Wednesday, September 2, 2020

Abbott Labs to the rescue? Free the tests!

Context: Cheap, fast, tests can stop this pandemic quickly, even if they are not very accurate.

Last week, Abbott Labs announced (more info here) that (finally)
the U.S. Food and Drug Administration (FDA) has issued Emergency Use Authorization (EUA) for its BinaxNOW™ COVID-19 Ag Card rapid test for detection of COVID-19 infection. Abbott will sell this test for $5. It is highly portable (about the size of a credit card), affordable and provides results in 15 minutes. BinaxNOW uses proven Abbott lateral flow technology, making it a reliable and familiar format for frequent mass testing through their healthcare provider. With no equipment required, the device will be an important tool to manage risk by quickly identifying infectious people so they don't spread the disease to others.  
Note the last sentence. Abbott gets it -- the point of this test is not to diagnose sick people, it is to keep most sick people from spreading the disease.  If every American got this test once a week for a month ($5 x $350 million = $7 billion = one drop in bucket of the fiscal and economic cost of this pandemic) it would be over in a month.
Abbott will also launch a complementary mobile app for iPhone and Android devices named NAVICA™. This first-of-its-kind app, available at no charge, will allow people who test negative to display a temporary digital health pass that is renewed each time a person is tested through their healthcare provider together with the date of the test result. Organizations will be able to view and verify the information on a mobile device to facilitate entry into facilities along with hand-washing, social distancing, enhanced cleaning and mask-wearing.
"We intentionally designed the BinaxNOW test and NAVICA app so we could offer a comprehensive testing solution to help Americans feel more confident about their health and lives," said Robert B. Ford, president and chief executive officer...
Even better. (It does not say who the app reports data to, which could make it better yet. For example, it could automatically notify your employer.)
"With lab-based tests, you get excellent sensitivity but might have to wait days or longer to get the results. With a rapid antigen test, you get a result right away, getting infectious people off the streets and into quarantine so they don't spread the virus."
Again, Abbott gets it.

But, what's this business about "through their healthcare provider?"
Under FDA EUA, the BinaxNOW COVID-19 Ag Card is for use by healthcare professionals and can be used in point-of-care settings that are qualified to have the test performed and are operating under a CLIA (Clinical Laboratory Improvement Amendments) Certificate of Waiver, Certificate of Compliance, or Certificate of Accreditation. Within these settings, the test can be performed by doctors, nurses, school nurses, medical assistants and technicians, pharmacists, employer occupational health specialists, and more with minimal training and a patient prescription.   
What is wrong with these people? (FDA) If most of us call our health care providers, you get  non-urgent appointment in about 3 weeks, insurance gets billed about $400, we pay $150, to get the necessary referral, and prescription (!) and on for more delays and costs to get the test.

What possible reason is there for all these restrictions? How can anybody be hurt by taking this test, and how will all these layers of bureaucracy help that anyway? Yes, a huge employer like Stanford can probably obtain a CLIA CoW, CoC, etc. and hire a "occupational health specialist" to administer tests, but how is a restaurant going to do it?

There is a pandemic on, folks. Regulators gotta regulate, I guess, to justify their existence. But not now.

Free the tests!


  1. I agree with you at a macro level about testing -- and have agreed with all of your prior articles on the subject up to this point -- but am not sure this specific Abbott test was the greatest of examples. It's NOT super easy to use, and the FDA trials included having both trained and untrained individuals administer the tests (i.e., it's not like they just defaulted to saying "no").

    The thing is, this specific Abbott test wasn't really designed to be an "at home" solution in the first place. The hope is rather that it can be a cheap/fast tool that pharmacy techs, school nurses and mobile health screening companies can deploy. The FDA has done what's needed to ensure those use cases can happen (the admin roadblocks might sound bad, but are not the least bit onerous for the intended user groups).

    So yes, overall government agencies need to get out-of-the-way of rapid testing solutions. But this one specific test seems to have been approved for precisely the use case it was designed to satisfy. And I'd not underestimate the potential value... if every kid attending a school big enough to have a nurse, everyone working at a company that pays for mobile employee health screenings, and everyone with a spare 15 minutes at a CVS gets this test, it'd a be a huge improvement over where we're at today.

  2. Mulcahy's Law: it is no fun to be a regulator unless you have something/someone to regulate.

  3. Not perfect but but better than the alternative...Karl Marx always urged us to ask Qui Bono?... When evaluating people's motivations, I'm at a loss who they think possibly benefits from such procrastination. Listening to Dr. Mina, it must be the case that the FDA in this case as many, simply needs a remedial course in basic statistics?

