A summary of the USTP FDA Reform Panel I organized
The US Transhumanist Party livestreams special “Enlightenment Salon” events every Sunday at 4pm on YouTube. Three weeks ago I organized a special “Enlightenment Salon” panel event on FDA reform. In addition to myself, the following people ended up being on the panel:
Prof. Alex Tabarrok, Bartley J. Madden Chair in Economics, George Mason University.
Dr. Max More, President Emeritus, Alcor Life Extension Foundation.
Jim O’Neill, CEO, SENS Research Foundation.
Dr. Edward Hudgins, Founder, Human Achievement Alliance.
Prof. Garrett Jones, Mercatus Center, George Mason University .
Willy C, Medical Student and Blogger.
The entire event was probably the most information dense event the USTP has done. I kicked things off by giving a short presentation which I managed to blaze through in 15 minutes (the slides can be viewed on SlideShare here). The presentation set the stage for a very informative and productive discussion.
My only regret was that we didn’t have any women on the panel. During the course of researching my presentation I found out about the work of Dr. Mary J. Ruwart however. Dr. Ruwart estimated the number of people who die every year from FDA delays to be around 150,000 per year in her book Death by Regulation. So, I am happy to announce that tomorrow (Sunday April 24th from 4 - 6pm) the USTP will be doing a special event with her.
Here is the recording of the FDA reform event, starting when at 00:05:00 when my presentation kicks off:
I’ve written a summary of the major points each of the panelists (and two others) made during the first hour or so of the session. These are heavily paraphrased (rather than direct quotes I shortened what was said in most cases while maintaining the core meaning of what was communicated). I have put my own comments in italics.
Prof. Alex Tabarrok:
The FDA can approve a bad drug (Type I) or fail to approve a good drug (Type II). If they approve a bad drug, people who were affected will go on Oprah and there will be huge backlash. If they don’t approve a good drug, there is no backlash. The invisible graveyard is a statistical reality, but it’s hard to see. This can be seen easily by asking people to name a time when the FDA approved a bad drug (or a drug with unexpected side effects). Lots of people can think of something (many point to Thalidomide, which is actually a drug that was approved in Europe and caused birth defects. Thalidomide, incidentally, as many important applications but was not approved by the FDA until 1998).
When you have a Type I error, you learn something — we learn about the harms of a drug and we change our behavior. With a Type II error we never learn anything. We can’t see the consequences of a failure to approve, and even worse we can’t see the many drugs that never even made it to FDA mandated trials in the first place because they were deemed to risky to justify the cost.
Reciprocity is a sensible reform that is one of the most feasible.
The FDA likes to think they are the “gold standard” for drug approval. Yet, people in other countries don’t worry about whether drugs are FDA approved. For food safety we already have reciprocity with Canada.
The FDA has been working for 40 years on new standards for approving sunscreens. So Europeans have much more advanced sunscreen than the US.
If aspirin were invented today it probably wouldn’t be approved.
One thing we have done already which is probably the smartest thing we’ve done in a while is PEDUFA (Prescription Drug User Fee Act). The drug developers pay an extra tax as long as approvals are sped up. The FDA was happy because they got top expand their bureaucracy and drug companies were happy because they could get to market faster.
In the EU the EMA “farms out” reviews to private companies. (So the EMA is more like a “an approver of approvers”). Private companies can do a good job -for instance look at Underwriters Laboratory in the realm of electrical devices (if you look at many electrical devices you may see a “UL” seal). Many major companies like Amazon won’t carry devices unless they are UL approved.
There is no formal process whereby if a disease is more deadly then the standards are lowered to speed approvals. For instance, for pancreatic cancer, which often kills within 6 months, the standards should be lower (and more risk should be tolerated) since patients have less to loose. For something like acne treatment the standards can be much higher. The FDA recognizes this to some extent in practice but it’s totally informal - technically it’s not supposed to happen. However they could do this formally and adjust the required statistical significance levels required. They could use Bayesian statistical techniques as well.
