Some thoughts to frame our minds:
In How Not To Be Wrong, Jordan Ellenberg opens with a story about up-armoring bombers in WWII. The Army Air Force wanted to enhance survivability for the bombers by adding armor to existing bombers and by incorporating armor in selected areas in new designs but they were limited in how much armor they added as it increased weight which would reduce ordnance payload and hurt takeoff performance. To determine where to add armor, they studied aircraft returning from missions determining bullet hole density in respective areas. Initial thinking was that the more densely holed areas were the most likely to be hit, hence these areas should be armored. Along came Abraham Wald asking where are the holes in the bombers not returning. He figured bullet density should be an even distribution across all bombers while statistical clumping would occur on individual bombers and thus those surviving actually showed where hits were tolerable. Instead of armoring where the holes were, he suggested, the bombers should be armored in the areas the surviving bombers lacked hits. Due to his assumption of even shot distribution, he hypothesized those not returning were the ones hit in more vital areas including the engines. Engines lacked holes on the survivors and proved to be the areas best to armor.
In the game Tigers and Goats, available in Ancient Games for those of us stuck at home looking for something to download, novice players playing the Goats often try to monopolize the corners. This tactic makes sense as the corners are absolute safe points, a Tiger cannot jump around corners to “eat” said Goat. Unfortunately it is a losing strategy as trying to occupy all four corners consumes Goats from the pool of Goats which could otherwise be used to support and defend other Goats on the board and reduce Tiger maneuvering space.
With these two thoughts shaping our thinking, turn to Covid-19. Covid-19 presents two resource challenges for us. One is test kits, the other is hospital capacity. Hospital capacity should be viewed in relation to normal load. This capacity can change slowly as more beds and ventilators get added or as normal load gets reduced as seen by delaying or canceling elective procedures.
The United States has suffered shortages of test kits for Covid-19 for several reasons. These include bureaucratic fumbles denying valid tests because they weren’t internally sourced, because lack of certification, and for privacy concerns. It includes waste in double testing. And it includes chasing a more perfect test rather than accepting good enough and quick. The logical conclusion to this has been a rationing of tests trying to apply them in the most effective manner. Unfortunately the mindset has rationed wrong. The mindset is like armoring where the bullet holes are on surviving aircraft. We’ve been testing likely ill persons rather than maybe ill persons. Likely ill persons should be assumed to be ill early in the game while excess hospital capacity is available. Maybe ill persons need to be identified and status determined quicker so as to prevent disease spread.
“This week, in Massachusetts, testing has been permitted only for patients with respiratory symptoms requiring hospitalization, or for patients with such symptoms who have traveled to endemic areas or had contact with confirmed coronavirus cases.” - March 14, 2020 National Public Radio
“Because of the scarce number of tests available in the U.S., doctors, upon the advice of the Centers for Disease Control (CDC), were following strict protocols for who could be tested: people who had symptoms and had traveled to a country where cases were endemic, or had been in close contact with someone who had tested positive.” - March 12, 2020 Time Magazine
““this is moving in the direction of having drive-through facilities at Walgreens, Walmart, and Target where you could drive up after getting an online approval. That means you would have a specific need for the test, which means you've got to have symptoms of some sort.” - March 17, 2020 The Atlantic
We’ve been thinking micro when we should have been thinking macro. Another way to put it is we’ve been tactical while we needed to be strategic. The immediate goal in early phases of infection spread should be reducing spread. If we accidentally use some excess hospital capacity, that is ok. It is not yet the limiting factor. If we send individuals home temporarily isolating them, again, that is ok. The loss in their societal involvement pales in comparison to another’s spread. Realizing disease spread is exponential, we shouldn’t test those we think likely sick until roughly forty percent of our hospital load excess above normal load is utilized(1) or until Test capacity has increased. It is at this forty percent point we need to start concerning ourselves with capacity preservation as we know capacity is about to be saturated. This is the point life saving ability becomes compromised. Until then, the asymptomatic and semi-symptomatic are our bigger risk as they’re likely not consuming hospital space but are participating in society. These are the ones who need to be flagged so as to curtail their contamination contribution. Our initial and current testing mindset would be good were we already in the second phase of concern with a goal of preserving life saving capacity. But in most areas we aren’t yet there. What we’re doing is like putting our Goats in the corners. We treat the obviously sick while testing to ensure they’re sick with what we think they are. This wastes potential to further support and enable others ceding opportunity to deny maneuvering space to the disease. Testing on likely persons whom could be isolated and treated based on symptoms and circumstance alone consumes limited tests that otherwise could have prevented maybe sick persons from exposing others. Think of it this way, if the doctor suspects you broke a rib, do you really need an x-ray to tell you this when treatment for bruised rib and broken rib will be the same? X-rays cost money, do you really want to pay for that?
