Wednesday, November 11, 2020

Virus over? Not quite

The news of a vaccine seems to be sparking an its-all-over sigh of relief. Not so fast.  Interesting and challenging corona virus policy remains on the front burner. 

Holman Jenkins makes a few good points in WSJ. The media and many governments (mine) are focused on new case counts, now 100,000 per day. But 

Brown University’s Dr. Ashish Jha estimated while we are identifying 100,000 new cases a day, “we’re probably missing 70%, 80% of all the cases out there.”

He mentions other guesses that say 90%. 

Why does this matter? Well, 100,000 cases per day x (say) 2 weeks of infectiousness means that 1.4 million people are infectious, or 0.5% of the population. Not bad odds for a dinner party, maybe not a rave. But if we really have 500,000 or 1,000,000 cases per day, that means 2.5% to 5% of the people you are going to run in to may be infectious. Yikes.  

If Americans knew they were being laughably misled, that the virus is far more widespread and their chances of encountering it are much greater than the confirmed case count (currently 10 million) implies, their behavior might be different. Especially we might get more mask-wearing by unwitting carriers to curb unwitting spread.

And a lot less partying. 

More intriguing, 

 a vaccine works by inducing antibodies. In the first year, Pfizer might have enough to inoculate 25 million people at a time when, at current rates, 100 million of us may have antibodies. 

This is a really interesting fact. He doesn't quite say it, but 100 million is darn close to herd immunity -- depending on how careful people are. (Herd immunity comes when the reproduction rate is below one. Reproduction rate = number of people an infected person contacts x chance the contact is immune. Lower contact rate and higher immunity rate work hand in had to determine when the disease stops.) It also says the marginal effect of a vaccine is not that high in the first year. 

On the other hand -- the other big unknown -- just how long does immunity last? There are reports of people getting it twice. This would seem like The Big Question. If immunity lasts a year, we will be on our way to herd immunity and the virus will be over before Pfizer has a chance to sell a vaccine. If it does not, we will really really need the vaccine. 

Everyone seems to see the vaccine as the magic cure. Wait a minute. It has to be distributed effectively. And our governments (Federal, state, local, all countries) have not been models of efficiency, for example on testing and contact tracing. 

So having an accurate picture of prevalence will be necessary to make sure we aren’t wasting vaccine (never mind the already politicized question of who should get early doses: mobile young people to curtail the spread or older people and minorities because this is what the media clamors for).

Joe Biden was quoted as saying the vaccine should be 

 “distributed equitably, and efficiently, and free for every American.”

Notice equitably comes first. Of course politicians or those with access will come absolutely first! But the hard fact is, give it to one millennial bar hopper and save the 25 people he would give it to. 

Free is interesting. If the government wants to pay for it, that's fine -- any imaginable cost is way less than the "stimulus" blowouts. I worry that means price controls. Let's be clear -- companies that come up with a vaccine should make billions of dollars in good old-fashioned evil capitalist profit. They have made a miracle for us, and we want them racing for the same miracle next time. (I also would like to ask Mr Biden, just what should not be free? Name something substantial that Americans should be expected to work hard and pay for, and not everybody gets it. Vaccine? No. Health care? No. Housing? No. Toilet paper? )

Mr. Biden now calls for a national testing board because it sounds take-chargey but let’s understand: Our current hunt for newly active infections is useful mainly for letting doctors know what disease they are treating when confronted with a severe case. It does nothing to control the spread. The results come too slowly and miss the overwhelming number of cases.

Dead on track with the way testing works now, and the conceptual breakthrough still not happening -- testing for public health is a totally different concept than testing for disease diagnosis and treatment. 

Don’t get me wrong: If a cheap, daily, at-home test were available, it would give us real leverage over the spread. We aren’t there yet and won’t be before a vaccine and natural effects have quashed the current destructive upsurge.