There’s no route to approve a drug for anti-aging. If a company wants to do R&D on anti-aging therapeutics, there’s not a clear route for approval.
Dr. Max More:
We should keep in mind full abolishment of the agency as a long term goal. [My response: I am against full abolishment but I agree with the general sentiment Max expressed. Everyone should be forced to at least consider abolishiment and if they are against it explain in some detail why the government needs to be involved vs using private sector companies and tort law. Going back to first principles regarding the role of government is healthy especially in places like DC where government institutions are taken for granted and not questioned as much as they could be.]
One thing we should keep in mind is something that Milton Friedman said, which is that expecting the FDA to behave differently than it does is like expecting a cat to bark (Note: he said this in a 1973 Newsweek column). We can’t just say “we want the FDA to do X”, we have to make sure incentives are in place so people actually do the things we want. Legal mandates can help, but it’s easy for people to skirt around them if the proper incentives don’t exist.
We are facing an enormous cultural barrier when it comes to reforming the FDA and CDC. We don’t have a proactionary culture anymore, we have a very fear-based culture, and that’s not something there’s a simple solution to. However, we have a good opportunity right now just like the AIDS activists had a good opportunity in the 80’s.
The proactionary principle is a “grab-bag” of tools based on a certain value perspective which basically says that progress is fundamentally good. We aren’t omniscient, so we have to learn by doing. As Alex said, you can’t really learn things without making mistakes. It’s impossible to make progress, like some rationalists believed, by just sitting in chairs and thinking carefully. We have to become empirical. You can “look before leaping”, but you also have “look while leaping” and adjust how you land, to use a crude metaphor.
Cost-benefit analysis is a basic thing that is used in many organizations but doesn’t seem to be used as much in government agencies. It shouldn’t be controversial. Mandating cost-benefit analyses would be a step towards using ideas from the proactionary principle.
We should institutionalize the Devil’s advocate procedure and institutionalize respectful disagreement. Instead of having the most powerful person in the room getting what they want railroaded through we should require debate and force decision makers to ponder both sides. There’s other things we could do such as reference class forecasting, structured argumentation techniques, auditing procedures, and auditing review panels.
Reciprocity seems like a no-brainer that is relatively easy to achieve and would greatly reduce costs.
Besides getting out these great ideas, we need to figure out how to get people to follow those ideas. Laws can help, but people can choose to not follow them. How do we put “bite” into laws? I think an annual audit on the FDA’s decision making would be a good idea. Importantly, the auditor’s report should be made public. The auditors should come from a variety of institutions, for instance a variety of think tanks from different sides of the political spectrum.
Gennady Stolyarov II:
The USTP agrees that abolishing the FDA should not be out of the question. In our platform, section 118 states:
Section CXVIII [Adopted by a vote of the members during March 25-28, 2020]: Given the extreme delays, bottlenecks, and expenses created by the mandatory approval processes on the Food and Drug Administration (FDA), the United States Transhumanist Party supports abolishing the FDA and replacing it with a Radical Life Extension Administration (RLEA), whose mandate would be to prioritize the rapid development of potential disease cures, treatments, and vaccines – including any possible cures or vaccines for COVID-19, as well as treatments to mitigate and reverse the disease of biological aging, the major risk factor for COVID-19. The RLEA would allow the marketing and collection of patient data on any potential cure, treatment, or vaccine which has passed affordable safety testing at a reasonably acceptable threshold.
Jim O’Neill:
I’ve had the pleasure of working with the FDA quite a bit and in my experience most of the people there are very smart and they actually believe in approving things, contrary to what it may look like from the outside.
The problem is that incentives matter and the FDA is a central point of failure.
When someone has a severe side effect from a drug, the commissioner get hauled in front of several congressional committees and is interrogated. When someone dies because something wasn’t approved, there’s total silence in Washington. We should blame congress, not the FDA for that incentive system being in place.
Individuals respond to the institutional incentives, but they also have personal incentives. A lot of people want to be the next whistle-blower who finds the next thalidomide and calls a halt to it. Both of these are pretty severe and would effect even the most principled person in ways they couldn’t even detect.
