Vaccine Testing May Fail Without Human Challenge Trials

In Why Human Challenge Trials Will Be Necessary to Get a Coronavirus Vaccine I asked, “What if we develop a vaccine for COVID-19 but can’t find enough patients–healthy yet who might get sick–to run a randomized clinical trial?” Exactly that problem is now facing the Oxford vaccine in Britain.

An Oxford University vaccine trial has only a 50 per cent chance of success because coronavirus is fading so rapidly in Britain, a project co-leader has warned.

…Hill said that of 10,000 people recruited to test the vaccine in the coming weeks — some of whom will be given a placebo — he expected fewer than 50 people to catch the virus. If fewer than 20 test positive, then the results might be useless, he warned.

As I wrote, “A low infection rate is great, unless you want to properly test a vaccine.” Challenge trials have issues of external validity and they take time to setup properly but they produce results quickly and they can be especially useful in whittling down vaccine candidates to focus on the best candidates.

1DaySooner now has over 25 thousand volunteers from over 100 countries.


Thanks, Alex. I think this post contradicts Reinhart's speculations in your previous post

Also, read

It's time to think about the possibility that the COVID-19 pandemic has been a fraud much greater than Madoff's.

Bernie Madoff didn't force one of his sleeper cells to release a bioengineered virus and kill hundreds of thousands of people just because the President was on the cusp of exposing shit that's a lot more corrupt and disgusting than a pyramid scheme. It's not really fair to invoke his name here.

Sorry. If you mean that the COVID-19 pandemic would have been a fraud but not greater than Mueller's, maybe you are right.

"So they'll just wake up then?"
"What's that?"
"The dead, they'll just wake up then? What with it all being a fraud and all?"
"Oh yes... They'll be right as rain, any time now!"
"Well, that's a weight off my mind. I say! What about Uncle Joe? We had him cremated!"
"Look, I'll level with you. When I say they'll wake up, I mean they won't in fact wake up because it's not actually the kind of fraud that makes the dead come back to life. Instead it's the sort where the dead remain dead, but I get to consider myself much smarter than everyone else for realizing it's all a fraud."
"Ohhhh... I get it! You're where Prince Albert keeps his jewellery."

If Jacob Sullum and the CDC modelers who think this is the "best estimate" are correct, over 6 million people in NYC should have been infected already, or about 75% of the entire city population. With random anti-body testing, this hypothesis should be tested soon enough.

Please read
The estimates are relevant to the U.S. and they were updated last Wednesday, May 20. Indeed, they do not apply to the stocks or total number of deaths and infected people of any particular area in the U.S.
Yes, average values are terrible to analyze diversity but marginal values should be used to analyze flows.

Infection fatality rate is, by definition, a stock measure. It is the total number of fatalities divided by the number of infections.

If the reason for CDC's relatively low numbers is that it thinks large numbers of high-risk people have been infected already, that would add some clarity. Jacob Sullum should not be going around claiming a new "infection fatality rate" estimate, though. Nor does it mean that there was a "fraud" (exposed by the CDC and reported by CNN, no less!) since NYC (total population = 8 million) did indeed experience 16,000 deaths.

Maybe what that is saying is that living in NYC is dangerous. Some combination of closed ventilation systems and density is dangerous.

Coming up with a rate of infection or rate of death outside the contexts where it occurs is useless information.

You might be conflating propagation rate and mortality rate. These are completely different effect in long run. The actually numbers are unclear in most places; random population testing occurs rarely and even them is poorly done. There appears to be underreporting of deaths. People who really should know better are then making motivated guesses of the infection rate and to a lesser extent of the mortality rate allowing them to say what they like.

Maybe you haven't seen this amazing graph:

Let me explain it differently. We have a space problem and a time problem. The space problem refers to the U.S. v. NYC but the ratios used by the CDC are relevant only to the U.S. (the average for the total U.S. population is useless to analyze diversity). Now the average ratios for the total U.S. population refer to stocks at a particular time (either a day or a daily average for a month) but during an ongoing pandemic, we are also interested in the flows to estimate "marginal" values of IFR and other ratios paying special attention to flows during recent periods (a week, or a month).

