered by the IRS, as specified in the Inflation Reduction Act. In the revised guidance, CMS outlined an expedited process manufacturers can follow if they choose to not participate in the negotiation program, which would enable them to withdraw their drugs from coverage under Medicare and Medicaid to avoid paying the excise tax"
I find it difficult to read that in a way that doesn't suggest that manufacturers can still sell to or through private insurers or pharmacies outside of Medicare/Medicaid notwithstanding their being subject to the government's jurisdiction (as US corporations or entities).
I'm not an expert either, and you're probably right about interpreting that quote. If it's accurate, then that does change my opinion on the IRA quite substantially.
I made a comment in reference to yours and I now realize that you were criticizing this method of negotiation, not praising it. Ok, yes I agree with you. I'm 3 hours back.
I think this reasoning implies that the healthcare regimes of countries explains the variance in health outcomes. That doesn't make much sense to me. Health outcomes are likely determined by the variance in pre-healthcare factors given that those vary more than healthcare regimes.
I had an experience that shows how the government can do this much better and also how perceptions of this cannot be separated from political affialiation.
For patients who require continuous home oxygen, most of whom are on Medicare, the government historically set a reimbursement rate by fiat. In 2011 it was a bit over $200 per month per patient. Then they switched to an auction system. Seven years later it was ~$85 per month. It drove massive consolidation in the Durable Medical Equipment (DME) industry that served those patients, typically by undifferentiated mom and pop providers. The data on how it impacted care are murky. It probably diminished it slightly. But the savings were significant.
I'll never forget the (very representative) DME owner who spent 30 minutes ranting to me about how Obama was killing his business with his socialist policies, despite this being a shift from a top-down to a market-based pricing mechanism and the mechanism being created by George W Bush in the Part D legislation.
I agree that the IRA mechanism is terrible and would have preferred a market mechanism. it would have been much better to focus on areas where multiple drugs with therapeutically similar drugs existed and get manufacturers to competete for volume with price.
Thank you so much for your comments in this thread. I've learned a lot from them.
This is why SB has become the only comments thread I interact with. I learn things and people are great.
Your proposal is too modest.
It's a starting point. Once implemented, it can be applied in an increasingly expansive manner until we successfully stamp out any profit-making enterprise.
How swiftly?
i get it
It makes sense for the federal government as a very large consumer of say hotels, hire cares or flights, to use its size to negotiate for a better deal. In fact it would be negligent not to.
It makes sense for a government healthcare system as a very large consumer of drugs to use its size to negotiate a better deal. To point to the utility it believes the drug provides, to refuse to pay more than that, and to threaten not to buy the drug if a fair price can’t be reached. This is what many governments already do (say Australia).
The historic ban on Medicare negotiating prices with drug companies is almost a criminal misuse of taxpayer money.
I agree pharmaceutical development is important - that doesn’t mean we should just throw money at incumbent pharmaceutical companies. Lots of things are important, no one suggests overpaying for electronics to spur more innovation in Silicon Valley.
The whole no-negotiating drug prices prohibition was inserted to circumvent litigation and a hold up problem. Like many institutional structures it’s purpose has disappeared yet it persists as a drag on the system.
Drug spending is 9% of heathcare expenditures yet it occupies the majority of attention ( https://www.americanprogress.org/article/following-the-money-untangling-u-s-prescription-drug-financing/#:~:text=American%20prescription%20drug%20spending%20totaled,United%20States'%20gross%20domestic%20product ). What about the other 91%? Any ideas about that?
I think it's because drugs are overwhelmingly provided by for-profit corporations whereas most other healthcare services are provided by non-profits. The public discourse is much less sympathetic to for profit providers. This is probably a mistake though as the non-profits are just as culpable for the overpricing they do and as you point out, are responsible for a much bigger share of the overall problem.
Agree. Taken as a whole, this could be one of the more cost-efficient dimensions of the American healthcare system. We spend so much more on things that provide no net benefit to health outcomes. I confess I was very disappointed with this post given the exciting title. Boosting innovation in the creation of highly effective drug treatments for deadly and debilitating diseases seems like one of the most important things humanity can pursue. It seems unlikely to me that fiddling the requirements for demonstrating effectiveness, thereby improving the risk / reward profile for a handful of drug manufacturing and marketing companies, is going to do much to "boost innovation".
