Daily Covid-Insights and Charts — March 26
March 25th was a brutal day for most countries by the number of cases, number of deaths and the death growth rate. However, the case growth rate is still trending down for most hardly hit countries, which is the only glimmer of hope in today’s edition. Please remember that the growth rate trending down does not mean we are past the peak of the curve, only that we might be getting closer to it.
Let’s start with today’s charts. My insights are after the charts.
Here are cases vs deaths.
And here’s total cases.
New today, I’ve also added a total fatality count chart similar to the total cases chart.
And the 7-Day Compounded Daily Growth Rate for Cases and Deaths.
1) The US had over 13,963 new cases and 249 deaths, both new highs. The case count is again highest for any country in a single day on record.
2) The US’s fatality curve is above China’s for a second day now. Here’s another way to say this: People are dying faster of Covid-19 in the US than they did in China. China’s population is over 1.4 billion, the US 330 million. US’s death curve will likely pass Iran’s in a day or two and so far seems to be shadowing Italy’s death rate, which is extremely concerning as Italy has the most deaths on record of any nation and the steepest death curve.
3) The death growth rate is increasing for most countries, which is expected as fatality is a laggard indicator. Since the case growth rate has been trending down, we should expect the death rate to follow suit at some point as it has in China and South Korea.
4) Canada had a significant spike in cases yesterday. As I’d mentioned last two days, ECDC, where I get my data from, was not counting all cases due to Quebec’s new methodology, but now they are. Canada has 3385 total cases as of yesterday and eight further deaths, a new high for Canada. Canada’s case curve has now passed the UK’s. It seems to be heading for France’s curve next, which is one of the hardest-hit countries. Please note that Canada is 11 days behind France, so we still have time to turn this around.
5) France’s death curve passed China’s yesterday as well.
A top question in everyone’s minds is when to start easing distancing and lockdown measures that are having a devastating impact on the economy.
Here is my updated back of the envelope estimation.
Based on how China and South Korea contained the virus and the World Health Organization’s guidance, the current consensus on a winning strategy seems to be: Lockdown, distancing, test, trace, isolate, treat. Countries like South Korea that were able to act fast, have been able to avoid lockdowns, but once you’ve hit higher death growth rates with critical mass of fatalities — say hundreds — lockdowns become inevitable (South Korea’s total death count is 131. US is at 1050).
China’s 7-day growth rate peaked around day 9/10. And they started talking about easing lockdown measures in Wuham about 60 days later.
We could use that 60 days as one guidance to estimate easing of lockdown measures for countries with high death growth rates like Italy, the US, Spain and France.
So, for instance, if we assume that yesterday’s 7-day death growth rate was the peak for the US, could be about another 60 days. However, that assumes that the US will do as good of a job China was able to do in all dimensions of that strategy during those 60 days: lockdown, distancing, test, trace, isolate, treat. Overall consensus seems to be that the US is not executing that strategy as well as China. So if we assume a 50–60% degradation in degradation, that brings us to 90–100 days.
90–100 days would take us to the end of June. Interestingly enough, California’s governor Gavin Newsom just gave similar guidance (source) — I’m sure based on infinitely more sophisticated models, but it adds up for me.
Note: Sorry for the late release today. ECDC was late releasing their data, which set me back a couple of hours.
Note 2: Updated the death rate analysis a couple of hours after posting this in case you read it then. I caught a Google sheet oversight on my part which has now been corrected.
