Category: Science

Solve for the rational Bayesian equilibrium

I now read quite a few public health experts on matters of the day, and I have noticed that none of them have condemned the British government for proceeding with the AstraZeneca vaccine trial, even after two adverse health events experienced by participants, noting that those events presumably have been examined and considered by the oversight committee.

At the same time, the American trial for AstraZeneca has remained halted.  I also have not read any public health experts criticizing that decision either.

What is the most likely equilibrium to be holding here?

1. Public health experts don’t express many opinions, especially these days.

2. Plenty of commentators think the British decision to resume is rash, Tyler just isn’t reading enough of them.

3. Most public health experts think it is fine for the British to keep on going.  But they won’t criticize the American trial halt, because their incentives and natural temperamental tendencies are to express mainly the risk-averse opinions, and rarely if ever say that the regulatory process should allow for more risk to be taken.

4. The mainly American experts actually are happy to see America free-riding upon British data, so they are content with things as they stand, but don’t want to quite come out and admit they enjoy exploiting the Brits.

5. In reality the commentators think the whole trial is so risky it never should have been started in the first place.

6. What they really enjoy writing is philosophical pieces about how social process have all these twists and turns, and natural bumps in the road, and so they don’t wish to work too hard to remove those bumps.

7. The public health experts think that Americans and British have optimally different tolerances for risk, and the split regulatory outcomes reflect that difference.

Your choice?

The public is fine with Human Challenge Trials

A vaccine for COVID-19 is urgently needed. Several vaccine trial designs may significantly accelerate vaccine testing and approval, but also increase risks to human subjects. Concerns about whether the public would see such designs as ethically acceptable represent an important roadblock to their implementation, and the World Health Organization has called for consulting the public regarding them. Here we present results from a pre-registered cross-national survey (n= 5; 920) of individuals in Australia, Canada, Hong Kong, New Zealand, South Africa, Singapore, the United Kingdom, and the United States. The survey asked respondents whether they would prefer scientists to conduct traditional trials or one of two accelerated designs: a challenge trial or a trial integrating a Phase II safety and immunogenicity trial into a larger Phase III efficacy trial. We find broad majorities prefer for scientists to conduct challenge trials (75%, 95% CI: 73-76%) and integrated trials (63%, 95% CI: 61-65%) over standard trials. Even as respondents acknowledged the risks, they perceived both accelerated trials as similarly ethical to standard trial designs, and large majorities characterized them as “probably” or “definitely ethical” (72%, 95% CI:70-73% for challenge trials; 77%, 95% CI 75-78% for integrated trials). This high support is consistent across every geography and demographic subgroup we examined, including people of diverging political orientations and vulnerable populations such as the elderly, essential workers, and racial and ethnic minorities. These findings bolster the case for these accelerated designs and can help assuage concerns that they would undermine public trust in vaccines.

Here is the paper by David Broockman, et.al.

*Editing Humanity*

That is the new book by Kevin Davies, and the subtitle is The CRISPR Revolution and the New Era of Genome Editing.  So far I am on p.74, but it is one of the best science books I have read in some while, maybe the best this year.  Excerpt:

…Cas9 normally takes about six hours to search through every PAM sequence in the bacterial genome, pausing at each prospective site for a mere twenty milliseconds to peer into the double helix to see if it has found the correct target.  But the packaging of DNA in a eukaryotic cell nucleus is far more complex than bacteria.  During lectures to his students at the University of Edinburgh, Andrew Wood shows a diagram of a bacterial cell alongside a winding, looping mammalian DNA fiber.  “Cas9 didn’t evolve to work in the environment in which we now put it,” he says.  “It’s mind-boggling that it is possible to interrogate hundreds of millions of nucleotides in a matter of hours.”

Recommended.  Emmanuelle Charpentier and Jennifer Doudna, the two Nobel winners from last week, are so far the central characters of the story.

Are Nobel Prizes worth less these days?

It would seem so, now there are lots of them, here is one part of my Bloomberg column:

The Nobel Peace Prize this year went to the World Food Programme, part of the United Nations. Yet the Center for Global Development, a leading and highly respected think tank, ranked the winner dead last out of 40 groups as measured for effectiveness. Another study, by economists William Easterly and Tobias Pfutze in 2008, was also less than enthusiastic about the World Food Programme.