I disagree with Alex that “FDA not recognizing aging as a disease is a major problem”. In order for the FDA reasonable measure success of any therapy, there must be metrics and biomarkers. It’s not FDA’s job to do all the scientific work to develop biomarkers for aging. That’s the job of the science community and NIH to some extent. There are epigentic clocks, but we need a lot more work on those. Those clocks can then be run through the FDA’s biomarker approval program.
The 2nd thing I disagree with is Dan’s idea of making the FDA independent from HHS. I think that would make things worse.
My favorite approval ideas call under the category of “progressive approval” or what Dan calls “tiered approval”. Contrary to what the FDA often thinks, doctors and patients are capable of processing information and making risk-benefit calculations using their knowledge about the specific situation they are in. The more information provided and the more transparency the better. The FDA should focus back on their original mission of safety and purity. I absolutely support repealing the 1962 Kefauver Harris amendments.
Prof. Garrett Jones:
I come at this as an macroeconomist. I think we can learn from what economists have learned about central banks around the world. The FDA should be as independent of congress and the president as central banks are or as judges are.
The Federal Reserve is a panel. That’s how we run the SEC, the FEC, the Federal Reserve, and the Supreme Court. There seems to be some magic to having a panel — it’s probably giving us a bit of the law of large numbers in decision making.
Another thing about these panels is they have long terms. They are probably going to be serving under a few different presidents. As I say in my book 10% Less Democracy, “short terms make short term thinking”. Political independence can lead to decision making independence and we have evidence that’s a good thing.
Discussions in institutional reform have “high marginal product” right now, as an economist would say. Congress moves slowly, but they tend to look for big opportunities for reform a couple years after a crisis. The Federal Reserve was born in 1913 was born out of the panic of 1907. Six years was how long it took between a huge financial crisis and Congress getting around to making some reforms. We saw something similar after the global financial crisis - it took about 2-3 years. The ideas that people are discussing now will be part of the information ecology of the next few years in Washington DC.
These ideas of long terms, independence, and panels are a good path for decision making. I am an unreformed Tabarrokian so I agree with everything Alex Tabarrok has written about FDA reform (chuckle). What I want to push here is institutional reforms that seem to work in a wide variety of settings. A little more financial and legal independence will lead to a situation where Congress is less a source of fear for FDA officials.
A lot of people on social media have told me that the president is in charge of the FDA. These people have never actually talked to anyone who worked on Capitol Hill — agencies live in fear of their Congressional overlords. They live in fear of the Senate Majority Leader and the Speaker of the House who have power over their budgets. They also know that if they make a mistake, they can be hauled up before Congress and fired ignominiously.
There is a risk that a more independent agency may misuse their freedom. However, in practice, if we look at the data, independent agencies with long terms. have high benefits and low costs.
It’s fun to complain about the FDA but it’s wise to complain about Congress.
Dr. Ed Hudgins
We’ve been talking about how FDA regulators are always in fear of Congress. What I want to see FDA regulators in fear of patients who want to get access to medications at less cost and quicker.
One of the most egregious examples of defining efficacy was when they decided that 23andme could not offer advice on whether someone was prone to breast cancer. Essentially they thought that women were too stupid to understand the information and would rush out a double mastectomy without getting a second opinion.
Another example is in 1989-1990 when they wanted to classify a urine sample cup as a “class A medical device”, in the same category as a heart valve.
In April 2019, the FDA wants to regulate artificial intelligence as a medical device.
There are many consultants now where their entire job is to help companies get through the FDA bureaucracy. So there’s a whole industry now just to help people get through the FDA — and that’s part of the problem now, too.
The “Free to Choose Medicine” idea should be at the top. Something like this was created around 1992 during AIDS crisis. Congress stepped in and put pressure on the FDA to do something. What they did was create a parralell track where sufferers could access a particular medication for AIDS during the three years it was being tested. 12,000 people took advantage of that, so there are 12,000 people who are not in the invisible graveyard as a result.
