Actual vs Confirmed Cases

This Medium article by Tomas Pueyo does an amazingly clear job of describing what's happening now (in reality) vs what is being report (confirmed cases). I will copy the most important graph below so you can see it, but please read the article.

This shows with retrospective clarity how many COVID-19 cases in Hubei were being confirmed vs retrospectively cases that were actually happening. You can see when the first cases were confirmed (the first orange bars) the spread was already much, much more. And only by acting quickly to stop the spread when it seemed small did they create the reduction that was confirmed two weeks later.

I'm very happy to see that my state of Michigan has decided to cancel schools statewide for three weeks as has Ohio. I appreciate the leadership this takes as I see local school districts saying they will hold classes because no confirmed cases exist. The cancelation of large evens like March Madness, St Patrick's day parade and other ongoing sporting events is also significant. Every day we are closer to beginning the actual decline in spread that we won't see for a few weeks. 

Venilators

We have 45,000 intensive care beds in the US and we might need 200,000 in a moderate outbreak. There are about 160,000 ventilators in the US. In a moderate pandemic it’s estimated we would need 65,000 ventilators and for a severe pandemic we would need over 742,000. 

The range is staggering. Makes you think if things get bad, we are in short supply of ventilators and that’s what people need when it gets serious for COVID19.

There’s a page on hacking an open source ventilator. Comments on the page suggest the description for what’s needed is wrong. I’m not sure this is the right answer, but the question is coming up more and more. Do we have enough?

Coronavirus - Italy vs South Korea

The difference isn’t in the spread. The trajectories of confirmed cases look similar. It has slowed in South Korea around 7,500 confirmed cases and continued to grow in Italy. But the spread looks similar.

The difference is in the death rate. The number of people who have died in Italy is 10x the number in South Korea. The death rate has been highly variable across the globe. Singapore has had low infection rates and no deaths. In Japan the death rate is reported around 2%. Why? What’s different?

South Korea is smaller at 50M vs 60M people. South Korea has had no regions locked down and 29k people in quarantine. Italy has restricted the entire country and closed restaurants, stores, everything.

South Korea also has a lot more hospital beds - 11.5 per thousand people, compared to 3.6 per thousand in Italy. Singapore has just over two hospital beds per thousand people and Japan has over 13 beds per thousand.

Testing is very different. It is reported that 222k people tested in Korea vs 72k tests in Italy. The ramp up in testing makes the scope of the problem more clear. It also enables the use of the data available to manage those infected.

South Korea has much more access to tracking data. Similar to Singapore they use all sorts of data . GPS, phone, credit card transaction data. Quarantine officers track data and ensure compliance.

Korea is also managing the pipeline of patients effectively. They are using their healthcare resources as effectively as possible. When a bed is available, the patient is notified and picked up via ambulance.

Is it the availability of care (beds), the efficacy of deploying those beds (pipeline) or the widespread testing and effective tracking (reducing the spread)?

Coronavirus Thoughts on March 12, 2020

Hospitals in Italy are overwhelmed at >12,000 confirmed cases. In Italy they have 3.6 hospital beds per 1,000 people meaning that they have 36,000 hospital beds for the 10M people in Lombardy. This puts the ratio of confirmed cases to beds at 35% and they’re saturated. It makes sense if you consider the likely localization (it isn’t uniform across the geography) and there are existing people needing those beds for non-COVID19 treatment (the hospitals aren’t empty waiting for this to happen).

Extrapolating to my home state of Michigan, we have a similar population to Lombardy (10M people) and fewer beds (25,375) in the state. Using the ratios from Italy and being more generous that maybe we can use 50% of capacity before we’re overwhelmed, we can handle 12,687 confirmed cases of COVID19 in Michigan.  In my four county area considered one “hospital group” we have 1,043 beds. That means locally we can only handle 520 confirmed cases with our bed capacity.

We have been talking about social distancing and other measures to slow the spread and flatten the curve. The curve of cases is compared with our capacity to handle them. It needs to be viewed somewhat locally. The curve for the US isn’t as helpful as the curve for Michigan or even the local hospital group that will treat a population.

As I think about saturating our hospital bed capacity, it dawns on me that there will be impact on others that need treatment and expect capacity to be available. There are people who will die because they were displaced by COVID-19 even though they don’t die from COVID-19.

Michigan has two confirmed cases, both in Detroit counties. We have one person who needs to be tested in Portage that has visited a big local employer, Stryker. Our testing has been very limited and large events have continued to be held up to this week, so the number of actual cases is probably larger. 

By next week I expect we’ll know the number is closer to twenty confirmed cases. The data show that once you detect one and start looking the number seems to grow quickly in a few days. Once it settles, the doubling time is estimated at seven days based on data from Wuhan. For Michigan, that means we’ll hit hospital bed saturation in eight weeks (mid-May).

I’m working on thinking through the impact of this future scenario and what that means I need to do. For what do I need to prepare to live in a place that looks like Italy does now. There are no available hospitals beds, we are probably restricted from free movement. It probably isn’t buy more toilet paper.