But we are! See recent Michael Mina thread on paper strip antigen tests. Once again, this represents the best opportunity we have, in my view, for reducing the spread of disease, now, at least before vaccine is so widely available that everyone can have it -- and that means everyone, as people travel. 

Meanwhile, the neglected antibody testing will become more important so recipients of the new vaccines, which are likely to be only partly effective, know they are actually generating immunity and can resume normal life. 

This is another important thought. The vaccine will not be 100% effective. Many people who test positive now refuse to quarantine. Many people may say, I have the vaccine I'm good to go. Always think of the behavioral response! How long does vaccine granted immunity last? 

In all the money and data being spread around, I wish we knew how many people in a completely random sample are infectious right now, and how many show antibodies. 

20 comments:

  1. Tremendous post. How are we not shouting for rapid testing? Perhaps, people are too busy slinging memes.

    Rapidtests.org has a template letter you can send your congressperson (doesn’t that even work?). Better than nothing I suppose.

    ReplyDelete
  2. Northwestern is running somewhat random antibody testing for those who live in Cook County (somewhat because you have request the test so there is self selection). I took mine last week (live in Hyde Park) but have yet to get the results. It is all by mail so low cost for everyone.

    https://scan.northwestern.edu

    ReplyDelete
  3. Behavior modification, as I've mentioned before, is the other side of the coin in defeating the virus. I still see people not wearing masks. Where are the incentives in not wearing one? It makes no sense to me.

    Preferences still appear to be severely screwed up. People want the economy to function again like before. but are unwilling to participate in its defeat. Either they don't care or expect someone else to fix the problem.

    While testing gets our arms around how bad it really is, it seems, like Dr. Cochrane has pointed to before, a vaccine is a miracle tech breakthrough that people are desperately waiting for, but all the while unwilling to behave properly. It's irresponsible at best to behave and think this way. No wonder we're getting ravaged.

    ReplyDelete
    Replies
    1. Unfortunately culture wars have swallowed everything, even masks. In some places wearing or not wearing a mask is highly visible signal of your red/blue tribe affiliation.

      Delete
    2. Mykel - you are assigning too much weight to the role of human behavior in thinking that places can reliably "control" an easily spread respiratory virus, particularly once it has reached a certain level of prevalence.

      Also, there's perfectly coherent set of preferences that can lead to some people doing not all that much about COVID risk. One way to frame it is this: the annual increase to all-cause mortality from COVID is likely on the order of 10% to 20% for 1 or 2 years. Hopefully the ultimate amount of that excess is lowered by a vaccine. Looking at U.S. age-adjusted mortality data from the CDC, that increase takes the age-adjusted mortality rate back up to where it was roughly somewhere between 2000 and 2006. https://www.cdc.gov/nchs/data-visualization/mortality-trends/index.htm

      Now, I don't personally go quite as far that argument might warrant, but it's framing that we hear very little about. Given the risk profile, it would be perfectly rational for a healthy person under some age (perhaps 25 to 35) to carry on life with limited disruption while taking some care to minimize close contact with people in much higher risk groups and being willing to limit community contact if symptomatic.

      It would clearly be a better situation if this disease didn't exist - it is in fact causing excess mortality. It's not useful, however, to hear arguments that oversimplify causality to nothing more than people being'"good' or "bad", ignoring the important role of factors such as seasonality and build up of immunity in a community. It's particularly galling when that line of thinking is used as the rationale for government mandates - such as business closures - that (1) don't appear to be rigorously scrutinized under cost-benefit analysis, (2) generally have limited evidence of effectiveness, (3) target endpoints that may not be realistic, (4) are massively disruptive to society, and (5) often posit a false dichotomy of "safe" vs. "not safe" rather than a spectrum of risks that individuals might choose to take or avoid based on personal demographics and risk tolerance.

      Delete
    3. Well, it's also a signal of clear, independent thinking. Multi-ply fabric masks are effective at stopping viral particles adhering to moisture. They are utterly ineffective in stopping those that aren't. Other sorts of masks (such as the laughable faux-surgical ones)? Useless in any milieu.