The most striking feature of the award is not that the Nobel committee might have gotten it wrong. Rather, it is that nobody seems to care. The issue has popped up on Twitter, but it is hardly a major controversy.

I also noted that the Nobel Laureates I follow on Twitter, in the aggregate, seem more temperamental than the 20-year-olds (and younger) that I follow.  Hail Martin Gurri!

And this:

The internet diminishes the impact of the prize in yet another way. Take Paul Romer, a highly deserving laureate in economics in 2018. To his credit, many of Romer’s ideas, such as charter cities, had been debated actively on the internet, in blogs and on Twitter and Medium, for at least a decade. Just about everyone who follows such things expected that Romer would win a Nobel Prize, and when he did it felt anticlimactic. In similar fashion, the choice of labor economist David Card (possibly with co-authors) also will feel anticlimactic when it comes, as it likely will.

Card with co-authors, by the way, is my prediction for tomorrow.

Does economic value reside in the brain?

I do not feel qualified to have an opinion here, but this piece, by Benjamin Y. Hayden and Yael Niv, seems of some interest:

Much of traditional neuroeconomics proceeds from the hypothesis that value is reified in the brain, that is, that there are neurons or brain regions whose responses serve the discrete purpose of encoding value. This hypothesis is supported by the finding that the activity of many neurons and brain regions covaries with subjective value as estimated in specific tasks. Here we consider an alternative: that value is not represented in the brain. This idea is motivated by close consideration of the economic concept of value, which places important epistemic constraints on our ability to identify its neural basis. It is also motivated by the behavioral economics literature, especially work on heuristics. Finally, it is buoyed by recent neural and behavioral findings regarding how animals and humans learn to choose between options. In light of our hypothesis, we critically reevaluate putative neural evidence for the representation of value, and explore an alternative: that brains directly learn action policies. We delineate how this alternative can provide a robust account of behavior that concords with existing empirical data.

Via Benjamin Lyons.

A Cost/Benefit Analysis of Clinical Trial Designs for COVID-19 Vaccine Candidates

I am very happy to see this new and urgently needed study.  They have heeded the stricture to show their work.  The authors are Donald A. Berry, Scott Berry, Peter Hale, Leah Isakov, Andrew W. Lo, Kien Wei Siah, and Chi Heem Wong, and here is the abstract:

We compare and contrast the expected duration and number of infections and deaths averted among several designs for clinical trials of COVID-19 vaccine candidates, including traditional randomized clinical trials and adaptive and human challenge trials. Using epidemiological models calibrated to the current pandemic, we simulate the time course of each clinical trial design for 504 unique combinations of parameters, allowing us to determine which trial design is most effective for a given scenario. A human challenge trial provides maximal net benefits—averting an additional 1.1M infections and 8,000 deaths in the U.S. compared to the next best clinical trial design—if its set-up time is short or the pandemic spreads slowly. In most of the other cases, an adaptive trial provides greater net benefits.

And what is an adapted trial you may be wondering?:

An adaptive version of the traditional vaccine efficacy RCT design (ARCT) is based on group sequential methods. Instead of a fixed study duration with a single final analysis at the end, we allow for early stopping for efficacy via periodic interim analyses of accumulating trial data…While this reduces the expected duration of the trial, we note that adaptive trials typically require more complex study protocols which can be operationally challenging to implement for test sites unfamiliar with this framework. In our simulations, we assume a maximum of six interim analyses spaced 30 days apart, with the first analysis performed when the first 10,000 subjects have been monitored for at least 30 days.

That means of course you might cut the trial short.  Kudos to the authors for producing one of the most important papers of this year.

Tyrone on clinical trials and how to keep them up and running

Tyrone — my evil twin brother — received so much hate and love mail from his recent pronouncements about QAnon that he felt emboldened to offer additional opinions.  As you might expect, he prefers to spew his hateful bile on matters of life and death.  In particular, he has been following the debates about Covid and whether new treatments should be accelerated in their availability.  Anyway, I told him I was willing to pass along another of his letters, as a kind of experiment (not quite a clinical trial) whether Tyler or Tyrone is a more beloved writer on MR.  I am sure you readers — and especially commentators — stand ready to defend my honor!

