Keep Lockdown

It’s interesting to look at the map of COVID-19 cases by state:

The bigger the red circle, the higher that state’s number of new COVID-19 cases per day per million population.

Downward-pointing blue triangle: rate is falling (congratulations, New Jersey).

Upward-pointing black triangle: rate is rising (watch out, Iowa and Nebraska).

Thanks to all for kind donations to support the blog. If you want to help, please visit the donation link below.

This blog is made possible by readers like you; join others by donating at My Wee Dragon.


21 responses to “Keep Lockdown

  1. One of the faster growth areas has to be Washington DC, which makes up eastern Virginia, the Baltimore area, and, of course, DC itself. My son is a physician at one of the major hospitals in that area, So far they do not have a large number of patients in the hospital. That could soon change.

    The only thing that can save people in Red States from their Governors is heat. If Texas heat slows/stops the spread of the virus, then the end of lockdown, which starts this Thursday, will look smart, even though it is an unconscionable and unnecessary gamble. The heat is also starting now.

    • Jeffrey Davis

      Places like the UAE have problems with COVID-19. I don’t know exactly what heat is supposed to do.

      • Keep people inside, I assume. The problem is that they won’t stay inside at home, but in crowded shops where the air conditioning freezes their brains. (There’s no other sensible explanation why they’re set so low you’d need a sweater inside when outside temp is more than sufficient for going naked …)

      • SARS-CoV-2 has a lipid bilayer which could, could, render it vulnerable to high summer temperatures. Nobody knows.

  2. Michael Sweet

    Is there enough testing for the numbers to be accurate? A state with less testing might have a lower rate than a state with more testing when the actual rates were opposite.

    • I can’t speak authoritatively to that question, but I note that the US has now performed over 6 million tests, and its testing rates are comparable, broadly speaking, to many other affected nations. (Surprisingly, we now have a higher test rate than South Korea–presumably because they no longer need to do widespread testing, having successfully contained the epidemic.)

      Moreover, every state has performed significant numbers of tests. The worst laggard in absolute numbers, per JHU, is Montana. They’ve done 13,528; their caseload is 451. On a per capita basis, the range goes from Kansas (9.7 per thousand population) to Rhode Island (56.9 per thousand).

      Given that the function of testing is now primarily clinical, not epidemiological, I suspect that we’re getting a reasonable picture of the relative infection rates among the states: if testing is mostly done to confirm diagnoses for clinical purposes, then asymptomatic folks are mostly excluded across the board, whether it’s Montana or Rhode Island.

      Just my two cents.

      • Let me add a cent to yours. You write:

        ” I note that the US has now performed over 6 million tests, and its testing rates are comparable, broadly speaking, to many other affected nations. ”

        Are you sure?

        US is below 20,000 tests per million; highly affected countries with a much higher case fatality rate tested better, e.g. Italy, Spain, Germany, Switzerland, all over 30,000 p m.

        Even Russia and Belarus are above the US in the test stat.

        J.-P. D.

      • Am I sure? No. I’ve seen some wide variations in testing stats particularly. But while my comment may or may not be correct, it’s not unfounded. Per the Worldometers tracker linked above, here are some representative numbers on national testing rates per million population (in descending order):

        Italy: 32,735
        Switzerland: 31,371
        Spain: 31,126
        Germany: 30,400
        Russia: 25,354
        Canada: 21,367
        US: 19,362
        UK: 13,286
        Netherlands: 13,184
        Turkey: 12,255
        South Korea: 12,153
        Sweden: 11,833
        France: 11,101

        Note that I’m not claiming there’s any rigor to that list; it’s just basically a stroll down the sorted table, picking out notable nations. And I purposely put the US in the middle to illustrate what I was thinking above, so there’s nothing probative about that placement. It’s just meant to illustrate.

      • One additional interpretive comment.

