Sunday, January 10, 2021

FDA vs. Astra-Zeneca; bureaucracy vs. evolution and exponential growth

 From Alex Tabarrok at Marginal Revolution, quoting Marty Makary, M.D., a professor of surgery and health policy at the Johns Hopkins University School of Medicine:

... the FDA needs to stop playing games and authorize the Oxford-AstraZeneca vaccine.  It’s safe, cheap ($2-$3 a dose), and is the easiest vaccine to distribute. It does not require freezing and is already approved and being administered in the United Kingdom.

Sadly, the FDA is months away from authorizing this vaccine because FDA career staff members insisted on another clinical trial to be completed and are punishing the company for inadvertently giving a half-dose of the vaccine to some people in the trial.

It’s like the FDA is holding out, pontificating existing excellent data and being vindictive against a company for making a mistake while thousands of Americans die each day...

My emphasis. Alex:

See also my post The AstraZeneca Factory in Baltimore. Thousands of people are dying every day. We have a vaccine factory ready to go. The FDA should lifts its ban on the AstraZeneca vaccine.

Alex understates the case. It is not just that "thousands of people are dying every day." It is that we are in the phase of exponential growth, and a new more infectious variant has just arrived bumping up the growth rate further. Every hour of delay means tens of thousands more will die.  

We are in a fight of bureaucracy vs. exponential growth and evolution. Exponential growth and evolution are winning. Just how many thousands have to be on the left side of the trolley switch before the FDA stops allowing Astra-Zeneca to pull it? What's the risk aversion coefficient that justifies months of delay and another clinical trial?  

More deeply, can the FDA ever figure out that the point here is to stop a pandemic? The mentality is traditional: we must provide a perfect vaccine to protect individuals, taking the disease as given, and people who die while we do more studies are worth the cost. That is simply not what's going on right now. The point of the vaccine is to stop a pandemic. The disease is growing exponentially, and mutating and evolving. The externality is everything. I know, it's awfully hard for bureaucracies to innovate and change mindset. Well, sometimes you have to.  

For years the FDA was focused on, don't repeat thalidomide. Drugs must be safe. AIDS forced a hard reckoning. The people who are dying while you wait matter. But this is a third, even harder conceptual change. Stopping the spread of the disease matters. And the FDA does not have the years it took to make the AIDS change of mindset. 

Alex, by the way, gets the award for most influential economist of the year. It looks like his campaign for first doses is going to be Biden Administration policy. Well done Alex! 

But it turns out vaccine supply is not even the short run constraint, it is unbelievably snafued rationing rules. Virginia Postrel, summarizing many stories you've heard: 

When the federal government turned state agencies into the country’s vaccine distributors, it bypassed the usual supply chains. Doctors and hospitals couldn’t get Covid-19 vaccines the way they order other inoculations.

Distribution also became politicized in ways that slow down vaccination. Every shot comes with a ton of paperwork, and the rationing rules are hard to understand. Who exactly qualifies as a health-care worker or an essential employee? Is it OK for hospitals to give shots to janitors or billing clerks?

n Minnesota hospitals, one doctor who asked to remain anonymous noted in an interview, “there was a lot of focus on scheduling appointments and dividing up by departments to be sure they were fair” even if that meant delaying vaccines and potentially letting some supplies go to waste. It’s a widespread problem.

As he threatens fines for hospitals that don’t use all their vaccines, New York Governor Andrew Cuomo also signed an executive order requiring providers to certify that every recipient qualifies under the current rationing protocol. Letting someone jump the queue now risks a $1 million fine and the loss of a state license. “If you wanted to make sure that rapidly expiring vaccines distributed in 10-dose vials end up in the trash, this is how you'd do it,” observed commentator Mason Hartman on Twitter.

Micromanagement is impeding the rollout. In South Carolina, for instance, a medical assistant often gives injections in a doctor’s office, and the job requires no special certification. For Covid-19 vaccines, however, the state says that even someone with decades of experience can’t administer a shot unless they have an official credential.

