Wednesday, April 01, 2020 ... Deutsch/Español/Related posts from blogosphere

A slowdown of a rise of cases saves almost no lives

"Models" and people claiming otherwise are deceiving the public

There are lots of data, rapid societal changes, conceptual principles to discuss during these Covid-19 times. Finally someone has made the same rudimentary yet absolutely essential point that I have been emphasizing for more than a month (but never dedicated a special blog post to):

If your package of restrictive policies only manages to "slow down" the growth, it saves virtually no lives.
This point was made by Maria Chikina, assistant professor of molecular biology at Univ. of Pittsburgh, and Wesley Pegden, associate professor of mathematical sciences at Carnegie Mellon Univ.:
A call to honesty in pandemic modeling
Thanks to Čebyšev for the URL. I think that you must have seen these claims many times because they're everywhere. "Our measures have already saved 100,000 lives" or something like that. The only problem is that all these claims are lies.



So these two scholars have pointed out the same idea as I have pointed out so many times – but so many people still seem so incapable of getting these elementary things: if you don't eliminate the virus by dedicating X months to some draconian measures, you haven't changed anything qualitatively because at the end of this X-month-long period, the "state of the system" (primarily described by the number of active infections and the percentage of the immunized population) is basically equivalent to what it would be without any measures at some intermediate point of the interval when the measures took place, but didn't have to.



It means that as long as the draconian measures only last a finite amount of time, and they better should, and assuming a macroscopic number of infected people still exist after the X months of these draconian measures, the normal growth of the infection gets restored at the end of the period of the draconian measures. In effect, the whole evolution – which may want to drift to some equilibrium, probably including herd immunity, at least in some regions and environments – was just delayed. It was just delayed by an amount of time that is shorter than X months, probably much shorter than X months.

The people claiming that "measures have saved lives" are only hiding the fatalities into the future which they account for inconsistently. At most, the number of fatalities in the whole future may drop by something like 1% because the policies help to build the immunity in a less fatal way (but only a small percentage of the people get immune so it doesn't make a big relative difference for the future fatality count). But nothing changes substantially about the "overall future" and the lives that will be lost due to the infection were just extended by that "epsilon" which is shorter or much shorter than X months!

Equivalently, these two faculty members point out, the models claiming that "lives have been sent" are deceitfully cut off at some future moment. They overlook the lives that would be lost soon after the cutoff in the scenario with the draconian policies, but that wouldn't exist in the scenario without the draconian policies. It means that an honest treatment must avoid such cutoffs. Instead, an honest modeling of these questions must always consider the time frame up to a moment when some equilibrium develops!

If your assumptions actually imply that at some moment, perhaps later in 2020, 50-80 percent of the people (or a similarly high percentage of cities and areas where the infection tends to grow) will have been exposed to the virus, it clearly means that it doesn't really matter much whether we get to this point with a one-month delay or without. And you just shouldn't obfuscate the essential point that the measures don't change anything in the positive direction.

For example, in the case of Lombardy, a 3-month shutdown may have cost hundreds of billions in economic damages (and created a million of unemployed people); 3 months of a pretty ruined life for 10 million inhabitants of Lombardy; and the extension of the life of (ultimately) 20,000 victims by 1 month – basically the same people will die, anyway, before some equilibrium is reached. I am assuming that the growth that takes place under the 3 months of lockdowns would normally be achieved in 2 months without a lockdown (and maybe lockdowns literally make things worse so I am very generous here). It's clearly not a good deal.

I join these two wise people's call for the scholars and politicians to stop lying. As long as the number of infections remains substantial at the end of the lockdown, the restrictive policies save basically no lives and they're nothing else than a waste of trillions of dollars and the economic suicide of whole nations! Restrictive policies only make sense assuming that one can basically make the virus extinct within a tolerable period of X months (as in China, if we trust them and our eyes), at least regionally. If that outcome cannot be made almost guaranteed, the restrictive policies only have huge disadvantages, no material advantages, and they should be completely avoided.

P.S.: Near the maximum of daily new cases, restrictions and slowdowns could make some sense and they would reduce the maximum of daily new cases. Also, when the hospitals are overloaded, a slowdown could help (perhaps the restrictions could be justified one week in advance). But it is completely futile to try to slow down the dynamics well before (whole months before) these two points are reached, i.e. "preemptively". Sadly, that's exactly what is happening in many countries today.

Add to del.icio.us Digg this Add to reddit

snail feedback (0) :

(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) })(window,document,'script','//www.google-analytics.com/analytics.js','ga'); ga('create', 'UA-1828728-1', 'auto'); ga('send', 'pageview');