Dr. Aaron Glatt on Covid Stats June 2020
Claim: Covid metrics as of June 11 2020 indicated a significant resurgence of covid spread.
This is part of a series to highlight the ineptitude and sheer illiteracy of the experts my community has relied upon for covid medical advice. I am only employing arguments that can be made from the data/studies and other information available at the time the claim was made.
This rebuttal was written a few days following the publication of Dr. Glatt’s article. It has been lightly edited for grammar and clarity.
Claim: Covid metrics as of June 11 2020 indicated a significant resurgence of covid spread.
Source: Rabbi Dr. Aaron Glatt COVID-19 Update, June 11, [2020] 9:00 PM
Background: Major lull in covid heading into the summer of 2020.
Dr. Glatt claimed that the number of infections is increasing in some areas of the country, as seen in twenty states’ “significant[ly] rising” number of infections; RD Glatt chose to highlight Texas and California as specific examples that illustrate this. The argument - though not articulated fully and relying upon the reader to work out on his own - essentially is that the rising number of covid infections indicates that covid transmission is rising, and is responsible for the increase in infections. This, in turn, indicates that covid is not in its death throes, or irreversibly dying out. If that is true, then it follows that a covid breakout can recur in a place that seems past it, like NYC, particularly if seeded by someone/s traveling from one of these new “hot spots”.
The Facts:
Exploring how the covid data are curated
All 50 states kept track of some basic covid statistics, including:
the raw number of positive tests, negative tests, total tests
the % of total tests that are positive
the number of hospitalized patients who have covid
the number of ICU patients who have covid
To understand what any particular data point represents in the real-world reality, it is necessary to begin with an introductory analysis.
The number of positive tests refers to the total number of covid tests conducted that returned a positive result. This is not the number of people who have a new, previously unreported covid infection, which is an entirely distinct group of people.
Firstly, the standard for being considered “recovered” from a covid infection is that you test negative twice, consecutively. This means that one person can easily be responsible for 4 or 5, positive tests, as people (especially those who developed symptoms after the symptoms have subsided and wish to be tested immediately) get tested repeatedly until they produce the requisite two consecutive negative tests, in order to return to work, leave quarantine, or even to feel comfortable resuming a less restrictive social life (all things considered, obviously). This has become considerably more prevalent than a month or two months ago, when testing capacity wasn’t anywhere near as large, and tests were being triaged and thus not as readily available on a whim for everyone.
Secondly, the PCR tests that are used also can pick up dead viral debris, sometimes even weeks after someone cleared the infection (I saw a case where a woman was testing positive 8 weeks after full symptom remission), thus some as yet unidentified % of the positive tests are in reality dead virus in people who already got rid of covid, and therefore cannot possibly be indicating of a current spike in active covid infections.
These factors skew the positive % higher, despite it being a mirage. And even after factoring a more carefully nuanced grasp of what the # of positive cases represent, there is another critical factor necessary to understand the context, insofar as determining the transmission rate:
What is the % of positive cases that exist at the time that are captured in the public infection surveillance (i.e. testing).
In other words, if we make a thought experiment, supposing that if two months ago (due to the much more limited testing capacity), only 10% of active cases were discovered, but now, 45% of active cases are discovered, then the number of active cases we see now versus the number of active cases discovered 2 months prior would have to be radically adjusted to get an idea of [what turns out to be] the plummeting number of active infections day by day. This is true even to compare now to a week ago – you have to compare the number of tests carried out to extrapolate a rough ratio of active cases from then to now.
For the most part, once taken into account, the increased testing usually by itself demonstrates – through a declining positive % of test results – a steady, downward trend of covid infection. This is emphatically true on a country-wide level.
Importing Mexico’s Covid into US Covid Data
There is a further confounding variable that, although rather scandalous that it is not reported straightforwardly, is critical, especially to the two states specifically cited by RD Glatt, among others:
There has been a massive surge – as in many thousands – of Mexicans who crossed the southern border to get better hospital treatment for acutely severe cases of covid, almost 100% of whom previously hospitalized in Mexico. This is in addition to the nearly 200,000 dual citizens or of other legal status in the US that work in Mexico and have crossed the border recently, from Mexico’s hottest hot zone.
Importing thousands of highly symptomatic – and therefore, highly contagious – individuals into the country leads the logical observer to anticipate that this phenomenon should manifest itself in the form of higher transmission, as these contagious individuals proceed to infect others. The importation of a more than 100k individuals from a covid hot spot likewise would lead one to conjecture that there will be a spike in the number of positive cases discovered, especially considering that these individuals are disproportionately likely to be tested because they came from a covid hot zone.
Hospital Stats
Another big data point being trumpeted as proof that covid infections are spiking is the hospitalization numbers.
Dr. Glatt is arguing that “more people in the hospitals with covid = rise in severe covid cases”. This argument collapses like a flimsily built sandcastle by mere cursory analysis.
“The # of hospitalized people with covid” refers to exactly that, without differentiating between those hospitalized because of covid, and those hospitalized with [typically asymptomatic] covid. Now that states have finally allowed elective procedures to resume, there is a surge of hospitalizations; this is a good thing, however!
