Not when you were previously seeing a doubling every three days it isn’t. 60k to 100k is in the noise. Two days either way at the start of the lockdown. That is why it is so important to get on top of things early.
No lockdown, and you would trivially have seen ten times the current toll. More likely significantly more. Seriously, two million dead by mid-year with no action taken was perfectly possible. Everything is exponential. Linear thinking just doesn’t work.
Avoiding the discussion of assholes versus the tendency of many people to take advantage of they can take advantage of, (especially when they know others are doing it) the discussion began with an apparent serious suggestion of doing 22 million tests, roughly 7% of the U.S. population, per day. Not home testing? That’s every American going into the lab for a test once per two weeks? Really?
And which test is being imagined being done? Nasopharyngeal swabs? Okay real time with infection (the antibody test is mostly after the fact of contagiousness). But clearly not an at-home process and in a low infection world (the proposed plan is to catch hot spots when rates are low), say 0.1% infection as baseline, a good specificity test with a low false positive rate of 2% would be telling 2% of the population that they were infected and had to stay home each time, while about 99.9% of them would actually not be infected. Based on what we currently understand the test would also be telling 30% of those with active infections that they were fine to work. At the current Medicare payment rate of about $50/test it would be an expenditure of $1,100,000,000 ($1.1B) per day, over $4 trillion a year in testing costs alone.
There is no way this idea should be considered as a serious suggestion.
Fortunately good surveillance testing requires much smaller population samples, provided reporting systems are efficient.
We are talking about now, not a hypothetical world with no intervention a month ago. Looking back, I rather doubt the millions+ fatalities was ever really likely but I have not objected to any of the lockdowns so far.
Is COVID less deadly now than it was a month ago?
Is there any reason to think a ‘no intervention’ world starting now or May 1 is significantly less deadly than one starting a hypothetical month ago?
What aspect of today’s world would suggest that 100k dead per month is ridiculous, when we’re already seeing a rate of 60k/month with heavy restrictions?
Still the point remains. The lockdown started at a time when the rate was a doubling every three days. If you wait say two days, the entire epidemic has about a 50% additional head start. That 50% stays. The entire curve is 50% higher, and all the reductions in rate have to start with that additional load. It isn’t a linear problem. So that 60k to 100k difference is just in the noise, it represents two days of dithering about.
This is the problem with dealing with exponential functions. It is the exponent of the numbers that matters.*
*A bit like cosmology. Where the errors bars are on the exponent. 
Yes, there’s reasons. The curves seem to be flattening under a variety of lockdown rules around the world. And I could turn that around: why are you so sure it’s as deadly as we feared months ago, when we really had no clue?
I mean, you realize we can’t literally stay in this lockdown forever/until a vaccine is developed, right?
So someone has to go to a lab every day to get tested and if they are positive they’ll get a certificate that allows them to stay home?
No, I don’t think people are all assholes. But you did refer to yourself as a pollyanna so I can see where you think everyone, all people, would be completely honest and never try to abuse a system like that.
Can you explain how surveillance testing works? I really have no idea how that gets implemented and how we enforce results from it. How does it keep 100 “Typhoid Marys” from putting back to early Feb?
We have better numbers, but we don’t have the luxury of waiting until we have a perfect understanding. We are a bit like a well known question. “I know what you’re thinking. ‘Did he fire six shots or only five’?”
Like I wrote before. Science isn’t a democracy. You make your choices, you wear the outcomes. The facts of how the disease operates don’t care. Everyone gets stir crazy, no vaccine, everybody says “to hell with it” - the disease has no concern. It will simply do what it does. Your call but, you’ve gotta ask yourself a question: 'Do I feel lucky?
We live in a democracy whether you like it or not. Science lists our options and one of those options is to slowly stop the lockdown and see how it goes. Science is not going to promise us a perfectly safe choice. It doesn’t even always promise whch is the safest choice.
But just to be clear, you are indeed suggesting the lockdowns stay until there’s a vaccine regardless of how long that takes?
No, I’m saying you can’t force reality to conform to your desires. Science isn’t a democracy. You can’t change what it predicts by force of will or vote.
If there is to be a release of lockdown it is either going to have to conform to some carefully structured plan, or, projections suggest, it will end in disaster. No amount of pleading that you live in a democracy will alter this. The virus does not live in a democracy. You can democratically decide to what you do. You can’t democratically decide what the repercussions will be.
You at least start with the systems built for influenza surveillance. https://www.cdc.gov/flu/weekly/#S2
This is a collection of regional to local monitoring systems that look at a variety of inputs, inclusive of but not limited to testing that is used to represent likely population wide percentages. IF the plan to have SARS-CoV-2 testing as part of that system in place early had actually happened the whole course here would like be very different.
We are at a mitigation strategy point. Maybe there was a point that containment was possible but it is long past. The goal now is to keep new cases, hospitalizations, and deaths rates at a levels below typical seasonal flu’s rate with the lightest touch that accomplishes that goal. As the brakes are eased slowly you monitor in real time for actual case rate increases by such sampling and the usual influenza monitoring model perhaps with reporting made more efficient. You respond according to results and make a next move be it more brake back on or eased up more based on both the results and the models informed by better data as you collect it. Let the data be the guide, not whether or not you feel lucky punk. Don’t mistake assumptions for facts.
You just blew my mind. :rolleyes:
The projections are all over the place so you can spare me the “we all gonna die if we lift lockdown now”. Of course it’s wise to do it in a measured manner. I’m not saying let’s just go stupid but we will definitely start testing the boundaries as soon as numbers start flattening.
Exponential growth will always elude some people.
I’m assuming the case rate monitoring is on a passive basis, meaning we look for people seeking medical help and those results are returned back through the systems outlined in your link. It looks like there are several.
Does this type of system have too much lag to be able to effectively prevent medical systems from being overwhelmed?
Thanks for the info on this. Obviously this type of system is actually employable and what is needed is for it to be streamlined and enhanced, not built from the ground up. It seems to me this system will not really be protective of the most vulnerable, short of waiting for a vaccine that may never come I don’t know that any realistic system will be. What are your thoughts on that?
It would be reasonable and achieve to supplement use of the established surveillance network with some active screening of asymptomatic population samples, both for active disease and for evidence of past infection with presumptive current immunity.
But monitoring out patient visit, hospitalization, and ICU utilization rates will be real time enough to make changes every few weeks.
The most vulnerable should try to stay as protected as possible for as long as possible.
Screening of asymptomatic population samples is testing patients that come to medical facilities for something unrelated, like a broken arm or to give birth?
That is one approach, the “convenience sampling” approach, but that approach is subject to selection bias. Better is to use similar methods as pollsters use, with similar sorts of sample sizes and correction methods.
Coming in to give birth has been one such convenience sampling technique used and has demonstrated that 15% of women giving birth in New York end of March to early April were actively infected, with 82% I believe it was, asymptomatic. Given that this was at a point that new infections had likely been peaking there were also likely some similar number to more who had had infections and had resolved already, but no antibody testing reported as done. By another week to ten days most of that actively infected 15% would likely test negative and be in the resolved bucket.
Another sample of convenience was blood donors in an Italian town in which about two thirds, if I recall correctly, tested as immune.
Again though these sorts of convenience samples are not necessarily representative of the wholes.
I’ve read your posts on those numbers as well as the ones you dug up on the USS Roosevelt. Thanks for doing all that research.
There are a few hopeful signs out there. The data from China is encouraging, as is the data you talked about here. This pandemic has a chance of being a warning shot for us.