This actually is where much of the leverage is. Drug developers spend an average of a billion dollars for each new approved drug. If that could be reduced by 10x (and it was 10x smaller just a few decades ago) we'd have much more drug innovation.
One important note: this doesn't just effect a handful of drug developers, it affects every single person who is developing a drug.
Indeed, the only other leverage points I know of are (1) improve academic research funding. That's also a great value, but all the academic research is stuck behind the bottleneck in clinical trials and approvals. (2) focus specifically on reducing the cost of a clinical trial.
That's fair, but isn't it more about the “hit rate” for truly effective drugs than the cost of trials per se? It seems far more likely to me that the “lower the gate” approach results in more ineffective but highly marketable drugs going into the stream of commerce, so the prescription drugs industry becomes like the supplements industry, but with much bigger advertising budgets. Opening a pathway for a large number of attractive-sounding but ineffective drugs that can make back their trial costs more easily does not seem like the best way to improve health outcomes. It should be much cheaper for Pharma to weed out the losers before they have to spend hundreds of millions on expansive trials, but that seems more like a technical problem that will be solved with better biochemistry and maybe better AI, which is where I thought the post was going…
Yes, I think those are both good points. The first is what I meant by calling it a libertarian solution. Mind you, the customer in this case is partly the doctor, who is very sophisticated, so this might work well, but there are many details and questions. Companies using AI to make better decisions about what to pursue is definitely a big angle within the industry now. I don't know how successful that will be, but the good news there is that those companies should be highly motivated to do that weeding out, since it's their money on the line. We'll see how it goes.
End heroic end of life care.
Give huge discounts to people that stay fit and are not overweight
Both absolutely huge ideas. And anything we can do to encourage healthy eating, exercise, and less overly heroic end of life care
Cue cries of eugenics and fatphobia!
(I think you're mostly right on the merits, but good luck convincing enough people. Also, the devil is in the details: how "heroic" is "heroic," and who decides? We use some kind of "$$ per QALY" metric like the British NHS, presumably?)
Yes, Congress has passed multiple laws that has lead to programs seeking ways to reduce healthcare costs, such as value based care initiatives from CMS, Medicaid Health Homes, ACO's, etc., creating large datasets for research like NIH's All of Us program, the VA/DOD are looking at ways of increasing physician efficiency to retain good doctors for less money, and on and on.
With drugs, it's a combination of utilization and cost. The cost side is much easier to attack by just not spending a ridiculous amount, it is the low hanging fruit. Just reducing that from 9% to like 7.8% (all else staying the same) would be huge.
Yep but basically nothing that would actually drive human behavior changes at the population.
For example, huge discounts for people being fit
That'd be an interesting plan, but ask your private insurer why they aren't doing that.
At my previous job, my insurer did a thing to encourage regular annual checkups. If you didn’t have a “wellness visit” before June 30, then your premium went up by $30 a month for the following year.
Unfortunately, if I came to the doctor for a routine checkup to re-approve a prescription, the doctor wasn’t able to write that up as a “wellness visit”, and a separate visit needed to be arranged.
In implementation, a lot of these good ideas often turn stupid, and I worry that “discounts for being fit” would do that. (I’ve often seen questionnaires about fitness activity, and they specifically usually ask how many minutes of moderate heart rate exercise you get that isn’t related to your job. I’m sure there’s some research reason that it makes sense to separate out people who get exertion through work and people who do exercise recreationally, but it could easily turn perverse if it became an incentive. I think my bike commutes often didn’t get counted by doctors.)
Use something lije military test for recruits to test fitness.
Number of push-ups, sit ups and 2 mile run (or better yet VO2 max)
Plus of course a dex scan for body mass
You can speak to your employer to see if they could get a program like that going, but that's likely beyond the levers available to the government to make that happen via provider/payer incentives.
I suppose Congress could enact something like the Presidential Fitness Test for adults, and the government could cut checks to health people or something, but, as Kenny Easwaran suggests above, that would absolutely become silly for a million reasons.
> "If you’re at very high risk of dying in the very near future from a fatal disease, then you probably aren’t as worried about the downsides of experimental treatments."