In Europe it is common to give yourself a shot, without a covid-risky trip to a doctor's office, stating in line, filling out paperwork and taking up half an hour of time of a highly trained person. I found this out one day many years ago, with a case of bronchitis. The pharmacist just handed me the shot and said go home, take this. It's not hard, folks. Amazon could just send it to you. Imperfect? Yes. Will there be problems? Yes. Will we beat evolution and exponential growth and save tens of thousands? Yes. 

Distribution is hard enough without these roadblocks. Start with the numbers. At Kaiser Permanente facilities, a single vaccinator can give about 10 shots an hour, with much of the time spent filling out forms. To get to herd immunity, the U.S. needs to inject two doses several weeks apart to something like 240 million people. At 10 injections an hour, that’s 48 million hours of vaccinators’ time, 4.8 million hours a week over 10 weeks to get to early March. We’d need 120,000 vaccinators working 40-hour weeks. ...

10 shots an hour is very efficient. Other numbers I've heard go up to .5-1 person-hours per shot, including the paperwork and monitoring afterwards. The 1 hour per shot comes from nursing homes. 

This is basic math. We are supposed to revere the "scientists" at FDA and CDC. Did anyone add up 240 million people (more likely 300, per Dr. Fauci), estimate person-hours per shot with all the restrictions, and figure out when the shots will be given? 

The last thing we need in these circumstances are special restrictions on who can administer vaccines — restrictions that send the perverse message that vaccines against this disease are somehow more questionable than those against the flu or measles.

A correspondent reports:

I actually got the Moderna vaccine first dose yesterday. In a lot of US, including where I live, only certain healthcare workers are eligible for appointments so far. The limited size of the cohort and the fact a lot of healthcare workers don’t want it after actually having the virus, means appointments are being missed and vaccines wasted. I simply queued for an hour in a standby line last night at my local Safeway pharmacy. They had 8 vaccines reaching expiry if unused, and I was no. 6 in queue. I would have given up my slot for an elderly couple behind me, but there was enough for them too and those that missed out were all early 20s. Mad, Soviet-like system (queuing literally in the frozen meat section of the store for an hour or so), limited eligibility and queueing and I felt kinda dirty getting it, but I’ll be damned if I’m going to pass it up when it will be otherwise wasted.

Note this is a state where Safeway pharmacies are allowed to give it, and Gov. Cuomo will not fine then $10 million for letting my friend skip the line 

In the U.K. I’m hearing even more insane stories. Only paying general practitioners cost price on delivery of the second dose, so some not bothering to offer it. And paying doctors nearly half the hourly rate they could obtain doing locum work. Penny wise and pound foolish. Literally: the U.K. currently has an incredibly expensive furlough scheme. 

I am often asked, "what can the government do to solve x problem?" My stock answer is, get out of the way. It's not always right, but it's often a good start. 

Update: From Jennifer Smith reporting in WSJ

Instead of seeking to halt the spread of transmission in communities, said Dr. Swann, who advised the CDC during the H1N1 pandemic, the focus has been on reducing mortality, especially among high-risk populations.

Reducing short-run mortality. When you halt the "spread of transmission" you reduce long-run mortality a lot more. Get on top of exponential growth, please. A lovely encapsulation of the mentality. Grumpy just spilled his coffee again. 

 

15 comments:

  1. Excellent point about state governments getting in the way. It's bad enough when a drug or vaccine has to jump through hoops created by the federal government, but now we've added more hoops from state and local governments. That's not a recipe for disaster, it's a guarantee.