Now some could argue that there is a third resource limitation in lack of personal protective equipment (PPE) for health care workers. This is valid and in this case, testing the likely sick may seem to help as likely sick found negative enables less health worker PPE use. Such thinking should be lauded as it is thinking like a planner while truly trying to find the critical factors. It doesn’t work though as likely sick are still sick with something to which PPE will still be required. And PPE is required for both likely sick and maybe sick until they’re found not so sick. Hence, this limited resource gets used no matter what. Unfortunately we really can’t improve this particular aspect outside of increased manufacturing. As it gets consumed either way, it is removed from the conversation regarding mindsets and test priorities.(2)
With test capacity starting to rapidly increase, the mindset will shift so as to more align to cover how phase one should have been. Though let’s not forget this lesson. And, as The Bay Area has instituted a lockdown, let’s recognize another lesson: lag effects often get misattributed. By this I mean that we should expect case numbers to rise in The Bay Area despite the lockdown and we shouldn’t fully see, appreciate, or assess the efficacy of the lockdown for a full three weeks after its imposition.
Notes:
(1) Hospital capacity should be assessed locally and in regions, different areas could be in phase one test the maybes simultaneously with other areas being further along into phase two test the likelies.
(2) This concern does, however, show five problems from our everyday mindsets which like the two mentioned above need to adapt to circumstances. One is that the general population has seized and hoarded some of this PPE. This should be eminent domained back to where it needs be. The mindset illustrated and in need of change is normal property rights. Private property concerns, similar to privacy concerns, needs to be tossed in the face of pandemic emergency. The Hot Zone showed us this with building and corporate owners preventing containment efforts. Two is on-demand supply chains and just in time logistics may be more economically efficient but they suck. Stores and reserves cost money to maintain but they gain hedge against risk. Risk needs to be mitigated and such expense should not be seen as unnecessary cost. In fact these costs should be mandated within essential services. We’re currently seeing how wrong the reduce all costs possible mindset is for society at large. Three, how we portray information matters. The Cobra Effect is real. Telling people not to take a limited resource encourages them to deplete the resource. Four, not every day is like the last. Five, not all aspects of new problems are different, the past still is instructive.
Thursday, Mar 19, 2020 · 10:42:18 PM +00:00
·
Fffflats
Excerpts from an OB/GYN physician via text chat below, but first, thoughts regarding the medical community.
- More and more specialties are going to be pulled to Covid-19. The OB/GYN’s workload, however, won’t decrease for at least a nine month lag effect. Perhaps the Orthopedics will recede quicker. We do, however, have in untapped resource - Veterinarians. Veterinarians should be pulled to help with the human crisis.
- The US Army often uses Veterinarians to fill in public health and medical planning roles. In fact, two such Veterinarians led the Ebola Reston clean-up and prevention efforts. This means that within their ranks we already have a pool of talent taught and practiced in viewing and working the Macro view rather than patient specific micro observations.
Doc: (Prior to reading the above article) Been saying all along we aren’t testing enough and our criteria for who gets a test is too restrictive, not helpful. Also dumb-founded the whole country is not on lock down. Schools out but everyone still at work, kids roaming around in big groups close contact like its extended spring break, sets up for disaster.
Doc: (after reading article and initial comments) xxx commenter is wrong here but following the directions we are stuck with.
Me: Yet somehow xxx fails to track PPE is irrelevant to testing choices and given the situation we can make better choices. I suspect he’s more representative to the medical community at large.
Doc: This is very true. Straight out of instruction by the CDC for whom to test. No one can or will go against that. Actually in many cases we can’t even order a test. A person completes a risk assessment then the state department of health gives test approval. They only approve those most likely to have it. Like the planes’ bullet holes, which is confirmation bias.(3)
Me: It’s a waste of a test.
Doc: I agree. However the decision makers, which is not me, are going to continue this.
Doc: This was from a midwife here after reading it [the above article]. Former public health background and losing her shit about how we are doing things.
Midwife: Agreed. Good article. A town in Italy tested all 3,000 of its residents. 50% were POS with no symptoms. That’s the group we need to identify to mitigate further spread. We know what’s coming... we’ve seen it play out in Spain and Italy. Sadly, these are going to be very hard lessons learned and healthcare providers are going to take the brunt of it as Americans “first string” team.
Doc: Did I mention they require a negative flu test to order a Covid? Meaning you did a wash which washed all the dna away first. Then you gotta wait, probably a day, to do the Covid test.
Me: And burned through at least two sets of PPE?
At a different point in time, the doctor made note of USNS hospital ships looking at being deployed to major US urban areas. With this in mind:
Me: Did anything come of this planning (or lack of planning) effort?
Doc: Nope
Doc: Wheels still spinning and knee jerks. Plus political impositions like sending the Comfort into New York Harbor.
Me: They can send it to harbor but New York has to request it for it to be able to provide assistance. Seems stupid to me when SeaBees could just pop up tent cities. More capacity, more space to move around and work, less risk transferring people around... That’s what Cuomo asked for.
Doc: At the midwife’s behest they looked at the EMF (tent hospital) supplies to raid from them for here. All the shit was moldy.
Me: Funny, you’d think tents and vulnerable supplies would be stored in the desert alongside plane boneyards.
Me: Seen the Florida beaches?
Doc: Yep
Me: FWD from a person with ties to South Asia: There is a million plus religious festival in India. Organizers say Ram will protect us. Modi is afraid of aggravating his religious base, so continue as planned. That is Game Over!
Doc: Had been wondering about India. They are never mentioned.
Note: (3) I’d argue surviving bomber sampling is actually a sampling error at the beginning data collection portion of an analytical process as it accidentally samples a unique subset while disregarding the whole population of bombers rather than a confirmation bias inserted during data processing inside the actual analysis portion of process.