Since all stocks are accumulated flows, stocks can change. A ratio of two stocks is an average over the relevant period of accumulation. As stocks change during a particular period, a ratio of the two flows is an average over this particular period and we may call it a marginal ratio.

How stupid are Americans, you ask?

Exhibit A

A German, of all people, might do well to consider the dangers of attributing to an entire nation the views of a few (even an unfortunately substantial few), to say nothing of making pejorative generalizations about entire populations.

A few? How many deaths have the Amerikans suffered due to their own arrogant stupidity? There won’t be a vaccine or treatment for American hubris or the imbecility of its people.

In Germany the virus has been under control for months. We trust science here, and in fact elect scientists as leaders.

Of course it took getting the crap beaten out of you losers in two world wars you started to achieve such enlightenment.

Imagine trusting the US CDC.

Re-opening so soon may be a blessing in disguise, a blessing for those who aren't infected in the second wave and get an efficacious vaccine attributable to human challenge trials with thousands of volunteers motivated by the second wave. As a bonus, the second wave might get rid of Trump.

"As a bonus"

The China Joe Biden side (here represented by rayward) is praying for more Americans to die, especially President Trump.

I would not wish a coronavirus death on anyone.

Indeed, most normal people wouldn't.

Leftists, however, will say and do anything to chase that dopamine hit and get their defective amygdala back to a tolerable state. In this case, as we see with our friend Ray Ray, that means rooting for a virus to kill as many Americans as possible and do maximum damage to everyone's economic health, all because their fellow citizens elected someone they don't like to be President and they're worried they might not be able to carry on their out-grouping charade anymore.

You can see why leftists are a cancer to civilization and why they must be exiled from the country ASAP. They're why we can't have nice things.

Liberals have special dispensation from the goddess to threaten violence if they don't get their way.

That's why they're always the ones pointing guns at people during their protest rallies.

The Trump tribe assumed that to "get rid of Trump" meant something nefarious, when all I meant is by popular vote. Popular vote is something the Trump tribe fears more than the coronavirus.

We know what you meant, Ray Ray. That's why the economy had to be taken behind the woodshed and mail-in ballots pushed so hard. So we can "get rid of Trump" using the "popular vote".

Actually, it's the electoral vote. Trump never won the popular vote. He's our affirmative action president.

"He's our affirmative action president."

Or maybe the player who read the rule book won the match?

I recall someone having quipped that the Russians hacked Hillary -- by removing her visits to swing states from her Google Calendar and replacing them with California visits. She let vanity or perhaps her desire to pile up money get in the way of refuting Trump's path to victory, which Nate Silver clearly explained about two weeks before the election.



The pharma industry was displaying its usual wisdom in not wasting money on vaccines? From the Guardian - The world’s 20 largest pharmaceutical companies undertook around 400 new research projects in the past year, according to Bloomberg Intelligence. Around half were focused on treating cancer, compared with 65 on infectious diseases.

There are eight potential vaccines for coronavirus in clinical trials, but there is no guarantee of success. One of the most promising, being developed at Oxford University, is said to have only a 50% chance of being approved for use.

The COE report says that rather than “compensating for market failures” by speeding up the development of innovative medicines, as per its remit, the IMI has been “more about business-as-usual market priorities”.

The report’s authors cite a comment posted on the IMI’s website, since removed, selling the advantages of the initiative to big pharma as offering “tremendous cost savings, as the IMI projects replicate work that individual companies would have had to do anyway”.

Here's an article in yesterday's NYT that is nominally about Moderna's inflated stock price but is actually about the rush among pharmaceutical firms to find a vaccine: As readers of this blog know, love of money is the root of all progress.

Then buy Merck! If the MMR vaccine pans out, the world will need just a few doses!

Rofl Europe and their american cast off groupies were doing so much chirping about American lagging in corona vaccines. Looks like some American drug companies will be able to acquire a few European firms for pennies on the dollar post bankruptcy.