    ReplyDelete
  2. It is inspiring to have a proponent of libertarianism arguing for pragmatism (aka, in USA, socialism). Are we seeing, in the (dis)organisation of administration of vaccine, just how the US government has been commandeered into performing corporate protectionism? Pfizer is a US headquartered multi-national corporation. Moderna is based in Cambridge (Massachusetts). AstraZeneca is based in the other Cambridge.
    Market distorting intervention on behalf of the US corporations?
    --E5

    ReplyDelete
  3. "The US had given AstraZeneca, which had little experience developing vaccines, $1.2 billion to secure 300 million doses of the drugmaker's vaccine back in May."--Business Insider [https://www.businessinsider.com/astrazeneca-slow-to-give-fda-regulators-trial-data-halting-process-2020-12]
    "Finally a concrete answer: almost certainly not. Astra’s base case assumption, according to its chief executive, Pascal Soriot, is that the US FDA will require US data, 'but of course it’s for them to decide'. Oxford University’s Sarah Gilbert added that because readout depended on events the researchers do not know when the US trial data might become available.
    "Moreover, aside from the FDA’s unwillingness to approve products with no US patient data, the standard dose regimen is not good enough. In the UK trial the efficacy figure in those receiving two standard doses was 60.3%, but the lower bound of the confidence interval was 28%, falling beneath the FDA’s criterion of 30%."--Evaluate [https://www.evaluate.com/vantage/articles/news/trial-results/astrazenecas-vaccine-looks-increasingly-also-ran]
    The situation with the AstraZeneca-Oxford vaccine is not straight-forward. The opinion piece by Marty Makary, MD, is a jumble. The three paragraphs quoted from Makary's article include unsubstantiated accusations. Alex Tabarrok: "Thousands of people are dying every day. We have a vaccine factory ready to go. The FDA should lifts [sic] its ban on the AstraZeneca vaccine." Tabarrok’s view comes despite indications that AstraZeneca has not yet provided the FDA with information/data supporting an application for emergency use authorization. Tabarrok's conclusion is simply not credible.
    Here's the counter-argument: Absent a credible regimen of vaccine trials, an expedited approval of a vaccine for general application but of doubtful efficacy and potential adverse reactions is not going to be accepted by the general public and healthcare professionals in particular: "The big news here is that investigators have finally admitted a case of transverse myelitis as being ‘possibly related to vaccination’. This occurred 14 days after the booster shot, and an independent neurological committee considered the most likely diagnosis to be of an idiopathic, short segment, spinal cord demyelination."–Evaluate. Telling the FDA to get out of the way is not the answer.
    Virgina Posteral's piece in Bloomberg News online is poorly informed. Contrary to her view, workarounds are being put in place locally to move the vaccines that have been given EUA to the locations where those vaccines are needed. Vaccines are shipped directly from the federal government to locations that are authorized to receive the vaccines and from there to locations where the vaccines are to be administered. If there are more vaccine doses in one location than there are individuals requiring vaccination, the surplus doses are sent to locations that lack sufficient doses to meet local demand. These arrangements are informal but fit within the protocol established for administering the vaccinations. This is the sort of adaptability that we should expect of a robust system.
    We are in for a lengthy period of living with this virus. The duration of immunization is not completely established. The virus has demonstrated the potential for mutation into more virulent forms. All age groups are susceptible albeit with differing degrees of disease severity. As transmissibility increases the virus will spread more rapidly through the population, and the mortality rate per unit time will rise with the rise in the case rate. In order to keep up, it will be necessary to build more hospitals and more intensive-care units in hospitals, and accelerate the training and recruitment of more healthcare personnel. Average life expectancy will decline. Triage will become more prevalent as the case load rises. Some very hard decisions will have to be made.

    ReplyDelete
    Replies
    1. Great well documented comment. Thanks.

      Delete
    2. Where has covid "demonstrated the potential for mutation into more virulent forms"? The new UK and SA forms certainly seem to be more contagious, but every health expert I've seen has been at pains to emphasize there's no evidence it's any more deadly.

      As for "If there are more vaccine doses in one location than there are individuals requiring vaccination, the surplus doses are sent to locations that lack sufficient doses to meet local demand.", this describes a government of far, far greater efficiency than any I've ever come across.

      Delete
  4. If you think that's nuts, go read up on the vaccination procedure in France. Summary: make doctor appointment, listen to doctor give a long explanation about how the vaccine works, go home for a mandatory 3-4 day reflection period, before returning to sign the consent form for the actual shot.

