Correct. If the ambient infection rate is “high” (Phase One) then it is beyond stupid to consider opening schools. Only when the ambient infection rate is “low” (Phase Two) is it time to consider opening schools.
So “studies” purporting to show how safe schools can be made with stringent precautions and mitigation factors only become germane when we reach the second phase. And I like most people hope that we can re-open schools (when the time comes) with a minimum of precautions/mitigation.
That is often missing. Any gathering multiplies spread. That multiplier is small when cases are small and large when cases are large. It’s not linear.
The original idea of having metrics and sticking to them has been completely abandoned in most places.
Steven Millman Update:
January 23rd COVID Update: FINALLY some good news.
Short version: The increase in death rates will peak around Groundhog’s Day and then drop precipitously for several weeks. There are too many unknowns to look out farther.
tl;dr as ever. What’s that strange feeling I can’t quite place. Like some distant memory. Is it hope?
Federal government officials are now telling us that things are going to get worse before they get better and for a change they’re right. The average weekly death rate has been rising since early October when about 5,000 people were dying per week to the current state where just under 22,000 died in the just last seven days. The model projections suggest these numbers will continue to rise until the first week of February. The good news is that at that point there is very strong evidence that death rate will not only start to fall, but that they will fall rapidly. What we’re seeing right now at the national level is a substantial decline of both new cases and test positivity without a concurrent increase in testing. In combination these indicate a decline in community spread of the disease. That’s REALLY GOOD news. What’s not good news is that if model performance continues to be as accurate a it’s been thus far, January will be by far the deadliest month of the pandemic. The model estimate for January is 96,290 deaths from COVID in the US compared to 82,000 in December, 41,000 in November, and 23,500 in October. The model predicts that 72,000 will die in February, registering the first decline in death totals since mid-September.
We’re not out of it yet, of course. The model suggests that between now and the end of February, another 100,000 Americans will die from COVID. More Americans have died from COIVD so far than the entire population of Tampa, Florida. By the end of February, more Americans will have died than the population of Atlanta, Georgia.
I have only projected out to the end of February because there are far too many unknown variables about what will happen as well as a lack of clarity as to the key drivers of the decline in new cases we’re seeing now. There are a number of things happening that could continue to drive the number down. First, let’s consider the number of people who are currently at least somewhat resistant to the virus. There are about 25 million Americans who have had confirmed cases of COVID-19. Estimates of total community spread have ranged in the literature quite a bit, although most range between five and eight times the number of those confirmed by testing. If true, that means that between 125 and 200 million Americans have already been infected at this point. We know that reinfection is possible, that resistance from prior infection is of unclear duration, and that there is always the chance that a strain will mutate in a way that’s resistant to existing antibodies. That said, the total number of available hosts for COVID is probably only about half the population right now. This is nowhere near the level necessary for herd immunity, but it does greatly reduce the available source of people for the virus to infect – at least for a time. Add to those numbers the fact that about 16 million vaccines have been delivered, although only three million people have yet had two doses for maximum disease resistance. Vaccine distribution is slower than we’d prefer, but still progressing and focused principally on those at highest risk of mortality. Taken together, this suggests that there will be smaller and smaller numbers of available hosts, and that those who become infected will be at lower risk of mortality. Case fatality rate has held constant at about 1.4-1.5% for the last two months, and we would hope that number will start to fall in the near future as the size of the most vulnerable populations gets smaller . Finally, the change in administration messaging on the pandemic and what should be done about it (wear a mask, take the vaccine) could create a change in the acceptance of taking the simple steps necessary to reduce disease spread. The President has, for example, signed new executive orders requiring all persons in federal buildings to wear masks, and to require all persons traveling by planes, trains, etc. wear masks. These new rules, if followed, should have a substantial impact on disease mitigation. There are a variety of other disease mitigation strategies the President hopes to undertake, but it’s too early to know which of these will be funded and how effectively they would be implemented.
On the scary side, the new UK variant of the coronavirus appears to be much more transmissible than the original virus, between 30% and 70% moreso. This variant is expected to be the dominant variant in the US by March and that will drive infection rates up. Even more frightening, there is preliminary evidence coming out of the UK that new variant is also perhaps 30% more deadly as well. Mortality data for the new variant are too early to be reliable.