I don't understand. The "problem" is that by the time a vaccine is ready for trial, the disease won't be serious enough for people to be willing to participate?

That's not what the article says. They have 10,000 volunteers lined up already but think that only 0.5% of their research subjects will actually be exposed to the virus. That means their trial will have weak statistical power.

The disease might not be running rampant but because of social distancing and lockdowns it might still be having serious costs.

Won't lockdowns and social distancing fade as more people have had the disease and it becomes harder and harder to contract within a given population? Seeing as it spreads much faster and more insidiously than the flu, I would have thought we'd have reached that point long before a vaccine can be rolled out at any sort of scale.

Lockdown and social distancing are simply not going to continue in the absence of a serious threat. Outside of a handful of strict authoritarian countries like China, that scenario exists only in the fantasies of the most extreme shutdown fetishists. The public's patience is limited: that was the key insight in Sweden's approach.

If the virus is not running rampant, restrictions will cease. Either they will be lifted in a controlled way by the authorities, or they will be ignored in a chaotic way by the public. And then the virus will either flare up again... or not.

The point is, a world where you can't find enough people to do a randomized clinical trial is a world where you don't need the vaccine anymore. See: SARS, Zika.

The problem is that regulatory delay will drag on for so long that the epidemic will be mostly over by the time they get to seriously trying out the vaccine. (The people behind the vaccine can't put it that way because they'd offend the regulators, who can be remarkably vindictive when criticized.)

Or maybe they're not planning to test in hot spots and on the people most at risk (doctors, meat packers...) like they should.

I like the spirit behind calling for challenge trials, but it doesn't seem to be the optimal way to do things. It's better than to go through the existing bureaucratic ways, for sure, but we don't need to be so extreme; we just need to jettison the time-wasting stuff that doesn't make sense. And extremism can backfire: suppose the first vaccine to go into challenge trials happens to be ineffective -- that being the most common way vaccines fail. Then you'll have injured tens or hundreds of people, and maybe killed a few; the press will be wallowing in their blood and calling for your head; and that boasted pool of thousands of volunteers will be drying up fast. It'll poison the atmosphere not only for challenge trials but for any sort of effort to get a vaccine out fast.

Oxford can't get more than 50 out of 10,000 volunteers to catch a virus so infectious and deadly it's required the world to be shut down for two months and counting.

Maybe they should send the volunteers to a barber or a beach? Oh I know—they should go out in public and stand only five feet away from each other without masks.

What a joke.

"can’t find enough patients to run a randomized clinical trial"
Isn't this why we have Brazilian slums, Syrian refugee camps?

Russian health care workers.

Most drug trials use mostly Indians, partly because India has so much drug industry production, partly because India is so British in economic structure, partly because of its large population.

Alex is the drunk searching for his dropped car keys under the lamp post because it's where the light it, not where he dropped his keys.

He's the reason there are no vaccines for malaria, dengue, etc, because you can't do testing in the US for such vaccines.

Yes that's why there isn't a vaccine for those diseases, really? It couldn't be that they are just more virulent than that cold you get every year. We have gotten really good at common cold vaccines too.

Now he's whining that the virus isn't infectious ENOUGH.

It's all over with, Alex. Don't worry, Trump will be out January 2025. You can make it! Hang in there!!!

We may already have a vaccine ready to go!

It seems that the old MMR might confer significant immunity due to similarities with the covid-19 virus and then cross reactivity.
As a recovered covid-19 patient I volunteer my blood today to enter a larger study of this.

With volunteers for a true pathogenic virus you have some major risk issues. Considering the value of speed by direct challenge, providing life and disability insurance to the volunteers at high levels should be required. Something in the range of 10 million life insurance would make economic sense.

