Based on the daily growth rate and cohort analysis I shared yesterday, I have also started tracking the 7-day compounded daily growth rate each country. So essentially, how case growth is trending for the last week at each country. I think it’s a better indicator of how the country is doing in it’s fight against Covid-19, and might be a useful leading indicator on what’s ahead.
- US added 8459 new cases vs. 7123 the day before. The highest for the US or any other country in a single day. US had broken that record yesterday as well. The 7-day Daily Growth Rate has increased to 37.58%. At the moment, US is the only western country with a daily growth rate that’s actually growing which tells us that we haven’t passed the peak yet for the US and more challenging times are ahead.
- 5560 new cases in Italy, about 1000 less than the day before. One of the deaths included a 34 year old, otherwise healthy woman. As I mentioned yesterday, there is a silver lining with Italy: their daily growth rate is declining steadily. A lot of their shutdown efforts, albeit a little late, seem to be starting to pay off. Their 7-day growth rate is down to 13.76% from a peak of 44% just under three weeks ago. China was around 13–14% growth rate about 40 days ago and South Korea about two weeks ago.
- 65 days after 150 cases, China has announced that they are easing some of the lockdown in Wuhan as that region has not had a new case in 4 days. However there are conflicting stories as to how real that is as China does not count anyone with mild symptoms that may test positive as a case (source). Hong Kong has also had a second surge in cases and have increased lockdown measures, travel restrictions and extended their school closures indefinitely which were scheduled to reopen in late April (source, source). I think this is the “Dance” phase that Tomas Pueyo talks about in his fantastic piece which is a must-read.
- Canada had 199 new cases vs 260 the day before. Both in terms of total cases and the daily growth rate, Canada is managing to keep below the curves of every other western nation. However, still early for Canada as we are 8 days behind the US and 18 days behind Italy.
Here is the daily chart of new cases for some of the key countries I’ve been tracking.
US is above China’s curve two days in a row, and Italy seems to be heading there too, however this chart tells only part of the story.
Here is the trailing 7-Day daily growth rate for each country once they passed 150 cases. I’ve included Germany and Spain in this chart and used datawrapper.de which makes interactive.
What definitely stands out is that US is the only country whose growth rate is still increasing, however, most other countries have climbed up a similar hill before coming down. Spain, China and South Korea all had daily growth rates of over 40% in their early days. The big question that many people that are a lot smarter and qualified than myself are trying to figure out is how long will that take for the US.
A very rough, back of the envelope estimate at this could be to look at China, and how long it took them to go from 38% daily growth rate to sub 1% they’re at now — the answer is just about a month. But key thing to remember is that China completely shut down a 60 million person region when they had less than 1000 cases (US has over 36,000 now) and they were on day 6 (US is on Day 19). It’s also fair to mention that the Chinese government was able to utilize many containment methods that are not as readily available to the US government.
As a western democracy, Italy might be a better example. Despite the tragedy in Italy, they’ve been able to reduce their 7-day daily growth rate from 35% to below 14% in 21 days. If we assume another 20–25 days to go from 14% to sub 1%, that’s about a month and a half.
So if US went on an Italy/China style shutdown, AND the US daily case growth rate started decreasing tomorrow, can we expect for the virus to be subdued by early June? I think that would be the most optimistic case based on some very optimistic assumptions.
There are many problems with this back of the envelope estimation but the biggest problem is that it assumes the growth rate will start slowing down tomorrow. We definitely do not yet see much evidence to believe that.
Higher the growth rate climbs, and later in the cycle it peaks, more deadly and costly it’ll be to come down.
PS: I hope to find a way to normalize the cumulative cases metric I’ve been using as it’s very susceptible to test rates. I’ve had some great suggestions such as # of tests conducted or icu/death rates or even per capita. I’m not crazy about per capita as it gives a false sense of security for larger populations and a false sense of panic for the smallest ones.