Model Assumptions Driving Action or Fear?
We are facing a ticking clock and each moment we pause because of politics, because of the devolution of human communication on social media forums and the polarization of our country in general — will literally cost us lives. I don’t care who’s fault it was, I don’t care what happened to get us here — we need to start rapidly testing (if CDC cannot provide push to public labs, or private labs, or the startups across the country that seem to be able to do this) — and we need to implement wide scale distancing by closing down the US for a few weeks.
Now even saying that — I will say there has been a push by many to leverage data in reporting in a manner to engender action by individuals and government that I believe actually causes individuals to feel panicked and that any efforts they may make are futile. I have seen a growing over-confidence in data and modeling around the Coronavirus that I think is starting to just freak people out without providing enough context to allow them to understand how the numbers work or the assumptions therein and I do want to be a voice of reason when it comes to some of these numbers.
This modeling feels super scary and real to people who just see the numbers and potential outcomes. But like any model it’s wrong in its prediction — and there are a bunch of ways it could be significantly wrong.
- We are making huge assumptions on every key data point without information based on case severity and lack of testing.
First we need to have a good estimate of the Total amount of cases of COVID in order to make projections about spread and about eventual impacts and we really don’t have that info.
Why? We know many have mild symptoms and are unaware — not everyone who is infected has been treated and not included in data modeling.
Plus we are idiots that can’t FedEx testing kits because our system is broken — many countries (third world countries like us) struggle to test a large number of cases.
2. I don’t know where the biggest assumption comes from outside of it was said as a scary big number by a Harvard Profession.
Like many others this article is using a number of 40–70% of population gets the virus — which came from Marc Lipsitch at Harvard in early Fed.
Lipstick said — if it is a pandemic — 40% to 70% of people world-wide could be infection in the coming year.
This is based on straight ongoing exponential doubling and what I would say is I don’t think that number gels with the data out of Wuhan, Italy or any other Country that has any scaled testing and distancing where rates slow and have actually stopped (yay for interventions). We know this from simple facts — It is obvious that 40–70% of China’s pollution would be significantly more that the 80,8015 people that currently are infected.
If China has slowed and only has 80k cases total or 100–120k — this 40% — 70% is WAY overestimated.
R0 — measures the vitality or spread (replication) and is a fluid thing, not a defined characteristic of a virus So in a country like South Korea where they’ve slowed the spread of the virus through social distancing measures, it’ll be lower than somewhere that didn’t act until it was too late.
This is the flatten the curve concept that really should be what everyone is talking about now.
So the best way to see the massive impact this has not only on overrunning hospitals but also deaths is this interactive |
Hopefully that gave you some relief — It is super real, and super scary — BUT — it’s not too late and it’s not necessarily that bad if we pull our heads out of our assess and get to work.
TO MODEL OUT A BAD OUTCOME lets assume, as was stated, 40–70% of the population contracts the disease (again I think this is preventable) bbbut if so the range of potentially infected people would be 138,880,000–229,040,000 infected in the US.
Now take the active infections and assume that 20% of cases are mild and don’t require testing much less hospitalization. (Only 16% of cases in China were labeled serious).
Subtract mild + perhaps untested cases and we are left with |
Total Serious Cases
111,104,000–183,232,000
Total Critical Cases — 5% — 15%
These are people requiring ventilation and is a number that should have Trump whipping out the checkbook to help the health care system get prepared ASAP, instead of bailing out the finance industry again.
Total number of of US Critical Cases — 5,555,200–9,161,600
Finally what is the CFR? — right now its high, according to the WHO 3.5% or so — but it will vary significantly. Case fatality rate (CFR) is time- & population-varying. Given that mild cases are generally identified at a slower pace than severe cases AND that care standards (& demographics) vary across affected countries, responsible discussions about CFR really should include context. otherwise it scares the shit out of reasonable people.
Note - In South Korea it is leveling at .8% and many think there are lots of mild or asymptomatic cases here. as well so it could be as low as .5 or .3. Is that likely — perhaps as likely as it will be 3.4% globally so we might as well evaluate those numbers as well.
Starting with .8% — 888,832 fatalities
Moving to .3% 333,312 fatalities
So this is the horrible outcome we are looking at without government and local interventions and it may be worse as hospital systems strain under the pressure of this patient population. Even if this happens, we will get through it I do think we will manage — but it will be sad, horrible, “never going back to 100% normal” bad.
However -If we intervene mildly now (which we are doing) and lock peak infections globally at 3.4 million The US CFR goes down to 320,000.
And If we enact aggressive measures now — it reduces peak infection to 533,000 and deaths to 53,000!
All of it is terrifying — even that we have to consider these numbers is hard to do — but I hope the context and understanding the assumptions behind the model provides you with the same feeling it gave me — there is hope, we will get through this, we need to fight and yell and demand our government act. The two things you can tell everyone is yell at everyone in government or Twitter or wherever you like to yell about this stuff — 1. to get our fucking tests in market and 2 — let’s push enact strict interventions to stop the spread.