Hospitals test every living thing that crosses the doorway. This means that every individual coming in for any surgery, procedure, or appointment of any sort is getting tested. Obviously, some % will test positive for covid asymptomatically. They are counted as covid hospitalizations, though (hmm… might juice the stats…).
Now, as of a few days ago, only 1.9% of ER visits in the US were because of “influenza-like symptoms” (which include covid symptoms, and let’s even concede that the vast majority of those are symptomatic covid). That is nowhere near remotely high enough volume to sustain the number of reported “covid hospitalizations”.
Similarly, there is mass confusion by many medical professionals between hospital bed usage, and hospital admissions.
The hospital usage stat refers to the cumulative number of hospital beds that are occupied on any given day, treating equally someone recovering from surgery 2 weeks ago and someone admitted today. That number has been steadily rising, and will probably rise until hospital capacity is a bit higher than average, as a backlog tens of thousands of procedures long, is finally addressed.
The # of hospital admissions of covid positive people, however, is dropping, even in some of the states that are supposedly the hospitalization hot spots. Yet it is usually the former stat that is cited by the experts.
Importing Mexico’s Hospitalized Covid Patients Too
It is also necessary to point out what I referred to earlier about the Mexican covid invasion, as particularly in regard to the covid hospitalization #, the Mexican invasion debunks it completely.
The NYT, of all newspapers, published a very lengthy, in-depth piece about this. (Staying true to form, though, they somehow neglected to mention the massive statistical impact that this represents.) They even discovered hospitals that were overrun and had to move covid patients farther north where there wasn’t a covid run on hospitals (before they started receiving “out of town” covid patients anyway).
This meant that there were thousands of Mexican severe covid cases being imported into the US healthcare system being counted as US severe covid cases (and even deaths).
That the official US data is incorporating another country’s cases, and worse, that these cases were themselves largely responsible for the bulk of the spike in hospital admissions because of covid is quite scandalous.
But it gets much worse. Very symptomatic covid patients - the sort that are progressing to hospital-worthy severity - are going to be far more infectious cases. Thus, they would almost definitely be the source of any genuine spike in covid infections too – particularly the hospitalized cases. So not only did the US data get inflated by foreign cases, the US population was exposed to an extra artificial surge of super-spreader patients too!
This is what Dr. Glatt would have us all believe is a signal of natural covid resurgence.
(This statistical chicanery follows the pattern established by the US death count, which included gunshot victims, suicides, alcohol poisoning, inoperable brain tumors, hospice residents, and drug overdoses, among other “clear alternate causes of death”, in addition to the many thousands of asymptomatic covid cases that were somehow killed by the asymptomatic covid infection.)
Experts Cannot Read Basic Hospital Data
Additionally, almost everyone I see is unable to distinguish between the date a case (or death) is entered into the published data, and the date on which it occurred or was discovered.
This is a far more confounding issue than one might think, because many states, especially now that the daily new reported cases have plummeted, have begun to sort out backlogged reports that were still waiting to be entered into the published data.
What they did not do, however, was to backfill the new cases to the date on which they were discovered on their dashboard displays. (Some do have a breakdown hidden in some random appendix type data document that is not perused by any except the most daring and patient data nerds).
This means that, depending on the state, an oftentimes significant % of their “new” positive cases are not exactly new, yet they are being counted as new cases! This creates the mirage of more cases than actually exist currently.
And this isn’t only due to bureaucratic incompetence. In Virginia, they managed to only enter the positive cases from the non-electronically reported lab conducted tests, while leaving out all of the more than 43,000 negative ones, and conveniently neglecting to inform anyone of this. This predictably completely skewed the positive test result %. The positivity # which also happened to be one of their critical indicia of covid reduction tied to their reopening phases.
There has been lots of data malfeasance in many states. A few states even managed to somehow combine the PCR tests with the antibody tests in the reported tests, which is an extremely illiterate thing to do. (Arizona, we’re looking at you in particular.) PCR tests (in theory) are measuring active infections. Antibody tests are measuring infections that have been resolved already, and are no longer active covid cases. Combining them denudes the positive test % and overall number of any coherent meaning, and is something that a competent 9th grade student would know not to do.
I have resisted the temptation (mostly because I was too lazy to do the hours of tedious work it would entail) to go state by state through their data detailing the various perfidies and entanglements of unrelated categories in their data sets. The broader ‘therefore’ from all of this, though, is that the data did not show any meaningful spike among US covid cases at that time, especially not one that would indicate that transmission has meaningfully risen anywhere in the country. The failure to realize this was a failure to read the elementary data, and a failure to pay attention to outside confounding factors that introduced rampant exaggeration to the critical categories that are used as proxies to determine covid transmission levels.
Unfortunately, I did not include in this section charts, and I don’t have time to spend a few hours reconstructing them from datasets at the moment. But I think (hope) that this still stands on its own.
At any rate, the arguments made definitely have aged well. Obviously, covid did come back in multiple waves with a vengeance. But that happened months later, and was certainly not indicated from any of the data contemporaneously available like Dr. Glatt was trying to argue. And the mitigation measures he was arguing in favor of definitely did nothing to hamper covid in any way.
What is quite clear though is that Dr. Glatt was functionally clueless and unable to properly read or understand the data and studies.