So here is his (as usual) fallacy-ridden missive:

Tyler, I don’t see why you let the defenders of FDA stalling get away with their dawdling.  They all end up with the same argument — if we let wonderful, salt of the earth Americans take beneficial medicines, treatments, and vaccines, we will not be able to set up informative clinical trials.  Why partake in the trial when you can just get the stuff through normal means?

That is so lame!  First, they could simply pay people to partake in those trials.  Isn’t that in essence what the NBA did with its Covid testing in the bubble?  If the value of those clinical trials truly is so high, it should be possible to internalize enough of those benefits to encourage participation.  If institutional barriers stand in the way there, let’s obsess over fixing those.

Why should we force so many Americans to be sacrificial lambs, just to subsidize the trial costs?  Let those costs be taken out of grant overhead!  (And admin. salaries, if need be.)

If the current medical establishment is not as able as the NBA, well OK, can’t they just admit it and plead patheticness?  We can send them to take care of Major League Baseball, and put Adam Silver and Lebron James in charge of our health care.

Second, there is another way to keep the trial up and running.  Approve use of the treatment, but allow the suppliers to charge very high prices!  Better yet, use the law to make them charge high prices and if need be forbid insurance coverage.

“What will it be sucker? Fifty percent chance of the placebo, or 100k for those monoclonal antibodies?”

I assure you Tyler that will restore a separating equilibrium.  Furthermore, in the meantime only the most meritocratic of wealthy men will get the treatment outside of the trial, all for the better.  If need be, you can pull away the price floor when the clinical trial is complete, in the meantime you have satisfied the Pareto principle.

And what about the Hippocratic Oath ?  “Do no harm”?  Is that not invoked so selectively by the public health commentators?  Surely you realize they court public opinion and high status by taking sins of commission far more seriously than the far less visible sins of omission?

Is it not harm to deny patients ready accessibility to a treatment with positive expected value?

Is it really such a great rejoinder to insist “We can’t let those patients improve their lot by raising pecuniary costs for the medical professionals running their trials!  That is true Hippocratic harm and must be avoided at all costs, because in fact we medical people would be too feckless to overcome that problem…”

Sigh.  At that point I had to stop reading and transcribing.  I am sorry readers, I didn’t know that Tyrone in his spare time was studying economics and indeed some logic as well.  Maybe he has even been reading MR.  That makes him less interesting, less funny, and maybe a bit too much like Tyler.  That is not why you come to read Tyrone, and indeed you might as well be reading Tyler.

What can I do to make Tyrone better and more eccentric again?  Perhaps try to get him premature access to some of those special treatments?  Stay tuned….

Is the treatment positive expected value, or negative expected value?

Well, which one is it?

If you consider the treatments of remdesivir or monoclonal antibodies for President Trump, their application is either positive expected value or negative expected value.

If they are positive expected value, you should be for using them!  (I don’t mean that as a political statement, sub in another patient’s name if you need to.)

If they are negative expected value, you should oppose the current widespread use of remdesivir in hospitals (not necessarily in every case, of course), and you should probably oppose the Advance Market Commitment already in place for Regeneron’s monoclonal antibody treatment, not to mention its successful advance through various trials.

I don’t see anyone taking those stances.

Instead, I see commentators — including highly esteemed public health experts — claiming there is not yet enough data, “expressing reservations,” referring to other public health catastrophes, referring to more general irresponsible habits of the patient under consideration, and serving up various other rhetorical devices to indicate a negative attitude toward the treatment without actually saying “I think this treatment is negative expected value.”

That is a very bad thought and writing habit!

Made worse by Twitter, I might add.  You are trying to create negative affect and mood affiliation without making the corresponding epistemic and predictive commitment.

Please just say you think it is negative expected value, and then apply that view consistently across the board.  Stand your ground and defend it.

Or if you think it is positive expected value, praise its use, and then of course it is fine to add qualifiers and reservations.

If you genuinely have no opinion (ha), it is fine to say that too, but then you can drop the negative rhetoric and maybe don’t tweet about it at all.