        I’ve come to think that the crucial issue is not so much how many tests you run over time, but how many tests you run early. I’m basing that on two comparisons: the US v. South Korea, and the US v. Canada. One is above the US in the ‘tests run per million population’ metric, one below. Surprisingly, the nation with the best-controlled epidemic is the one *below* the US!

        Yes, South Korea, pioneer of drive-through testing; the same nation that yesterday reported exactlyfour new cases of CO19; that has an active case/cumulative case ration of only about 14%; that is actually holding a professional baseball season; has now tested at a rate 43% below the US, where the curve has been flattened, but not nearly enough to say the epidemic is controlled.

        But there is a commonality between the two comparisons. Both Canada and South Korea have per capita caseloads considerably lower than does the US. And the American comparability–if you accept that it exists–in testing is relatively recent. Early on South Korea and Canada were both doing much more testing than was the US. And the US testing paradigm was such that it nearly ensured that early cases of community spread would be overlooked, unless and until they resulting in clinically serious illness.

        Early on, you can use tests to isolate not just patients, but their contacts. This is what China reportedly did. At this point, testing is an important public health tool. But as caseloads climb, thorough contact tracing becomes impossibly labor intensive. Testing remains valuable, of course, but not as a public health tool: its importance becomes primarily clinical. You need to know if Patient X has Covid, or something else, in order to manage care, and also to isolate him or her. But that’s it. Testing in this situation contributes much less to actual control, or the lack thereof. Basically, all you can do is social distancing-style measures.

        So I’m thinking that there is a window during which testing is potentially highly effective. Miss it, and well, that ship has sailed, until you get to the winding-down phase of things. Then you can once again do contact tracing, and effectively ‘mole whack’ the disease.

      • at Doc: Yes, I think you are correct about early testing and contact tracing. After we move to active pandemic infection rate within general population, the infection testing is about treatment and triage, quarantining, etc. It is no longer useful to prevent moving to the spot where the pathogen is spread widely within the general population. Once you get there, the pandemic prevention ship has sailed and all you can do is flatten the curve with familiar techniques.

        Once we are at this point, the more important control projects become antibody testing with expectation of a variable level of immunity from secondary infection and development of vaccines to achieve a similar antibody production level from a less dangerous form of encounter with the pathogen. That is where we are at now imho. I am on a Medicare HMO program and I am looking forward and hoping that organization will offer a covid antibody test sometime soon.

        The US screwed up in a major way by not proceeding with testing based on the WHO testing setup. It has gone downhill from there. Mr. Trump doing what he does best: getting it wrong in a major way.

  3. Robert Eidson

    Heat effects people not the virus you fools.

  4. If the metrics are bad, hide the metrics. That will solve the crisis. Apparently Florida is trying this tactic.

  5. On what time scale are the rates being evaluated? My plot of New Jersey daily confirmed cases looks mostly flat and very noisy since the end of March and only in the past week has it been going down. Maybe the difference is my data source which is the Wolfram Data Repository which is compiled from John Hopkins CSSE data or maybe there is a bug in what I am doing.

    Most states have a bent curve of confirmed cases when plotted on a log scale but Nebraska looks like a mostly straight line. Hopefully it bends soon.

  6. I think it’s important to separate the number of tests that the Harvard study suggested (5 million tests per day scaling up to 20 million tests per day) would be required to manage the pandemic in the US from the total number of tests that have been done. Otherwise, we could start sounding like Donald Trump and then, we will start trying to insert light through our skin and drinking bleach to clean our insides. To be clear: Harvard study said we need 5 million tests per day scaling up to 20 million tests per day.

    It is unlikely the US can do anything close to this level of testing, so we are likely to engineer a path through this pandemic that kills a lot of Americans.

    Also, to be clear: UV lights in injections of disinfectants are dangerous ideas, not effective treatments or ideas to be put into testing. The testing has been done. These treatments can make you sick or kill you. They are not reasonable ideas for responding to the Covid pandemic.


  7. That many viruses are seasonal is well established. What causes it is not well established.