All in, its still too soon to know what will happen to disease spread and mortality in March and beyond. Remember, IT IS IN OUR CONTROL WHAT PATH THE VIRUS TAKES. If we ALL social distance, wear masks, and take the vaccine when it becomes available, this virus would be largely under control in about six weeks. This has always been true. We just need to do the things we obviously need to do, and then take the vaccine to ensure the virus can’t come roaring back.
But as it stands, 100,000 people will die in the next five weeks. There is, finally, hope – but that hope will be dashed if we don’t each do our own small part.
As always, not a medical professional, just a professional statistician that hopes there may be a light at the end of this tunnel.
Citations:
Johns Hopkins University CSSE / CCI; Updated: 01/23/2021, https://91-divoc.com/pages/covid-visualization/index.html
https://academic.oup.com/…/10.1093/cid/ciaa1780/6000389
https://www.economist.com/…/almost-one-in-five…
Steven Millman model comparison of predicted deaths to actual deaths:
January 24th COVID Mini-Update: How’d the model perform over time?
Given the New Year, I thought it might be interesting to look back over my early predictions and see how they panned out. Naturally, I’ve revised the models every two to three months to account for new information and changes in the underlying assumptions, but the first version of the current modeling structure appears to have performed very well over time.
My June 1st estimates were for 333,604 to 373,362 confirmed COVID deaths by the end of 2020. (black arrow defines the range) Seven months later, the actual confirmed COVID death count according to Johns Hopkins turned out to be 345,955 (green line shows actual). I’d say “not bad” but this is clearly horrible.
Why? Presumably they test everyone in class if a kid has it. If an adult in the kid’s house develops symptoms, the kid will be tested. If they have it, their class will be tested.
You’re leaning super hard here on the inability to prove a negative. We can’t prove that there were no undetected asymptomatic cases which were acquired in school, but the very low number of student to student transmissions and the zero student to teacher transmissions are extremely suggestive that transmission is low. Not only that, but this is the result we would expect from the Iceland study.
We have here an infectious disease paediatrician - an expert in the field - conducting as good a study as can be conducted under the circumstances and going from being a schools-open skeptic to advocate, and your response is “sure but what about this thing I thought of which is unlikely and for which there is no evidence but can’t be definitively ruled out”. It’s been the same since day 1 in this thread, anything which says schools are unsafe is met with immediate acceptance and “see we knew children would transmit it, duh” and any evidence that schools can be opened safely is met with apparently limitless skepticism. The random-sample testing you’re proposing would not help, since even if it uncovered a bunch of asymptomatic cases in school, that would say nothing about whether transmission was occurring in that setting.
Lol zero chance
i’m holding out hope that things won’t go worst case and those at risk of dying had at least the first shot of vaccine and get a benefit from it in february, which means deaths decline could come earlier.
Any guesses on where the upward trend on this graph will peak? CVS has said it expand to up to a million per day and I think Walgreens is supposed to have the capacity to do slightly less.
One of the articles linked here suggested Modenra and Pfizer would manufacture 12-18m doses/week worldwide. Wasn’t 100% clear whether they meant each.
You’re not in Kansas any more.
Abolish the CDC!
Lol how? We are way over icu capacity right now and there’s no way LA is better
Never singling out another R bumpkin governor for advocating dying for the economy. What is the functional difference between Newsom and DeSantis at this point?
One difference is that California has a statewide mask mandate. Florida doesn’t and they prohibit local governments from enforcing their mask mandates with fines or other punishment.
Continuing my habit of answering rhetorical questions sincerely, CA seems past peak for cases and hospitalizations, though only just and still much higher than the summer peak.
A difference between CA and FL is the restrictions in CA are still much tighter.
I don’t know where you looked in to it, but it’s not a particularly remarkable claim. It has been reported for a while using quotes from Gates himself, and was when they needed money. The Gates foundation had given then a small grant, then later told them to partner with big pharma and gave them companies to speak to.
This article is from last July.
During the search for money, Bill Gates pushed Gilbert and Hill to partner with a big pharmaceutical company, and as a CEPI founder he had leverage. “We went to Oxford and said, you are doing brilliant work,” Gates recalled in a call with reporters in early June. “You really need to team up, and we told them a list of people to go and talk to.”
Illinois reopened indoor dining yesterday.
Gavin has reservations at the French Laundry coming up though. OPEN FOR BUSINESS!
