Yesterday, I quoted John Ioannidis’s description of how the pandemic was changing the norms of science:

There was absolutely no conspiracy or preplanning behind this hypercharged evolution. Simply, in times of crisis, the powerful thrive and the weak become more disadvantaged. Amid pandemic confusion, the powerful and the conflicted became more powerful and more conflicted, while millions of disadvantaged people have died and billions suffered.

First, note that no one person or thing (or a few countable ones) was responsible for the transformation, but rather that it was a natural property of the system. Then note that the pandemic sped up (or “hypercharged”) the system’s transition to its (breaking) endpoint. The American clinical trial ecosystem is even closer to its breaking point than science overall, and for the same reasons: its postulates make it untenable, and the influx of ZIRP-conjured pandemic-stamped money hypercharged the transition to this.

People go looking for ways to speed up clinical trials like they would for speeding up house construction. They ask: what is one key piece of legislation — the parking and zoning requirements of clinical research — that could either be tweaked or removed to break open the damn, so that trials would be as fast as they were in the 1960s? And why won’t the clinical trial hobbits tell us what those obstacles are, so that we can crush them?

But there is no one, or two, or two dozen such obstacles, the serial removal of which could speed up trial design, approvals, execution and readouts. The clinical trial path does not have boulder problem. Boulders would in fact be great: at least you can, while climbing over or making a 3-day detour around, fantasize about crushing them.

No, the problem with the path is that it runs through quicksand while being covered by a dense fog. Move too fast or stand in place and you die; move too slowly and you don’t get anywhere. Not that you know where to go: with all-encompassing fog you can only see five inches ahead anyway. There is not much time to think about anything else when you are in that kind of a quagmire, and how exactly would you imagine dealing with the fog? Drive it away with a big fan?

By “quicksand” I mean the moral, ethical and physical safety risks inherent to any clinical trail ⊕ This is why every introductory course on the topic must start with the ignominious history of experimenting on humans… and the systems we developed to deal ⊕ …immediately followed by a description of Institutional Review Boards and other regulatory matters. with them.

By “fog” I mean that much of the deliberation is not a matter of legislation but of opinion, values, principles and — the most loaded of words — comfort of people sitting on these IRBs and in the regulatory agencies. Depending on comfort level of individual members a question on whether a trial can proceed may take much deliberation or none at all. We can hardly know our own minds, and good luck about reading other people’s.

This is why I am skeptical that uncovering the FDA’s records would be of help to anyone but historians: the people who wrote them and interpreted them are no longer in play; different brains run the show, and the archives won’t reveal their thought process. Sure, sometimes there is one person sitting there that makes things worse than they should be, by being too slow, or thoughtless, or obstinate. But is “find better people” a tenable solution, when those people could be doing anything else, and with more reward of all kinds?

The pandemic hypercharge made everything worse. In addition to wading through quicksand while blinded by fog you now have to deal with many, many invisible neighbors elbowing you in the chest and kicking you in the groin, some intentionally, some out of carelessness, confusion, or not knowing any better. IRBs are overloaded and so are the regulators. What difference would any legislative change make?

Except, of course, it is to abolish one or both. It would lift the fog, sure, but would need a new kind of system to avoid making everyone sink to the bottom of the ethico-moral pit. Tearing down institutions is easy, building them is hard, as we see one but not the other being played out under our inattentive eyes.

• Bart’s Watch Stories on YouTube: How Casio Made the Most Sold Watch Ever. It is about Casio F-91W, my default watch, and I can attest to its — excuse the pun — timelessness.
• @StatisticUrban on X: What states do Americans most approve of? DC is the lowest at +1 net approval.
• Ruxandra Teslo: Your newborn is not Hepatitis B vaccinated because of wokeness. See also her exchange on X with Curtis Yarvin, aka Mencius Moldbug, aka Mr Head-Up-Ass who loves the sound of his one voice even as he recommends disastrous remedies for obvious problems.
• Marioun Fourcade and Kieran Healy: Authenticate thyself. The subheading is “The sovereign individual and the paradox of the digital age”. It is more highfalutin than what I usually read — “The combination of epistemological self-centredness and hyperconnectivity makes people susceptible to diffuse forms of ‘supersense’-making (to borrow a term from Hannah Arendt).”, reads one sentence — but the essay does some important sense-making of its own:

As the deployment of digital technologies continues to generate ever-more stratospheric concentrations of wealth, the masses sink deeper into the void left by the evisceration of social solidarity and the rise of automation. The often-missed point about sovereign individuals is that not everyone gets to be one. But everyone should aspire to be one, and in the meantime follow one, as they walk down the road to selfdom.

Worth reading for that last sentence alone.

Out today in Annals of Clinical and Translational Neurology: Durability of Response to B-Cell Maturation Antigen-Directed mRNA Cell Therapy in Myasthenia Gravis. It only took 18 months to get here from the pre-print but hey, we were able to get longer follow-up!

Whatever you think of medicalization of moderate obesity, the GLP-1 inhibitors semaglutide and tirzeparide (aka Ozempic, Wegowy and Zepbound) are truly groundbreaking. It takes a lot for me to admit something approaches imatinib in innovation and importance, and they are there! Incredibly, the drug companies that developed them are considered losers in the upside-down world of American finance:

Since their peak last year, the decline is more pronounced. Novo Nordisk has lost $367bn in value since its peak in June 2024, a fall of more than two-thirds, while Lilly has fallen 29 per cent from a record valuation last year, wiping $250bn off its market capitalisation.

In a decision that was short-termist and reactionary to the extreme, Novo Nordisk even fired their longstanding CEO over it.