100%. My paper 'Pandemic Ethics and Status Quo Risk' - https://academic.oup.com/phe/article/15/1/64/6499206 - offers a sustained exploration of this kind of argument:
"A pandemic reverses the usual asymmetry of risk. Now it is the status quo that is immensely dangerous, and a typical sort of medical intervention (such as an experimental drug or vaccine) is comparatively less so. As a result, we should expect to find many cases in which the potential benefits of innovation outweigh the potential risks. Doing nothing new, and allowing the pandemic to continue unabated, should be recognized as a far riskier prospect—for many individuals, and especially for society at large—than trying experimental or otherwise uncertain solutions for which the risks are orders of magnitude lower than the risk otherwise posed by the virus."
(Of course, it's also importantly true of many individuals in non-pandemic contexts too!)
The FDA process is pretty damn good. One of the issues they have to suss out and why it takes so long is the FDA doesn't do the testing. The companies producing the drug do. There are MANY statistical manipulations the companies are allowed to use in regards to adverse events(aka side effects). One of the primary ones is if a participant is using another medication, has an unrelated health issue, contracts an unrelated disease, the company can remove this adverse event data from the test population. This makes sense on the first read as all these things are not necessarily under the drug companies control when they are running the trial. However, the company doesn't really have to justify it either, meaning they are under no real obligation to prove it wasn't their new treatment target that caused the adverse event. This is why the FDA is so cautious. Basically a drug/treatment usually has to show efficacy above or equal to current treatment technology to be market viable, side effects are huge part of that efficacy math. So companies are highly incentivized to dump adverse effect data under any pre text. As Matt already stated even with 10000 participants you won't catch all the side effects(adverse events), especially longitudinal ones. The FDA process as restrictive as it seems isn't flawless either. Look no further then Fen-Phen, the Opioid Crisis, or problems with host of other drugs that end up making a media splash.
I do think negotiating prices is the right call. Medicare/Medicaid are huge buyer. They should be able to work with pharmaceutical companies for better pricing, especially for life saving small market drugs that often are the target for crazy price increases.
>FDA doesn't do the testing
Would it make sense to change that? E.g. for phase 3 trials, you have a government agency coordinating the trials and healthcare providers carrying it out; the manufacturer of the drug candidate is not directly involved, and the FDA will not have to worry that a business with billions of dollars at stake is trying to fudge the results. The studies can be paid for by application fees from the manufacturer, enough for a reasonable Phase 3 trial (e.g. 1000 patients or enough power that the manufacturer confidently expects p<0.01; several visits during the trial, in which you measure outcomes, ask about various side effects, ask if there were any unexpected effects, etc.) The agency can get additional public funding for further studies if it finds some sufficiently interesting unexpected results.
The thought is that this way, the efficacy study will not be designed and managed by people strongly motivated to want a specific result. It might save time and effort that would otherwise go into trying to get away with manipulating the results, and trying to catch any manipulations. And financing it by application fees would mean that the cost of bringing a drug to market, from the company's perspective, should not change dramatically.
There might be additional benefits to theoretical efficiency (and logistical difficulties) if the drug assessment agency were an international consortium. It could test once for the world; recruit 1000+ patient samples each to be representative of the US, UK, EU, China, and the world, and do subgroup and pooled analysis, all to be included in the final report; and the patients can be recruited preferentially in countries where good medical research can be done at a relatively modest cost. The FDA and EMA would then be able to request (or perhaps conduct) follow-up studies if they think more information is necessary before approval.
The national or international medication testing agency could, in principle, also help to coordinate and fund post-marketing, Phase 4 trials to assess things like rare or long-term adverse effects, head-to-head comparisons of different drugs approved for the same indication, maybe studies to look at correlates of why some patients respond better to one medication over another (genetic, metabolic, etc).
I'm not an expert in medical research or government agencies, though I know a little about both, but separating the testing from the interested parties sounds like a good idea to me.
I'm just not sure we want to pay for it at that level but it might be a good idea. Right now the system actually functions pretty well was my point. It's not nearly as restrictive as it's portrayed. It's highly regulated sure, just like financial markets but that doesn't mean it's stifling innovation or costing us more per drug target than the risk mitigation is worth.
It would be interesting to what the cost of your suggestion would be though just gross dollar wise. I suspect that would well I'm excess of any political will to establish it.
Much of the cost would theoretically be borne by pharma companies -- they would pay a huge fee, close to the cost of the clinical trial, for the agency to conduct it, instead of directly working with healthcare providers to sponsor the study.