India’s health ministry has revealed that their initial trial of hydroxychloroquine prophylaxis was successful and that the pool of people for which such prophylaxis is recommended is expanded. The ministry reported:

“A retrospective case-control analysis at ICMR has found that there is a significant dose-response relationship between the number of prophylactic doses taken and frequency of occurrence of SARS- CoV-2 infection in symptomatic healthcare workers who were tested for SARS-CoV-2 infection. Another investigation from 3 central government hospitals in New Delhi indicates that amongst healthcare workers involved in COVID-19 care, those on HCQ prophylaxis were less likely to develop SARS-CoV-2 infection, compared to those who were not on it. The benefit was less pronounced in healthcare workers caring for a general patient population.
An observational prospective study of 334 healthcare workers at AIIMS, out of which 248 took HCQ prophylaxis (median 6 weeks of follow up) in New Delhi also showed that those taking HCQ prophylaxis had lower incidence of SARS-CoV-2 infection than those not taking it.”

So as news of this success slowly leaks out in the west, the number of potential vaccine testees will decline even more rapidly.

Human challenge trials means sacrifice, something Americans can certainly appreciate on this Memorial Day. Today's news is that social security is, well, insolvent, in the sense that payments out exceed payments in. Not surprising since payrolls, and hence payroll taxes, have collapsed during the pandemic. Not to worry since there is an almost $3 trillion social security trust fund. Of course, it is and has always been a fraud: the largest working American tax increase, adopted during the Reagan administration, has been more than offset by the largest investor class tax cut, adopted during the Reagan, Bush, and Trump administrations, depleted the "fund" to zero. The non-existent "fund" is measured by the difference between payroll taxes collected and benefits paid, or almost $3 trillion. Do you feel rich or poor? Well, the rich feel rich while the rest of us feel, well, poor. America is about to experience a coming to Jesus moment, but I don't expect Americans will come to Jesus. The devil is more likely. Suckers!

The non-existent fund is as non-existent as any other investment in US debt. If the US defaults, all bets are off.

In 2009, the vaccine makers were left holding the bag on H1N1 vaccines as H1N1 ran its course and countries cancelled their orders en masse, leaving the vaccine makers high and dry. There's a window for the vaccine makers: The vaccine is most useful on day 1, but its value generally diminishes a bit every day as people adapt and blunt Ro, get infected, build immunity, etc.

A vaccine generally sells for a buck or two per dose. Your potential upside for selling the entire US a vaccine might be $500M. The downside, however, is huge: A big chunk of your R&D is redirected for a year onto a solution that won't bring in much revenue, you will undoubtedly have a lot of legal engagement needed as people claim side effects, and there's a very real chance most of the world will cancel orders if the virus peters out. In 1976, a flu vaccine gave some recipients Guillian-Barre syndrome. The H1N1 vaccine apparently induced narcolepsy in people. The federal govs will usually take on the risk for these events, but it's still a lot of time and effort for the vaccine makers sort out.

I'm glad I'm not in that business. But I do understand the sense of pride the scientists take in working to be first. How exhilarating that must be. Seldom are their Olympic style events for the professional class.

First let us try it on the rats.

It doesn't go into humans until it is first tested in the test tube, then in animals. Then Phase I testing is a small scale safety study in humans.

And if the vaccine promises to be 50% effective, what do you do then?

To turn it around, how effective would the vaccine have to be for you to lose all caution going forward, Alex?

No vaccine is 100% for all takers. Best guesses and effort, and than tincture of time to see its large scale and long term effect.

And the problem is?

A big problem with challenge studies is that the effectiveness of the vaccine may be different for people with characteristics that induce a high case fatality rate, which is exactly the group we most want to protect. No conceivable ethics panel this side of China would allow trials on people highly likely to have serious permanent adverse consequences if vaccine fails.

It is true that if almost everyone who is likely to survive a COVID-19 infection gets vaccinated and becomes immune the high risk members of the community will be protected by herd immunity, but that would leave potential hot spots such as retirement homes and nursing homes.


One idea is to use a vaccine already proven safe and effective.
That being the MMR in the very new news. Safe and cheap and no need for human trials. There is apparent cross reactivity of antibody effect between rubella and covid 19.

Today I donated a bit of blood as a recovered covid 19 patient with positive IgG antibodies.

We might know more in just a few weeks.

Comments for this post are closed