    ReplyDelete
  5. Efficient vaccination could hit 100 shots per hour for the vaccinator if all they did was give shots. Just have 10 other (not medically skilled and cheap) filling out paperwork or have each person fill it out and have it be reviewed. Or make a smartphone app used by the person who wants to get vaccinated and the can transfer the data to a server and then show a QR code to a device that the vaccinator has.

    Between physician, RN, DDS, phlebotomist job categories that is about 4 million people in USA. With all these people we could vaccinate the entire population in one day.

    ReplyDelete
    Replies
    1. To get well above 4 million potential vaccinators also include pharmacists, in some states pharmacist assistants, veterinarians, and physician assistants.

      People who can jab people with needles in America number in the millions and, given a supporting org of sufficient size,could inject tens or hundreds of millions of people per day. Granted creating a management structure to oversee this would be a challenge.

      We can and ought to scale up our vaccination capacity to whatever the vaccine delivery rate will be 1, 2, 3 months from now.

      Delete
  6. Also interesting:

    Open Orphan PLC (LON:ORPH) said the first volunteer has been dosed with a new needle-free, nasally-administered coronavirus (COVID-19) vaccine.

    The phase I trial of the next-generation single-dose inoculation will take place at the company’s hVIVO’s facility, a specialist centre for human challenge studies.

    The advance has been developed by a US company called Codagenix and will be subject to a randomised, double-blind, placebo-controlled dose-escalation assessment.

    This will evaluate the safety and tolerability of the single-dose intranasal vaccine candidate in 48 healthy young adult volunteers at hVIVO's state-of-the-art centre in Whitechapel, London.

    ---30---

    The laborious phase 1, 2, and 3 outlook...sure, good enough for most situations...but should there not be a hurry-up plan for pandemics?

    ReplyDelete
  7. John, not so fast. As you yourself point out

    "But it turns out vaccine supply is not even the short run constraint"

    Ok we have 2 vaccines that are approved and we have vaccines piling up in warehouses that can not get into peoples' arms.

    Lets not jump the gun and jump up and down every time the FDA wants to slow down and be cautious. We have 2 vaccines that work and are in good supply so there is plenty of reason to take care and make sure that any new vaccines are safe and effective.

    ReplyDelete
    Replies
    1. Are they safe and effective in the UK but not in the USA?

      That's weird. Maybe the virus asks first the nationality of the body it is about to infect, but I doubt it.

      Delete
  8. It's like the FDA fell out of a plane, but they want 5 studies back on the efficacy of parachute deployments before they'll yank on it

    ReplyDelete
  9. To the points of Postrel and the correspondent about slow distribution, as well as John's point about short-run mortality:

    About 81% of U.S. reported COVID deaths are among the ~16% of the population who are over age 65 ( https://www.cdc.gov/nchs/nvss/vsrr/covid_weekly/index.htm#AgeAndSex ). That group constitutes just over 50 million people. They are also more likely to be hospitalized than younger people,though that ratio of increased risk is considerably lower than the increased fatality risk ( https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/older-adults.html ).

    Importantly for a mass vaccination program, that's a group that can be readily identified by the simple metric of age. It's a criterion that's easily communicated to the public, as well as any change to it over time (e.g., at some point stating that now everyone over 55 is eligible to be vaccinated). Based on the experience in states (e.g., Florida) where everyone over 65 is currently eligible for vaccination, the demand from this age group more than meets current levels of vaccine supply. That's hardly a surprise based on the size and risk profile of the group. That's even more the case if frontline healthcare workers are also eligible.

    The pragmatic solution here seems to be to distribute the vaccine widely and stick to simple criteria of age. (And, apparently, healthcare workers, who have already been put at or near front of the line in every state.)

    My prediction: for all that things have been rocky so far, we will - much like with PCR testing - see a steady and fairly impressive increase in the daily numbers of people being vaccinated.

    ReplyDelete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.