Yeah similar to the way User Fees are already treated. The problem is that would likely raise costs(in the short term) and put a number of current testing groups out of business or in tighter margins. It's why I said I don't think the political will exists. It's not just about finding the money, it's also about the political will to piss off the current structure of the market.
If you’re saying that a drug company doesn’t need to report an adverse reaction in a study subject because it believes that it was due to another drug or some other cause, that’s false. They can make that argument to FDA but they still have to report it. FDA is usually agnostic about causation in controlled studies because such unrelated events are equally likely to occur in the control group so they cancel out in comparison.
It's not false. You nicely point it out right the end of your comment. The FDA is agnostic to causation in controlled studies because unrelated events are equally likely to occur in the control group.
A participant reports an adverse event. The group conducting the study then determines if that event is cause for removal of the subject from the study. Then it is determined if the data can be retained and used or not. It's very possible that the data just gets dropped from the final submission. Are they supposed to do this? No. Do they do it? Yes. Scientific American had an article about it a few years back. In some cases it just reported as drop, in some cases it's not reported at all.
This may have been the case 20 or 30 years ago, but procedures have tightened considerably. Investigators are required to keep enrollment records. It's illegal to remove or drop records. And FDA routinely conducts investigations.
It was still true as of 2017. Especially as more trials we're done outside of the US. Again, I'm not really faulting the FDA. I think they do a great job. I'm just saying the economic incentives highly inform the process and can lead to massaged, manipulated or outright manufactured data. Especially for a drugs that are in tight market or late stage trials. They represent a huge investment and cost especially for the smaller companies.
I don’t know where you’re getting your information; maybe the Scientific American article was from 2017 but references incidences from years earlier. I have been working with this data at FDA for 20 years and never seen anything like this or even how it would be possible.
This is the chicken and egg of trial cost verus trial quality.
Given pharma's incentive to frame data in a favorable way, the more transpareny the better. The more endpoints the better. But all of this drives study costs up.
There’s no additional cost to reporting everything. In fact, spending time deciding and explaining whether or not an adverse event is due to the study drug is actually cost and generally a waste of time and money for the reason that I explained.
As far as endpoints go, FDA requires endpoints to be specified in advance to prevent biased framing.
The cost is in the data collection, especially when collecting the data requires more patient encounters, additional tests, and makes recruitment and study completion harder. Agree that the data analysis is trivial.
As far as consistency goes anyone who argues that women have a right to bodily autonomy and hence that the government can't prevent doctors from giving them drugs/treatments to enable abortion should also believe that same justification protects the right of patients to try experimental or untested medications.
Sure, maybe you can justify restricting genuine snake oil consistent with that moral position but the underlying principle is that when deciding what kind of medical treatments you are allowed to get the government doesn't get to tell you how you should weigh values/risks be they the moral status of the fetus or the chance you paid for a medicine that just harms you.
I mean if your bodily autonomy doesn't let the government stop you from legally accessing drugs to remove a fetus why does it let them stop you from drugs designed to remove a tumor? What if it's an effective drug but just hasn't passed the regulatory hurdles here? What if abortion drugs only worked 50% of the time, 85%, is there some threshold of probability of success at which it becomes moral for the government to substitute their judgement re: risks/benefits to your body over yours?
--
To be clear one can, as I do, support abortion rights for good old fashioned consequentialist grounds but it's the attempt to claim a special moral status that's dubious.
I 100% agree with your first paragraph, which is why I also roll my eyes at all the lefties who claim that the U.S. Constitution mandates anarcho-capitalism for a handful of specific medical procedures/products.
(Seriously, I would love if the Constitution contained an express "right to privacy," but it doesn't -- I say that as someone who has co-authored a book on privacy law -- and if you're going to interpret the "penumbras and emanations" of the Constitution to create such a right, then it should at least be applied in a logically consistent manner. If there's bodily autonomy for abortions, birth control, and puberty blockers, then there should certainly be bodily autonomy for taking any other sort of medication you want to take, at least so long as there's full disclosure about the experimental/untested status of it.)
I don't per se disagree but I'd argue, a la Balkin's originalism, that the OPM of the constitution included an understanding of the courts as engaged in a common law practice of substantively explicating what the law means -- even to the extent of finding new rights (I mean the rule of lenity, suppression, and virtually all 1a law aren't written in the constitution and the later was clearly not OPM either).
