Definitional inexactitude

I have been going on about the absence of clean, reliable, consistent data as a reason to be cautious about what is being reported as happening at the moment as well as being very cautious of any projections. WE DON'T KNOW WHAT IS GOING ON! Because we don't have the clean, reliable, consistent data which would help tell us what is going on.

Based on what we are seeing in other countries, particularly in Europe, we know that one of the core issues is simply definitional. What constitutes a death from Covid-19? It appears that the number and percentage of deaths from Covid-19 as the single causal factor is very, very small. Certainly in Europe, the great majority of deaths are Covid-19 with co-morbidities such as suppressed immune system, diabetes, cancer, etc.

However, just as pneumonia is frequently only the proximate cause of death to an underlying and chronic mortal condition, Covid-19 is a factor in the mix of death, not the sole, or even primary cause of death.

From Europe we know they are struggling with even the basics. Initially, some countries were only counting Covid-19 deaths as those which occurred in hospitals to admitted patients. Die at home from Covid-19 and you weren't in the numbers.

Italy seems to have been an example of an expansive definition. They were one of the countries only counting deaths which occurred in hospitals (which would reduce the real numbers). On the other hand, it appears that their definition was expansive. A death from Covid-19 was counted as attributable to Covid-19 when a patient with a tested and confirmed case died, when a death occurred which demonstrated symptoms consistent with but not tested, when there was reason to believe the person might have been exposed, and for all deaths from underlying conditions which might have entailed some suspected (but not tested) Covid-19.

Ideally, only deaths would be counted where the deceased had been tested and confirmed to have had Covid-19 and with no underlying co-morbidities. In a separate, but additional category, you would have all deaths where the patient had been tested and confirmed to have Covid-19 in addition to other additional underlying co-morbidities. These would have Covid-19 as the proximate or contributing cause of death but not the primary. We would, ideally, not be counting any deaths from patients who had not been tested and confirmed.

Were we to be using this stringent definition universally and consistently, I think there is good reason to suspect that among the existing some 10,000 recorded deaths in the US, only perhaps 500 would meet the criteria.

These concerns stem not only from what we have been seeing from Asia and Europe. It is a well-known industrial engineering phenomenon. The Hawthorne effect:

The Hawthorne effect (also referred to as the observer effect) is a type of reactivity in which individuals modify an aspect of their behavior in response to their awareness of being observed. This can undermine the integrity of research, particularly the relationships between variables.

The original research at the Hawthorne Works for telephone equipment in Cicero, Illinois, on lighting changes and work structure changes such as working hours and break times was originally interpreted by Elton Mayo and others to mean that paying attention to overall worker needs would improve productivity.

Later interpretations such as that done by Landsberger suggested that the novelty of being research subjects and the increased attention from such could lead to temporary increases in workers' productivity. This interpretation was dubbed "the Hawthorne effect". It is also similar to a phenomenon that is referred to as novelty/disruption effect.

There is a measurement variation of the Hawthorne effect. In fact, it a combination of the Hawthorne Effect, the Observer Effect and the Uncertainty Principle. The mere act of defining something or measuring something for the first time causes a change in the perceived prevalence of the thing being defined and measured. You see this in medicine and psychology all the time. We saw this with changing definitions of ADD, Asperger's and Autism - the more we focused on them and the more defined them, the more cases we found even though the underlying reality of any of those items probably did not actually fluctuate.

So with Covid-19. We know how to test for it but have not really done much testing and in the meantime are using loose guidelines to distinguish Covid-19 from the already common conditions of pneumonia and flu which it can easily pass for. With such looseness and the magnitude of deaths from pneumonia (~50,000) and flu (~25-50,000) each year, there is plenty of definitional flux for there to appear to be a deadly threat from Covid-19 when in fact what we have are definitional fluctuations. This is where all deaths morbidity number become important. If it isn't changing, or if only by a small amount, then that suggests that all we are doing is reallocating deaths definitionally rather than seeing a real increase in deaths.

This is not a call to ignore the pandemic or change any of the existing containment or treatment approaches. It is a call for mindfulness that WE DON'T KNOW WHAT IS GOING ON! Covid-19 may be catastrophic and might also only end up being a small blip on all-causes deaths. The more drastic and costly the measures we take to address the feared pandemic, the more we need to be clear that it is also perhaps much ado about nothing (metaphorically).

Are COVID-19 Deaths Being Overreported? by Julie Kelly is one of the few articles which I have seen to focus on this critical issue of definitions. It is useful in that she focuses on the US numbers. All my concerns so far have been based on European and Asian examples and the long-established Hawthorne phenomena. At least with her account, there is some American specificity. There is a bit too much partisan jabbing but it at least deals with factual content. Bureaucratic laziness and sloppiness in definitions does not necessarily constitute malice or "rigging".

I would also note that there is a real chance that Coid-19 deaths are being under-reported.

An example for why there might be concern that America is going to have the same problem with definitional inexactitude as everyone else is illustrated by the guidelines.

Guidelines recently released by the Centers for Disease Control bolster concerns that the death toll is being rigged to show a higher fatality rate.

“In cases where a definite diagnosis of COVID–19 cannot be made, but it is suspected or likely (e.g., the circumstances are compelling within a reasonable degree of certainty), it is acceptable to report COVID–19 on a death certificate as ‘probable’ or ‘presumed’,” the agency advises. “In these instances, certifiers should use their best clinical judgment in determining if a COVID–19 infection was likely.”

That clinical judgment, alarmingly, does not require administering a test to confirm the presence of the virus.

“Ideally, testing for COVID–19 should be conducted, but it is acceptable to report COVID–19 on a death certificate without this confirmation if the circumstances are compelling within a reasonable degree of certainty,” the guidelines state.

The CDC provided three examples to help officials determine how to properly document the cause of death. One scenario described an 86-year-old female nonambulatory stroke victim who developed a fever and cough days after being exposed to a sick family member later diagnosed with COVID-19. Even though the decedent wasn’t tested, the coroner nonetheless determined that the woman’s underlying cause of death was COVID–19, “given the patient’s symptoms and exposure to an infected individual.”

Let’s just say that kind of bureaucratic guesswork is unacceptable while the economy is in chaos, tens of millions are suddenly out of work, and power-hungry government tyrants arrest surfers and pastors for daring to violate “social distancing” decrees handed down to their local authorities by Beltway lifers.

It is going to be long while before we discover the real numbers.

On the other hand, for all the economic damage being done (and I do not discount that), if the true all-causes death count turns out to be low for Covid-19, it will have served as an excellent dry run. Our bureaucratic infrastructure was clearly ill-prepared for a real pandemic, both in terms of speed of decision-making, stockpiles of critical supplies, modeling and in forecasting. Strategically, this might end up being a low cost stress-test of our global and national systems which, if we pay attention and learn our lessons and make better organizational changes, might mitigate or preclude a future and more lethal pandemic.

But there will have to be consequences to the institutional incompetence and malfeasance which has been too often on display. Careers will have to be ended, people fired, accountability exacted. And we are not especially keen on those sorts of things.