The idea of a parallel track has been put forward by Bart Madden. Data from people on that track would be put into a public real-world database.
There are alternatives to RCTs. If observational data is put into a public database, then doctors can look at that data and make informed recommendations. Drugs would be able to fail quicker, too.
In the case of AIDS, it was patient groups the besieged the FDA’s buildings. In light of COVID, and people seeing that the system isn’t working for them, we have an opportunity now to push for change.
There’s momentum for FDA reform building off the right-to-try legislation that has been passed in many states. In Texas and North Carolina there are strong pushes to broaden right-to-try to people like patients with Alzheimers.
Willy C.
The AIDS-FDA story is more little nuanced then was described by Ed Hudgins and Max More. We all know about the militant groups like ACT-UP which pressured the FDA in the 1980s. However in the mid-90s there developed a few counter-movments against that. One group was called Treatment Action Group, and they pressured the FDA to move slower because they felt the FDA was approving HIV/AIDs treatments that didn’t actually have much benefit.
A good book is Malignant by Vinay Prasad. It documents how, over the last twenty years or so, the FDA has lowered the standards for many cancer drugs. They often are approving drugs based on surrogate endpoints and biomarkers and then the drugs don’t go through follow-up studies to show if they have actual clinical benefit. So there’s been a natural experiment where we tried to lower the standards for cancer drugs and it doesn’t seem to have worked very well. Of course, I’m not an economist, so there might be a way of adding up the costs and benefits where the marginal benefits have outweighed the costs.
The FDA had many failures during COVID-19. The first big one was with testing, both with the FDA and the CDC. Others were the decision to delay the approval of Pfizer and Moderna (by about 4-6 weeks), and the approval not to approve AstraZeneca, which hasn’t had any transparency. Finally, there was a complete lack of experimentation with human challenge trials. What all of these share is there has been very little transparency and not much good reporting on these issues. There hasn’t been thorough investigations from journalists and we don’t really know what’s going on. Before attempting reform we need to first go and find out what went wrong during COVID-19. We need a non-partisan investigation of all of these issues. We need to utilize Freedom of Information Act requests. We need to find out how Trump was involved, why approvals took the amount of time they did, etc.
Dr. Natasha Vita-More
Cosmetics does not need FDA approval pre-marketing. It only needs post-market approval if the company says something in their marketing materials that could be misleading. There are many doctors pushing crack cosmetic treatments and behaving in a very “loosely goose-y”. I have a hard time understanding how they get away with these things, unless there are big monied interests pu
We all know about Theranos. In 2015 they got FDA approval for one of their tests. There’s clearly an imbalance here - many life-saving treatments struggle to improve but a company which is completely fraudulent like Theranos was able to get approval. [This is a great point! Theranos did receive approval, but only for their Herpes test. If I recall correctly, this test was done with conventional laboratory equipment rather then their special “minilab” device, a fact which Theranos hid from investors. Theranos also utilized a loophole to sell tests without FDA approval.]
Reform ideas mentioned by the panelists
Here is a summary of the reforms brought or alluded to by the panelists during the first hour, for what it’s worth:
Dan Elton
In favor (at least in exploring these ideas) - Improved transparency, reciprocity, independence, rolling reviews, adaptive licensing, free-to-choose medicine, tiered approval, structured cost-benefit analysis (proactionary principle).
Alex Tabarrok
In favor - reciprocity, adaptive licensing, using Bayesian analysis, aging as a disease, approver-of-approvers model.
Max More
In favor - reciprocity, improved transparency, structured cost-benefit analysis (proactionary principle), mandatory audits on decision making.
Garrett Jones
In favor - reciprocity, independence, run by panel not a commissioner, longer terms. Jim O’Neil
In favor - eliminating efficacy requirements, tiered approval.
Against - independence, classifying aging as a disease.
Ed Hudgins
In favor - Free to Choose Medicine
Willy C.
In favor - improved transparency, non-partisan commission to investigate COVID failings.