But what matters is coherence. What makes the court not a legislature is the need to follow or overturn precedent. So I'm fine with the court extending Griswald to Roe but it's got to take the principal seriously and apply it to drugs generally etc.
I think it's wrong for the government to forcibly extract someone's kidney and give it to someone else to save their life. But I still think the government can ban medical treatments that don't work. (It's easy to say "patients should be allowed to make their own decisions" but that doesn't really address the institutional problems you create when companies can make profits by pushing treatments with weak or no evidence.)
Exactly the same thing goes for abortion. A person's body is not a public resource that can be forcibly exploited for the benefit of others. But I don't have a problem with regulating abortion methods that don't work or are particularly unsafe, or imposing malpractice liability on doctors who botch abortions, etc. The trouble in that sphere is that most of the safety concerns turn out to be pretextual trumped-up stuff driven by motivated reasoning.
But that's not what the government is literally doing when they ban abortion. If you break it down they are banning the use of certain medications and surgical treatments because they judge the harms exceed the benefits. Sure, making it literally murder is a different level but just banning abortions requires no more than saying those drugs/surgeries are not approved for that purpose and no off label.
Sure, when they judge those harms they are substituting their values about the fetus's life for your values. But how is that different than when they come in and say: nah if we think the evidence says it's only a 5% chance of a cure but a 20% chance of a nasty side effect and a 95% chance of losing your money you can't legally take the med any different?
You're looking at it at too high a level of abstraction. The intent of an abortion ban is to prevent someone from terminating a pregnancy. The intent of medical regulation is not to prevent someone from extending their life. It's literally true of course that banning a medical treatment infringes on "autonomy" in that you can't choose to get the treatment, but the substantial autonomy interest at stake in abortion is not the specific procedure but the outcome. The same thing is true of life-extending drugs (the autonomy interest is in being able to choose to do treatment to live longer) and precisely for that reason regulation aimed at effectiveness is different in kind from regulation that rejects the goal as such.
Ok, suppose I want to be in the state of being really fucking high -- not just drunk but feeling the intense state of reward provided by many hard drugs. Now you can't argue that the government isn't systematically trying to block all the treatments that put my brain into that state. Fuck, it's specifically legal to use many of the drugs that produce this effect as long as that isn't the purpose of the use.Sure, maybe it's legal to be happy but that's analagous to it being legal to have a miscarriage.
The whole idea of the abortion autonomy debate is that the individual gets to say: I want to change my body in a way that wouldn't happen without intervention and you don't get to tell me I can't.
--
But even if we leave that issue aside I think you have another problem. Your changing the argument from a right to bodily autonomy to a right to be free from interference with a certain kind of intent.
I think it's going to be very hard to formulate any such principle in an attractive way. Exactly what is the kind of intent that triggers this right and why is that an attractive principle.
The appeal of the autonomy argument was always the simple: but I get to decide what happens with my body. Once you start stacking on extra qualifications it's no more persuasive than just saying I think this particular thing should be allowed.
--
However, if you can formulate a compelling principle that rules out the examples I raise that would be the kind of thing that could convince me. Like tell me what exactly it is that one has a right to.
I don't think it's true in reality that bans on hard drugs are about preventing intense highs--they're primarily about the other effects of those drugs. You can take some of them for medical reasons but that's because we think the dangers are worth it in those cases. But you're getting the point--a ban on drugs because they produce intense highs directly implicates an autonomy right to have an intense high. I'm not sure such a right exists though.
Intent does matter a lot to rights. An incidental burden is different from a direct burden. It's one thing for the government to say "you must pay $20 to go to the Legalize All Drugs rally" and another thing for the government to charge you the normal highway toll to get there.
It sounds to me that you are getting stuck on the difference between a slogan that you want to read to its furthest semantic potential ("my body my choice") and the actual views non-libertarian pro-choicers hold. But that's just to say that reality is always more complicated than slogans. The case for abortion rights has a lot to do with autonomy involving the body but it doesn't hinge on or require thinking that every choice involving the body in some way is equally protected. I'm sorry you don't find my argument appealing but the idea that there's a difference between safety regulations on medicine (or bans on hard drugs), on the one hand, and bans on terminating a pregnancy, on the other, is obviously very widely accepted.