      Viral particles are on the order of 100 nm. You need at least a well-sealed N95 to stop them effectively. Even multi-ply fabric masks worn inside have only a marginal effect.

      I'm amenable to wearing a mask indoors simply on a go-along-to-get-along basis. Outside: never. Masks do nothing outside except impair breathing, reduce O2 saturation, and trap bacteria. Anyone wearing one in the ambient is almost by definition thoughtless. One would have to work very hard outdoors, over a long period in close proximity to the infected, ever to achieve a dangerous viral load.

      This is a fascinating and wholly depressing social panic we're living through.

      Delete
  4. "In all the money and data being spread around, I wish we knew how many people in a completely random sample are infectious right now, and how many show antibodies."
    Many have raised this point often since at least March of this year.
    Perhaps someone can explain why mandatory random tests are an affront to individual liberty, while lock downs, police entering homes to ensure a small gathering, forcing businesses to close, and preventing people from using their labor are not.

    ReplyDelete
  5. when you say 100 million is close to herd immunity you will need to specify what is meant by "close". In some sense it is not. In a homogeneous system herd immunity threshold for covid (Ro~2.5) is about 60-70%. But that is only roughly half way through the epidemic. For the full epidemic it is closer to 90% total infected. So there are 200 million to go. Also, heterogeneity plays a huge role. Right now most infections are occurring in the midwest...etc where there wasn't a lot of spread in the last few months.

    ReplyDelete
  6. "Many people who test positive now refuse to quarantine". If we could have decentralized ubiquitous testing this would not even be a big problem. You refuse to quarantine but you are not allowed in the office, in the school, in the theater, in the supermarket, or not allowed to check-in for your flight... I guess these people could just wander aimlessly in the streets.
    Again, businesses are already "testing" everyone with whatever tool they have available and are allowed to use. This translates typically in a useless $5 thermometer.

    ReplyDelete
  7. I'm kind of curious if Pfizer or any of the other big pharma companies have a deal in place already with the government. After all it certainly hurts your incentives to pour a bunch of money into developing a vaccine only for the government to shortchange you on the price and then tarnish your reputation in the media for being heartless (something the pharma companies are already dealing with and will likely be aware of)

    ReplyDelete
  8. I'm kind of curious if Pfizer or any of the other big pharma companies have a deal in place already with the government. After all it certainly hurts your incentives to pour a bunch of money into developing a vaccine only for the government to shortchange you on the price and then tarnish your reputation in the media for being heartless (something the pharma companies are already dealing with and will likely be aware of)

    ReplyDelete
  9. You might consider the following short news report found in the "U.S. WATCH" secton of today's Wall Street Journal (WSJ: 12 Nov 2020)...
    https://www.wsj.com/articles/asymptomatic-covid-19-cases-show-need-for-wider-surveillance-testing-study-suggests-11605132000?mod=searchresults_pos1&page=1 [Attn.: paywall limits access to subscribers]

    "U.S. WATCH
    "SOUTH CAROLINA
    "Study Shows Need for Surveillance Testing

    "A study of Covid-19 testing among nearly 2,000 young adults found symptom monitoring missed nearly all cases of infection, suggesting regular, widespread surveillance testing is
    needed for both asymptomatic and symptomatic people to get the pandemic under control.

    "The research, published in the New England Journal of Medicine on Wednesday, looked at 1,848 Marine Corps recruits between the ages of 18 and 31, who had been required to quarantine at home before arriving at the Citadel military college in Charleston, S.C., where they underwent a second 14-day on-campus supervised quarantine.

    "At Citadel, they were monitored every day for symptoms and were scheduled to be tested three times. By the 14th day, 51 of the study participants had tested positive for the coronavirus.

    "What was surprising to researchers: All 51 cases were picked up by the prescheduled tests. None were detected from additional tests given to those who reported symptoms.

    "Despite the stringent measures and initial testing, the virus still slipped through the cracks and spread among recruits, the researchers said.