  4. Bureaucrats panacea; If it's not broken, fix it until it is!

  5. On to vaccines?

  6. Testing every American once per week for one month at $5 per test, if it were possible to do so, would yield an uptick in revenue of $28 billion for Abbott Laboratories, i.e., assuming that "Americans" number 350 million and that sufficient infrastructure exists to support that rate of testing and that those tests can be administered to every "American" every week for a "month" comprising four successive weeks..., etc. The $5 cost per test would not cover the marginal cost nor the average cost of testing. But say it did, and you were able to compel on average 50 million "Americans" per day for one month to be tested, a major mobilization effort involving state coercion on a scale that is seldom undertaken except under war-time conditions, what is the likelihood that it would "end" the pandemic in a "month"? Answer: Highly unlikely. Firstly, infectees identified by such testing would have to be removed from the general population (quarantined or placed into isolation) which would require a certain amount of coercion in itself to achieve. Secondly, the term "American" requires qualification--does it mean American citizen, or an American citizen with health insurance, or every person regardless of citizenship or nationality residing in the U.S.? Thirdly, how would compulsion of, on average, 50 million persons per day, and a different 50 million each day, be effected given that law enforcement agencies are all ready stretched to the limit in the enforcement of laws already on the books in each state and federally (to avoid doubt, you are talking about universal compulsion, by the way without exception) and those agencies are unlikely to be sufficient to the task envisioned? Fourthly, the numbers compelled to isolate or be placed into quarantine will require monitoring and if a certain fraction break out isolation or quarantine then those will need to be tracked down and returned to isolation or quarantine, requiring an additional level of enforcement effort, etc. And so on.

    Basically, you talking about replicating the PRC's program that was implemented in Wuhan, China, to suppress the disease there. Think about. It takes a totalitarian government to effect such an effort. Do you really think Americans would roll over without batting an eye or raising an objection? Really?

    1. I think you might be getting too caught up in this idea of the police compelling people to get tested. I don't think it would require anything of the sort. If my current employer offered these tests as a condition for getting back to on-site work, I would be first in line on Tuesday morning. The point being that state and country health department could require these tests as pre-requisites for full capacity building access and decentralize the adminstration of the tests.

  7. Abbott Laboratories production schedule was announced alongside the announcement of the EUA covering the BinaxNOW™ COVID-19 Ag Card rapid test for COVID-19.

    Production will commence this month and tens of millions of units will be produced in September; production will continue to ramp up to a level of 50 million per month at which level production will be maintained.

    We should expect to see somewhere between 30 to 40 million units produced in the month of September, and then an additional 50 million units every month thereafter. This level of production is substantially below the level of production required to test 350 million Americans each and every week for one month. At that rate of testing, Abbott Labs would have to build an inventory of 4 x 350 x 10^6 units and distribute that inventory at a rate of 350 x 10^6 per week for four weeks. The build time for the first 350 million test units at a production rate of 50 million per month is 7 months. It will take 24 months at a rate of 50 x 10^6 units per month to reach an inventory on hand level to support at testing rate of 350 x 10^6 persons per week for four weeks. The inventory at the commencement of testing would have to be 1,200 x 10^6 units (1.2 billion units). In the meantime, no testing is performed for the first 24 months. Then in the 25th month, the entire U.S. population is tested once each week for four weeks, after which time the test rate drops to the production rate of 50 million test units per month. The proposed testing scheme is impractical--who would want to wait 2 years before commencing use of Abbott Laboratories' BinaxNOW COVID-19 Ag Card? No one. The purpose of the EUA is to get the test into the health care system rapidly and now, not two years from now.

    How could the 350 million tests per week for four weeks be effected if it is urgently needed now? Build 28 Abbott Laboratories' production lines, each at a rate of 50 million units per month (12.5 x 10^6 units per week). This gives you a production rate of 350 x 10^6 units per week. With a one week lead-time to build the first 350 million units (two weeks if distribution lag times are accounted for), you can start the testing scheme. But by the end of the 4th week, production is shut-down except for 1 production line out of the 28 production lines. The construction lead time for 27 facilities has not been factored in, as the scope and scale of the production facility needed is not known. Would the price of $5 per test unit still be the price charged by Abbott? Almost surely not. Conclusion: the testing schema proposed in the blog article is impracticable, almost surely.

  8. How can anybody be hurt by taking this test? False positives. If 4% of tests are false positives, each year 88% of the population would be put out of work for a fortnight. That's 6.6 million man-years of work sacrificed each year.


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.