But really its just inconsistency that bugs me. If people want to respond -- ohh totally it's just a slogan that's what I want.
Sure, they don't intend to stop you from getting high just like they don't intend to stop you from becoming non-pregnant -- just block the associated consequences (addiction/intentional death of a fetus).
And look I'm a huge fan of abortion rights on consequentialist grounds and "my body my choice" is a fine slogan but people frequently advance the bodily autonomy consideration as both a serious moral and legal argument (though my point would be valid regardless -- but only relevant to those who do accept this argument). For instance.
1) See Thompson's violin argument. A famous philosophical argument claiming to argue for abortion rights on something more than consequentialist grounds. This is covered frequently in college philosophy classes.
2) The legal argument for Roe has always turned on the notion of 'privacy' as articulated in Griswald which really means something very close to personal autonomy.
---
And sure, I'm happy to admit that you can understand the notion of autonomy in a more consequentialist fashion -- reproduction is especially salient and important to women's self-identity and ability to fully participate in careers etc like men do. And I believe all that.
The problem is that once you admit it's this muddy notion that turns on disputable value judgements (it's important that women can compete for jobs with men, it's really important that people get to control their reproduction etc) then you can no longer blast supreme court decisions or people who go the other way as acting illegitimately or similar invectives. You can strongly disagree and wish they brought your values to the table but it's no longer an illegitimate deciscion.
And while it's a bit off topic I think that distinction is hugely important. Trump trying to steal an election was illegitimate. The supreme court going the other way on Roe is a legitimate result of our country's system of government that is dealt with by winning enough elections for long enough to change the court (or just passing a federal law protecting abortion).
“Euphoria” is considered a side effect of some drugs that counts negatively, because it leads to abuse potential.
You’re forgetting that people are idiots. They will be exploited by fraudsters selling snake oil to a degree the public won’t find acceptable.
No, that's a perfectly cogent consequentialist argument for treating the positions differently. As I said, I accept such arguments.
My point is that you can't short circuit the whole abortion debate and push the anti-abortion position from the board without getting into the messy consequentialist details unless you are willing to do the same thing with drugs generally.
The existence of some other reasonable argument isn't at all helpful for the whole right to bodily autonomy position.
Should the same agency regulate safety and efficacy? I could imagine FDA regulating only safety and what efficacy claims could be made and Medicare using CBA to decide whether to pay for and how much to pay for new drugs.
I think that there's something to that. I don't think that you can totally get the FDA out of the efficacy business. You don't want FDA approved child leukemia drugs that do literally nothing. But the FDA approval doesn't do enough on relative efficacy of different treatments or sub-populations differing efficacy. And drug approval isn't the place to do that, but right now there doesn't seem to be enough incentive to do those kinds of efficacy tests post-approval, so a price negotiation CBA is an interesting approach.
But doesn’t the fda allow the off-label prescription of any medication that is approved for any use?
Yes. I've hear that argument since at least 1972 and it is so Econ 101 that I doubt it was new then. And yes, there is the bureaucratic "embarrassment" effect (the same reason the Fed won't reduce the EFFR until they are SURE they will not have to raise it again in a month or two), but is that really the reason not to use cost benefit analysis, or rather to use cost benefit analysis in which bureaucratic embarrassment figures among the costs?
After all, FDA is not the only regulatory agency that does not use cost benefit analysis or we would have an XL pipeline more nuclear power plants and infill commercial and residential development. Did CDC use cost benefit analysis or provide information so that public decision makers could use CBA in deciding when and for how long and in what conditions to close schools, and restaurants and churches?
Now I'm not opposed to looking for agency specific solutions an OMB memo to all agencies saying "use CBA in making decisions" would not work, but at the same time we ought to think about systemic reasons for regulatory "misbehavior."
FDA is prohibited by law from using cost-benefit in its decision making
Oh that is good news. I assumed it was something much harder like bureaucratic culture, outdated attitudes or something like that. CBA is something that we ought to be able to get bipartisan agreement on.
But that’ll kill grandma!!!!1!
I'm curious since you post a lot about these things. In what sense are you aware that cost-benefit analysis is used to inform regulatory policy at all? I see things like CBO estimates and that sounds like CB analysis I guess, but I'm not sure.
AS far as I can tell, it is NEVER used. I just think it should be.
