The badness of the data is a good reason to stop basing draconian policy on it.

My refrain over the past 18 months, in the face of every confident forecast and definitive statement has been

We really don't know what is going on yet with Covid-19.

This has been largely driven by the fact that neither the US nor any other country has robust and reliable data (though a couple are getting close).  From the beginning we have used loose definitions, varying definitions, and patterns of communication inconsistent with the underlying data being reported.  Beyond bad data collection, we also have the issue of bad protocols.  For example, our PCR tests are way too sensitive and unreliable, returning false positives and false negatives at too great a level.

We still do not have a good mechanism for determining who has the disease nor distinguishing those who die from Covid versus those who die with Covid.  

The second best measure is Deaths from Covid, marred by the failure to distinguish "with" and "from".  

The best measure we have is All Causes Deaths and the corresponding Excess Deaths measure.  There is statistical noise in the measure but broadly, a given stable population has a given death rate over time and you can track changes in that death rate to monitor external impacts.  

For example, if you usually have 100,000 deaths in a year, Covid-19 strikes, and your All Causes Death Rate jumps to 105,000 that year, then it is reasonably logical that your Excess Death Rate of 5,000 that year might be attributable to Covid-19.  But even then you have to be careful about jumping to conclusions.  If the past five winters have been mild with few respiratory deaths among the elderly and the ill, but Covid-19 hits in the year when there is a hard winter and a lot of respiratory deaths, then the Covid-19 deaths will be exaggerated, not because Covid-19 is more lethal but because there are more elderly and ill than in the average year.

But most countries are reporting primarily "cases" of Covid-19 infection, the least useful number because it is primarily a function of rate of testing (the more you test, the more you find), poor testing tools, and weak testing protocols and definitions.  

Given this severely compromised testing regimen, 

We really don't know what is going on yet with Covid-19.

That is also the message from Do Covid Vaccines Stop Covid Spread? by El Gato Malo.  The subtitle is "the answer is complex and honestly, we may never know."

With this much data, we all want answers.  But 1000 reams of bad data are less useful than one well done study and the data quality here is just terrible and getting worse.

As disappointing as this may sound, there are an awful lot of things we may never know around covid and much of what remains of the data cannot support strong claims.

This is a good reason to stop basing draconian policy mandates upon it.

He provides an example of one of the confounding issues.

Covid vaccines have been shown to suppress symptomatic and severe covid in numerous trials.  Yet, out in the real world, they do not appear to attenuate spread of reported cases.

This has been repeatedly blamed on “variants” and “unvaccinated spreaders” but such claims are mostly false.  The real issue is that there is is a fundamental mismatch in the way “cases” were counted in the vaccine trials and how they are counted by health agencies.

The Moderna trial looked like this:

 

Click to enlarge.

If health agencies counted this way, covid “cases” would be 70-90% lower.  Using mass PCR testing on a scale never before even imagined in human history to run 1-2 million mostly asymptomatic people a day in the US alone was always madness.

This test was not suited to purpose.  It cannot discern live or even complete virus from fragments and non viable “dead” virus, and is being run at amplification levels 60-1000X above the maximum at which viable, replicating virus was ever able to be cultured.

We created an industry in the US testing for covid whose annual revenues were ~8 times the size of quest labs’ entire business.  This case mill could only ever produce casedemic. and it’s doing so now.

This spike in cases, among the vaxxed and unvaxxed alike, is not proof the vaccines are failing or that the new variants have evaded them and that vaccine efficacy is degrading.  It’s just evidence of definitional variance:

The trials for the mRNA vaxxes only tested symptomatic individuals and counted positive PCR as a “case” only with symptomatic confirm

Health agencies count any positive test as a case.

See the mismatch? it’s a completely apples and oranges comparison.

Casedemic is the term for an epidemic of cases as opposed to an epidemic which is an epidemic of actual illnesses.  Further:

What’s getting called “breakthrough” is mostly an artifact of a ludicrous testing and definitional system.  The test is just too sensitive.

We can see very clearly that vaccines had no effect on reported case counts in the US.  They were all nice, smooth gompertz curves that were well into seasonal decline before vaccines became a factor and they showed no slope shift from vaccine ramps.  All the age cohorts showed the same curves despite different temporal vaccine patterns.

(Data from CDC HERE, graphic by longtime gatopal™ @justin_hart)

Click to enlarge. 

It’s obvious that rise in vaccination % came well after case drops, did not affect the rate of decline, and that those vaccinating late got the same curve as early.

From this, one might be tempted to conclude that “vaccines do not work” or that “variants are evading them” but neither claim can be substantiated from this data and this is where we need to be careful.

The vaccines never really said they stopped “spread” and the NIH was quite clear about that.  It was the politicians and the CDC and folks like Fauci that were misleading here.

The whole article is a good recapitulation of all the measurement issues which are still preventing us from understanding the nature of and the likely course of Covid-19.

It is an example of government policy making based on garbage data in, garbage policy out.  Despite all the chock headlines, based on unreliable case numbers, it is challenging to remember that (from El Gat Malo):

“Cases” are a meaningless metric because high Ct PCR is unfit for task. 70%+ of reported cases appear to be non-clinical

Deaths are down 71% from same time last year and are dropping vs rising then

Hospitalization is 50% lower

The US is at or within a week or so of seasonal peak.

Community resistance is widespread and effective

Excess deaths have reverted to normal levels

It is doubly frustrating because the CDC's baseless recommendations to mask up again and lockdown the economy again are both now known to be ineffective in the real world.  There is no linkage between lockdowns and masks and reduced Covid-19 deaths.  Just as there was no linkage between vaccinations and death declines.  

The only thing we should know by now is that base of data is too fragile to support strong claims.  Otherwise, 

We really don't know what is going on yet with Covid-19.

I am surprised that there has been little discussion in the press about Wicked Problems.   I posted about wicked problems several years ago in Characteristics of Wicked Problems.  I noted the characteristics of a wicked problem.

The problem is not understood until after the formulation of a solution.

Wicked problems have no stopping rule.

Solutions to wicked problems are not right or wrong.

Every wicked problem is essentially novel and unique.

Every solution to a wicked problem is a 'one shot operation.'

Wicked problems have no given alternative solutions.

Go down the list with respect to Covid-19.  Every characteristic is checked.  We have been treating Covid-19 as a rote checklist of emergency response procedures rather than acknowledging it as a Wicked Problem requiring a different approach.  And requiring much more useful and robust data.