    —Sarah Toy ((c) 2020 The Wall Street Journal) [Fair Use Doctrine]

    51 out of 1,848 recruits tested positive for coronavirus after 14 days, but had no associated symptoms. 51/1848 = 0.0275974..., or 2.76% (rounded up) of the cohort (ages 18-31 yrs) were asymptomatic coronavirus carriers.

    The ensemble average for this cohort of 18- to 31-year olds fits within the broad estimate found in your blog article ("...that means 2.5% to 5% of the people you are going to run in to may be infectious.")

    What it means is that, short of a PRC-type top-down direction to seize and test every person in the USA for testing and quarantine, the pandemic will not be over for some time to come, perhaps for as long as it takes to innoculate a very large proportion of the U.S. population and every foreign traveller arriving at a U.S. Port of Entry.

    It also means that foreign countries will continue to limit entry by U.S. persons to those countries. It may, furthermore, mean that so-called "essential workers" travelling between the U.S. and a foreign country will be restricted in ways that they have not so far been restricted.

    ReplyDelete
  10. I wonder if your last sentence is not the most important.
    "In all the money and data being spread around, I wish we knew how many people in a completely random sample are infectious right now, and how many show antibodies."

    Surely we can construct a properly random sample of the population to gauge what seems to be a key piece of data for taking actions?

    ReplyDelete
  11. My comment disappeared. Maybe because I included a YouTube link? Look for MIT coronavirus lectures, lecture 8.
    It’s excellent on antibody v PCR testing, rapid less sensitive tests v more expensive more accurate PCR tests. And the CDC mandating low throughput outdated PCR test equipment..

    ReplyDelete
  12. I went to a dinner party. There were four people there. Three out of four tested positive for Covid. I tested negative. Either, I had Covid and didn't know it, wasn't exposed enough, or there was a faulty test. I have not had an antibody test. BTW, all the people "survived" and it was simply a bad cold. All of us are close to 60. Typing that last sentence hurts more than Covid.

    ReplyDelete
  13. "On the other hand -- the other big unknown -- just how long does immunity last? There are reports of people getting it twice. This would seem like The Big Question. If immunity lasts a year, we will be on our way to herd immunity and the virus will be over before Pfizer has a chance to sell a vaccine. If it does not, we will really really need the vaccine."

    This is quite important, and made me wonder what the odds are that this virus ends up mutating into multiple different strains with the vaccine only effective against one, similar to the flu where we have to guess which strain will be most prevalent this season. But also if immunity only lasts a year, even if we approach herd immunity, that could fade away, and so the vaccine might still be useful further down the road. (If the virus isn't completely obliterated)

    ReplyDelete
  14. Or it could be that the SARS COV-2 virus is already endemic, you won't find many people antibodies as the T-cells take care of before they need to be generated, and testing thousands of healthy or already infected people will increase the false positives due to how the rt-PCR process works and how many amplification cycles are run before a positive/negative verdict is given.

    ReplyDelete
  15. At the risk of making you more Grumpy, I believe your formula for "Reproduction rate" is incorrect. Shouldn't the second factor of the formula be the COMPLEMENT of the chance the contact is immune. Let me add that your posts, and things like GoodFellows at Hoover, are among my best sources of thoughtful comment and analysis. I don't know how (/if) you find time to sleep

    ReplyDelete
  16. Being vaccine damaged and having two sons who are also vaccine damaged, one at 6 years and one while in military, our research has shown the danger of vaccines. There are so many factors, but when one in 50 children now have autism, you may want to do your own research. These vaccines for covid are well tested and the quoted 90% efficacy is misleading. Not much other info so that you have no informed consent.

    ReplyDelete
  17. John, here is an article about the new vaccines. https://www.sott.net/article/444442-Recipients-being-misled-to-a-criminal-extent-Whats-not-being-said-about-Pfizers-new-Coronavirus-vaccine

    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.