Ehhhhh... there are a lot of clinical trials run at least partially at foreign sites already. Any gains in price reduction there would be marginal.
And I'm not sure your hand waving of the innovation argument holds water. Drug discovery has a lot more upfront costs and failures that hotel chains do! Companies will not engage in a research project if the market isn't there (in fact, I've been on projects where the research has gotten pretty advanced and then the company has pulled the plug due to changes in the market).
On a related note... if you haven't been following the Relyvrio mess (see https://www.newyorker.com/magazine/2023/06/26/relyvrio-als-fda-approval and
https://www.science.org/content/blog-post/als-drug-fails-again), it's a strong argument for the FDA to NOT approve drugs without strong data since payers will be encouraged to take on exorbitant costs for no benefit.
"it's a strong argument for the FDA to NOT approve drugs without strong data since payers will be encouraged to take on exorbitant costs for no benefit."
No, it's not. Let people have the right to try. If necessary, one could devise a system in which cash payments for experimental drugs are fine, but insurance won't cover them.
One example of a drug that doesn't work, doesn't obviate the argument that those of us, like myself, with fatal diseases shouldn't be able to try.
If you want a real mess, look at petosemtamab, which I'm on: https://www.onclive.com/view/petosemtamab-generates-early-clinical-activity-and-tolerability-in-hnscc and which should be available to patients immediately after our first tumor-removal surgeries. Yet instead of being approved with incredible phase 1b and phase 2 trial data, it's heading to phase 3. Madness.
I'll say up front that I am extremely sympathetic to you and anyone who suffers from a disease with no available treatment, or in the case of many cancers, inadequate treatment. It's literally the reason I do the job that I do (drug discovery research).
Having said that... the industry is littered with drug candidates that looked amazing in Phase II only to fall apart in Phase III. You need to run the trial to get the statistical power to determine both efficacy and safety. For cases of extreme unmet medical need, the FDA can approve on a conditional basis before Phase III is complete, and they've been doing so more regularly in response to patient advocacy groups.
Relyvrio is instructive because it shows the downside of that approach when the underlying data is questionable (and make no mistake, it was). This wasn't even a case of a company with years of expensive research seeing a payoff... it was literally taking two generics together and seeing an effect. Which would have been incredible if it had worked! But it didn't, and payers had to cough up a lot of money for the drug in the meantime, and those costs get spread to everyone in terms of higher premiums. You can argue that insurance should be decoupled from FDA approval, but then they'd need to have their own scientific review boards and that just adds time and expense to the whole process.
Why was the AstraZeneca covid vaccine never approved in the United States?
They didn't get the emergency use authorization because they screwed up the statistics on their original trial. I don't remember the exact timing but at some point the Pfizer and Moderna vaccines got full approval, meaning the EUA pathway was no longer available. And then they eventually withdrew their application.
I remember that there were significant issues with their original application. But I never understood why they never got full approval, the way that Novavax and perhaps some other latecomers eventually did.
"So basically whichever country in the world gets the best deal on the price of a drug, the United States has to also get." Your HIV drugs in Africa example is right as an example of why this is dumb. But would it be less dumb if a purchasing power parity formula was used to set a global price?
This is called price-refernencing and it aready exists, just not in the US. The result is that nearly all drugs are launched outside the US in the same sequence of countries. Every drug company has spreadsheets that solve this pricing problem. If the US did this, expect the sequence to change and US launches to get pushed back. Since the US is still the biggest market (higher usage, even if the EU, India, and China are bigger by population), the strategy would be to set the highest price in countries that the US references to maximize the US price.
Because the US contributes the greatest share of drug profits today, many development proejcts would be cancelled and innovation would suffer.
I generally think that Econ 101 arguments don't translate well to healthcare, but this is a case where they would.
I think global prices for drugs would be very good, but a global price would have to be higher than what Europe, Canada, Japan, etc. pay for drugs currently, so they are not going agree to such a global price regime without a lot of kicking and screaming.
This article is missing any discussion of whether the pharmaceutical industry is making outsize profits or margins. Certainly seems that way when Semaglutide blows up for Novo Nordisk, but what about all the failures. I’ve heard that pharmaceutical profits aren’t all that impressive compared to other industries, and that needs to be discussed before we engage in aggressive pricing controls.