The kicker comes from a healthcare fund manager quoted near the end of the article:

“If you’re a generalist investor, why are you putting money here, versus buying an AI stock, [given] the headwinds of both tariffs and the most favoured nation policy?” he added.

What are we even doing here?

The Occasional Human Sacrifice puts faces, personalities, anxieties and neuroses to the names of people who acted as whistleblowers to some of the biggest ethical failures in clinical trials. Some are textbook, like Tuskegee or the Willowbrook hepatitis studies, but many were either new to me, or just barely registered when they were briefly covered.

Elliott was himself a whistleblower in the case of Dan Markingson so he is hardly impartial to their cause — caveat lector — but the cases presented seem truly egregious. And not all of them are ancient history: Paolo Machiarrini experimented on humans without oversight as recently as 2014.

The picture you get is bleak and does not fill one with confidence about clinical research anywhere in the world. Physician-scientists are careless at best, selfish profiteers at worst, people who sit on ethics committees are a bunch of box-checkers, institutions are insular and protective of their own. Of course, there is major selection bias going on: yes, institutions protect their own but then there are many of their members who are accused daily of misconduct by conspiracy theorists, biopharma lobbyists, and the occasional psychopath. Some IRBs are indeed approval mills, but then there are those which truly protect research subjects, though alas what they do doesn’t make it into a book. How to tell where the equilibrium should lie?

In fairness to the book, it does not pretend to be a grand unifying theory of what is wrong with medical research. It is a collection of vignettes, no more and no less, and as such is an important source of “real-world” information to the research community. It is also a big honking red flag to any person thinking to blow the whistle on wrongdoings medical or not: it is a difficult path to take, with no vindication at the end.

Is lithium deficiency an important factor in developing Alzheimer’s disease? A recent paper in Nature provides some convincing evidence, mostly from mice. For example:

Replacement therapy with lithium orotate, which is a Li salt with reduced amyloid binding, prevents pathological changes and memory loss in AD mouse models and ageing wild-type mice.

I know quite a few doctors who would say lithium (carbonate) deficiency is responsible for many behavioral issues in adults, but this is not what they had in mind! (↬Derek Lowe)

Well that was fast:

“At the FDA’s request, Dr. Vinay Prasad is resuming leadership of the Center for Biologics Evaluation and Research,” HHS spokesperson Andrew Nixon said. “Neither the White House nor HHS will allow the fake news media to distract from the critical work the FDA is carrying out under the Trump administration.”

Maybe lobbying isn’t as effective as I thought? I am sure there are stories to be told about what happened during these two weeks.

Here are a few good comments on what recently happened at the FDA:

• Washington Post: What the ouster of a top FDA regulator shows about Trump world divisions
• The Atlantic: The Man Who Was Too MAHA for the Trump Administration
• Science: Vinay Prasad: That Was Fast

V(inay) P(rasad)’s ouster was clearly death by lobbyists, but then they had plenty of fodder. The current administration does seem to be going for a mid-to-late 19th century vibe in many ways, and this is one of them. Sure, Ulysses Grant probably didn’t coin the word, but isn’t that the peak period when those who had the president’s ear could get things done quickly and blatantly? Whether that excites you or scares you, well, that depends on what kind of person you are and what you do for a living.

Selfishly speaking, it will be good to see VP back publishing oncology papers. Here is a recent one about informative censoring in clinical trials, with a lay summary here. More of that, please.

A few choice excerpts from a NYT investigation, Medicare Bleeds Billions on Pricey Bandages, and Doctors Get a Cut:

For one patient in Nevada, Medicare spent $14 million on skin substitutes over the course of a year, according to billing records reviewed by The Times. The wound of a patient in Washington State persisted after Medicare paid $6 million for the coverings. A man in Texas got $1.3 million of bandages despite having no wound at all.

Five years ago, the most expensive skin substitute cost $1,042 per square inch, while some were as cheap as $45. Today, the three most expensive products on the market each cost more than $21,000. (Samaritan Biologics, a company in Memphis that sells the three products, did not answer questions about why they cost so much.)

For the first six months of a new bandage product’s life, Medicare will set the reimbursement rate at whatever price a company chooses. After that, the agency adjusts the reimbursement to reflect the actual price paid by doctors after any discounts.

To circumvent the reimbursement drop, some companies simply roll out new products.

The doctor who earned the most for skin substitutes last year was Dr. Aaron Jeng of Southern California, according to Early Read’s analysis. Medicare paid him $117 million. (Dr. Jeng declined to comment.)

Another high earner, Dr. Stephen Dubin of Las Vegas, was paid $17 million by Medicare for skin substitutes in 2024. (He estimated that after expenses, he took home roughly $4 million.) Dr. Dubin retired at the end of last year, in part, he said, because of increased competition for wound patients. Sometimes he would show up at a patient’s home only to find that someone from a different clinic had placed a new skin substitute the day before.

The article is 3 months old but still relevant: a day after it came out the administration announced that it was indeed delaying implementation of the new reimbursement rules until 2026. Wouldn’t it be neat if there were a government department that deals with this kind of fraud, waste and abuse?

Microsoft claims their new medical tool is “four times more successful than human doctors at diagnosing complex ailments”. Unsurprisingly, what they meant by “diagnosing a disease” was the thinking-hard part, not the inputs part:

To test its capabilities, “MAI-DxO” was fed 304 studies from the New England Journal of Medicine (NEJM) that describe how some of the most complicated cases were solved by doctors. 

This allowed researchers to test if the programme could figure out the correct diagnosis and relay its decision-making process, using a new technique called “chain of debate”, which makes AI reasoning models give a step-by-step account of how they solve problems.

If and when deployed, how likely is it that these algorithms will get a query comparable to a New England Journal of Medicine case study? Most doctors don’t reach those levels of perception and synthesis, let alone the general public.