Three Models that Inform Why We’re Taking Drastic Action Today
Wow. The number of articles, posts, and comments that I’ve seen out there are overwhelming. There’s a lot of noise and a lot of confusion about what is really happening and WHY our state and national governments are keeping us home with ever increasing constraints. There’s one report that summarizes it all but is sort of hard to read. There are many people who have been summarizing, extrapolating and sharing information and I’m going to do my part to help in that process too. My goal is to help inform why we’re taking actions, and not to question the validity of the choices, or to compare point and counterpoint all of the things wrong with it. Plenty of experts are likely to do so and dissect this for years to come.
The “study” by the Imperial College in London took data that came out of China & Korea and began rigorous data modeling with those inputs. There were enough cases and data to do that and we’ve had past models to lean upon as well. According to the New York Times, the CDC is taking this seriously and you can see how it is currently informing both federal and state actions. I’ve attempted to summarize and reframe the three models, various people’s explanations of them and their forecasted impact for ease of understanding. [Original Report]
Model 1 - NO ACTION: Treat this like the flu.
The math goes as follows. 80% of Americans get infected. .9% would die. 4-8% of seniors > 70 years old would die. Death count: 2.2 Million.
There is a cohort that will reach the severe stage of this virus of this will need ventilators. Without ventilators, they will die. With them, 50% will die. The other 50% will spend 2-3 weeks on those ventilators. The “survivors” will be immunocompromised and susceptible to a whole host of other complications and will have subsequent death rate (probably not accounted for in the virus’s death count).
Since the demand for ventilators would be far greater than the supply (30x greater), and since 100% of patients who need a ventilator and can’t get one would die, the actual impact would be higher. In 3 months, the virus could adversely affect seniors > 70 yrs old (both of my parents are in this demographic). Death count: > 2.2 Million
Model 2 - Partial Action: Social Distancing for Seniors & Case Isolation.
Quarantine anyone symptomatic for a minimum of 14 days.. The goal here is to keep the most “at risk” segment of the population out of the mainstream and move anyone who is suspicious out of the mainstream immediately. Let's call this the “do something” action.
This model slows down infection rates, and spreads it out over time. It’s a moderate version of the “flatten the curve” mantra and the initial death count expectation is 1.1 Million.
We will see the need for ventilators drop by 66%. However we’ll still need 8x as many as we have. Again, most people who need them will not get one.
Final estimates after you take the natural dispersion model, then account for ventilator capacity are greater. Death count: > 1.1 Million
Model 3 - Full Action: Social Distancing, Case Isolation, Quarantine & Shutdowns, the full suppression strategy.
In this model there is a significant improvement in all of the outcomes
The impact of the virus peaks in the next few weeks. We’ll still need all of those ventilators but demand for them will roughly equal supply.
We wouldn’t have the huge impacts we see in model 1 or model 2. We might look back at this and think of this as just another bad flu. This model seems like a reason to celebrate and you can see why its what we’re doing. Death count: Several Thousand (let’s say 2-10K)
BUT… THIS IS THE IMPORTANT PART, quoted from the publication,
““The major challenge of suppression [Model 3] is that this type of intensive intervention package – or something equivalently effective at reducing transmission – will need to be maintained until a vaccine becomes available (potentially 18 months or more) – given that we predict that transmission will quickly rebound if interventions are relaxed.””
This may be our way of life for 12-18 months!
Figure 1: Demand for ICU beds for each of the three models
- Imperial College COVID-19 Response Team
The gray area is the period of action and effect of each model
You can see the bounce back period if we call victory too soon.
Figure 2: On & Off again model for Model 3
- Imperial College COVID-19 Response Team
Turning model 3 on and off periodically
Shows we’ve got a long ways to go
Countries have already tried all three approaches. The data was plugged into the model. We have a sense of the consequences. Word on the street in Italy, who is ahead of us in this process, is that they went through all three phases and are now in an ever-tightening version of Model 3. It’s possible they needed to go through the three phases to get people slowly used to the change. Yes, they have more seniors, and population demographics vary by location.
I think its easy to look at the numbers and react in a few ways that go like this. “Oh, I’ll have a minor reaction, it won’t hurt me.” or “these numbers are full of holes”, or “how can they really project these figures, its all hypothetical.” But consider the ICU capacity in this country. If it is completely occupied by the at risk populations, I pray that you or a loved one doesn’t get in a car accident, have a traumatic injury at work or home, have a heart attack, or encounter some other one of the millions of viruses & diseases out there. There won’t be any beds for you.
Decision making when faced with the unknown or ambiguous is challenging. Those of us who do it for a living start getting comfortable with making big, difficult decisions with these unknowns. I’ve learned over time that being directionally accurate is better than facing the consequences of inaction while we await perfection. We can probably poke holes in the models; they aren’t perfect and death isn’t patiently waiting for us to achieve perfection. Politics and our own fears are building momentum, suggesting that we’re overreacting. If we’re successful, it might certainly appear that way in retrospect. That’s precisely because we avoided the catastrophe. Listen to the scientists and heroes on the front lines who are looking out for all of us. There’s only one tribe. The human tribe.
Notes: Again, i’m not a scientist, economist, or frankly, even a writer. I left my job in October and am not working full time right now, and know that many of will not be for the foreseeable future. I have time on my hand and writing about, thinking about, and planning for scenarios seems like a decent use of my time.
The Imperial College Report, CDC & Other Stats:
The report,was not released in a peer-reviewed journal but was authored by 30 scientists on behalf of Imperial College’s coronavirus response team and it simulated the role of public health measures aimed at reducing contact. This was done because of the need to inform action.
New England Complex Systems Institute has a critique written about this research found. The fact is, it doesn’t much matter because the actions taken across the globe are tied to the Imperial College report.
We estimated the case-fatality risk for 2019 novel coronavirus disease cases in China (3.5%); China, excluding Hubei Province (0.8%); 82 countries, territories, and areas (4.2%); and on a cruise ship (0.6%). Lower estimates might be closest to the true value, but a broad range of 0.25%–3.0% probably should be considered - CDC
Journalistic Reports
The Virus & ICU Capacity
3 charts that helped change coronavirus policy in the UK and US
Peter Attia M.D. has a podcast called “the Drive” which does an outstanding job getting technical, but relatable. He has a couple of podcasts episodes on the drive and I did get some insights on ventilator effects from his last episode on the Covid-19.
A back of the envelope calculation assuming …
New York state has 3,000 ICU beds at full capacity (and assuming no one else needs ICU bed that is not a coronavirus patient; a generous assumption)
Taking a reported 421 diagnosed cases in the state, a 1.3x growth rate, 20% hospitalization (assuming all ICU bed space)…
X= 13.6 days to ICU bed failure (critical capacity)