I note that the average R&D spending by industry is less than 4% and for the pharmaceutical industry it's 22% (and 44% for biotech). (https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/R&D.html)
In the immortal words of Walter White, I would advise we tread lightly.
David Abbot linked this in another comment:
Windfall profits will exist as long as pharmaceutical patents grant an unregulated monopoly. This JAMA article says pharmaceutical companies have average profit margins of 13.8%, which is 79% higher than the 7.7% average for other large corporations. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7054843.
I haven't read the article but there's the link.
The "limitations" section as noted in that analysis are significant enough to make any conclusions very, very suspect.
Also ... why do I care - at all - about 13.8% profit margins? Seems shockingly cheap for the risks inherent in their business model. Facebook's net profit margin is up to 35%. I can find companies > 30% all day just due M&A consolidation and massive barriers to entry.
Other commenters said most of the cost seems to come from Phase 3 trials.
At least this limitation:
===============
First, this analysis focused on large, fully integrated pharmaceutical companies that generate revenue and profit primarily from drug sales. It did not consider small or midsized biopharmaceutical companies or biotechnology companies engaged in discovery research or early-stage development, which typically have little revenue and negative profits (losses).18 As such, the pharmaceutical data set was not representative of the broad biopharmaceutical industry and the results cannot be extrapolated to the industry as a whole.
=================
would be less relevant in that case, as I _assume_ the smaller companies in early-stage development are mostly doing Phase 1 & Phase 2 trials.
What’s the risk profile of pharmaceutical companies compared to other industries?
That's a fair question - the limitations section (the only section I looked at, after John from FL's comment) says:
==================
Third, this analysis did not consider whether companies in the pharmaceutical data set had excess profit. In accounting and finance, excess profit is defined as profit over and above a “normal” return on capital invested in the company—a return that is commonly associated with the risk of the investment. Future research can be directed at examining the relationship between investment risk, returns on capital investments, and reported profits.
===================
So I don't know the answer, and they didn't consider it.
An interesting political problem here is that a lot of voters/activists see pharma companies as profoundly unsympathetic - in many ways analogous to the “big developers” you see as housing boogeymen opposed by activists who ostensibly want a non-market solution. In practice, this leads to policies that make the cost/benefit for drug development worse without actually addressing any of the approval-related costs you mention. For example, the IRA recently reduced patent timelines for small molecule drugs, which, without any other reforms, just makes it less economical to develop them.
When you’re developing a drug, you’re sort of making a leveraged bet with a really long horizon whose success or failure likelihood can’t really be quantified - the signals appear way too slowly. A few years back, there was some interest in an X-Prize type system where you’d get a large up front payment for a successful approval (or reaching some milestones), then forfeit some of the longer-term monopoly rents. I think this has fallen out of favor, but I’m not really sure why.
For small biotechs, this is how a lot of the financing actually already works - you typically do deals with big pharma partners where you’d get payouts depending on whether or not predefined milestones are met, both pre-clinically and clinically. So I think a lot of the biotech ecosystem would be compatible with such a system, assuming it was sufficiently transparent.
First, the US has done a much better job over the past decade at supporting smaller trials for regulatory submission when there is unmet medical need and/or high burden with or without terminal risk. In oncology and rare disease, Phase 3s are often skipped altogether. The offset, however, is that it's harder to find patients and run the trials on a per-patient basis, and it can take multiple Phase 2s to get the trial design right in a disease that is not well understood (rare) or where the basic science is in rapid flux (oncology).
Regarding international trials: it's not just the patients and their genetics. It's the diseases they have (or don't), which are VERY different. It's the practice of medicine and the ability to actually execute a modern protocol with modern diagnostic practices, which sometimes involves specific medical training and sometimes also involves a great deal of technology and infrastructure, which can be vary greatly. There are many variables. The US and Europe are like siblings. Things are close enough. But the rest of the world is more like cousins (in the flexible sense of the term.. might be first cousins, or second cousins once removed, third cousins...).
It's a time value of money problem and a patent enforcement problem. Venture capital wants to make big investments in biotech today (yay!) in exchange for a big possible payout upon success (boo!) and unfortunately, that currently often relies on pricing in the US. The best way around this? Wait a few years until the patent exclusivity expires, and actually enforce it without the IP shenanigans that we so often see from big pharma who eventually acquires the successful IP.