To be sure, there are various heterogeneities and I am not applying the appropriate qualifiers in each sentence above, for reasons of expositional convenience.  For instance, is Trump different from other patients?  Are the treatments being applied at the right time?  Who exactly has the private information here?  And so on.  Incorporating those factors should not change the basic analysis above, though for the most part they should push you toward a more positive attitude toward the treatments.

The problem with rapid Covid testing

Mayank Gupta emails me:

The absolute number of false positives would rise dramatically under slightly inaccurate, broad surveillance testing. At least initially, the number of people going to the doctor to ask what to do would also rise. One can imagine if doctors truly flubbed and didn’t know how to advise patients accurately, a lot of individual patients would lose trust in the medical system (testing, doctors, or both). The consequence of this would be more resistance to health public policy measures in the future.

I see this as quite similar to what happened on three mile island. There was a clear utilitarian benefit to taking on some small amount of nuclear accident risk. The public was never taught how or why to internalize and cope with this risk. When the risk manifested, individuals saw the risk but not the public benefit and turned against nuclear. This change of public opinion was reflected in public policy after.

I’m sure there’s some mismatches in this analogy, but I’m using it to point out that in general I find thinking on the margin without thinking about the public’s ability to think on the margin can result in setbacks on the margin.

In an age where science is both more capable of solving social problems and more complex to understand, but the public has too much complex information to sort through, the central problem of governance seems to be how to solve public choice, without creating a monopoly on information.

To be clear, I do favor rapid testing, but it is worth giving this problem further thought.

The elasticity of science

By Kyle Myers:

This paper identifies the degree to which scientists are willing to change the direction of their work in exchange for resources. Data from the National Institutes of Health are used to estimate how scientists respond to targeted funding opportunities. Inducing a scientist to change their direction by a small amount—to work on marginally different topics—requires a substantial amount of funding in expectation. The switching costs of science are large. The productivity of grants is also estimated, and it appears the additional costs of targeted research may be more than offset by more productive scientists pursuing these grants.

Here is the full article in American Economic Journal: Applied Economics.

Ho hum

Compact Nuclear Fusion Reactor Is ‘Very Likely to Work,’ Studies Suggest (NYT)

What else have we got this year?  Visiting alien drones, life on Venus and Mars, super-acceleration of vaccine production techniques, Tesla valued at mega-levels, and plummeting prices for solar panels.

Remember what I wrote in the final section of The Great Stagnation?  I predicted you’ll all be super-pissed off when it ends.

Decentralized serological testing?

I would like to know more, but here is one new paper on the topic, by Lottie Brown, et.al.:

Serological testing is emerging as a powerful tool to progress our understanding of COVID-19 exposure, transmission and immune response. Large-scale testing is limited by the need for in-person blood collection by staff trained in venepuncture. Capillary blood self-sampling and postage to laboratories for analysis could provide a reliable alternative. Two-hundred and nine matched venous and capillary blood samples were obtained from thirty nine participants and analysed using a COVID-19 IgG ELISA to detect antibodies against SARS-CoV-2. Thirty seven out of thirty eight participants were able to self-collect an adequate sample of capillary blood (≥50 μl). Using plasma from venous blood collected in lithium heparin as the reference standard, matched capillary blood samples, collected in lithium heparin-treated tubes and on filter paper as dried blood spots, achieved a Cohen′s kappa coefficient of >0.88 (near-perfect agreement). Storage of capillary blood at room temperature for up to 7 days post sampling did not affect concordance. Our results indicate that capillary blood self-sampling is a reliable and feasible alternative to venepuncture for serological assessment in COVID-19.

Via Alan Goldhammer.

Where do economics journal editors live and work?

Over half the journals we consider have over two thirds of their editorial power located in the USA. A large majority of journals have a tiny editorial contribution from academics located outside of North America and Europe. Any one of the states of California, Massachusetts and Illinois has more power than the four continents of Asia, South America, Africa and Australasia combined.

That is from a new paper by Simon D. Angus, Kadir Atalay, Jonathan Newton, and David Ubilava.  Here is a useful visual showing the actual distribution.