George Rebane
The discussions about the procedures, reliability, timeliness, types, … of testing in the media, and even in these pages, can bring a smile to the lips of those with a modicum facility with numbers. People still believe that testing for active infections at current levels (about 5%) of C19 prevalence is the secret sauce to stopping the epidemic within these shores. The Left uses this argument to lambast Trump – if he’d only started test making and testing earlier, then none of this would have happened. Any realistic look at the timing, what we knew when, and the logistics involved would reveal these arguments as sophomoric, destined for even more simple ears.
Contagious diseases spread through a population by a process in science known as diffusion, a physical process that is described by some really snarly partial differential equations. As we have discussed in these pages, epidemiologists have broken it down into a couple of graspable concepts (to some) in the forms of “the basic reproductive number”, R0, and herd immunity. R0 can be understood as the number of vulnerable people an infectious person will infect in a relatively short interval after becoming contagious, which is of the order of a week or less in a low prevalence population like ours today. According to the CDC (‘High Contagiousness and Rapid Spread of Severe Acute Respiratory Syndrome Coronavirus 2’), the Wuhan population diffusion data indicated an initial R0 = [2.2-2.7] which subsequent analysis upped to R0 = 5.7, (95% CI 3.8–8.9).
The CDC paper gives even the technical reader a bit of a headache as to how their analysts handled the multiple levels of modeling complexities, each requiring the use of blatantly brown numbers to fill in for the unknowns and the barely perceived. But CDC’s timing the receipt of the China data and its subsequent massagings again makes clear that there was no chance for anyone in America (or Europe) to have been sufficiently prescient to comply with the butt-stupid accusations of Trump’s mishandling preparations for the onset of C19 on these shores. Unfortunately, it takes a technician with some appreciation for the math and probabilistics involved to understand how facetious these know-nothing charges from the Left are – however, the sheeple remain clueless and at the mercy of the loudest shouters.
But back to the numbers. Today, 10may20, we are told that there are approximately 20K/day of new confirmed C19 cases, each of whom have already contributed their reproductive number of new cases that will become apparent in about a week. We may safely assume that at least another 20K/day will remain asymptomatic, with each doing their R0 subrosa infections, and recovering on their own. If we take the CDC’s low end R0 of about 2.5, then we have an infection doubling time of about a week. That means that the 20K+20K infected individuals have now become 80K, with the symptomatic ones waiting to be discovered. Now consider what level of infection testing is required to ferret out these 40K infected persons from a population of 330M containing, say, 20 high density, interacting (diffusing) sub-populations, before some of them become symptomatic, how many thousands will be missed by even the most comprehensive testing programs (which I have covered in detail), and what was the contacts fan-out for each of these infectious persons before they were taken off the streets.
It is this fan-out of contacts which now must be traced, tested, and quarantined in very short order. And it doesn’t stop there, the fan-outs have fan-outs have …, I hope you get the idea how this expanding ripple-cum-tsunami works. There aren’t enough tests, or timely and reliable testing procedures, or fan-out tracers to even make a pin prick in the numbers required to "control" C19 or any other human-to-human contagious disease in dense, high population areas, except perhaps at the earliest level when there are substantially fewer than maybe 100 infected, and when all response systems are assembled, alert, properly positioned, and ready to leap into action (think of the cost to maintain such systems at the ready in perpetuity, like the fire department or EMT). And at such a low number of infecteds, the authorities will take a week or two to confirm what’s happening, then scratch their asses, and puzzle out some initial inevitably ineffective response. That’s what happened, and will happened again and again.
OK boys and girls, can we all now spell ‘H-E-R-D I-M-M-U-N-I-T-Y’, and then ‘V-A-C-C-I-N-E’?
[11may20 update] Well, well, well – we finally have someone from the medical community starting to think strategically about the realities of testing, something we’ve been doing on RR for almost two months. Philadelphia cardiologist Dr Anish Koka offers his quantitative take on testing in ‘Why More Testing Won’t Solve The COVID-19 Crisis’ that concludes, “Inherent limitations of the tests we currently have make it unlikely that applying testing to the entire population is going to be very useful.” His incomplete analysis ranges between confusing and clutsy, mainly because he doesn’t understand the role of Bayesian inference in computing the results from unreliable tests applied to populations with an estimated prevalence of a disease. But not to put too fine a point on it, the published attempt is a start of actually dealing with the realworld factors of testing, factors which continue to be absent in today’s national C19 dialogue. Innumeracy at our highest levels of leadership, let alone the know-nothing media, is such a curse in our nation, mostly because no gives a crap about such ignorance. (RR shibboleth: If you don’t understand the numbers related to a social problem, then you don’t understand the problem and can only emote about it.)
rebanesruminations.com 
Leave a comment