Daily Covid-Insights and Charts — March 24
Today’s chart is new and improved based on some of the excellent feedback I’ve received and includes death rates for each country. This chart can now provide us with a window into how a country is doing in identifying it’s infected people, how many infected people they’ve identified, and how well they are doing caring for those people. Please note that the left vertical axis is cases (straight lines), and the right vertical axis is the death count (dotted lines). Also note left axis, cases, goes from 1 to 100,000 and the right axes, deaths, goes from 1 to 10,000. Multi-axes charts can be confusing, and please let me know of any questions. (If curious about the details of changes see the end of the post).
1) US’s surge in new cases continues with its largest-ever tally once again with over 11,236 new cases yesterday — also the largest for any country in a single day. The 7-day CDGR (Compounded Daily Growth Rate) increased again for the US, from 37.58% to 38.42%. The US is on Day 20 of past 150 cases. No other major country had such a high growth rate on their Day 20 — including some of the hardest-hit countries like China, Italy, Spain or Germany (see 7-day CDGR chart below).
2) The US recorded 119 deaths yesterday vs 131 the day before. It is now a sliver away from crossing China’s death curve, which will likely happen in the next few days. As you may remember, I had predicted that the US would exceed China’s cases curve as well, which occurred a few days ago. Given the lag between infections and a sick patient dying, the lag in between the two curves is expected. The silver lining at the moment is that the US might be tracking China’s death curve more closely than Italy and Spain, which have had some of the highest death rates. Amesh Adalja, an Infectious disease specialist from John Hopkins, was expecting lower death rates in the US due to a higher level of emergency and intensive care — I am hopeful (source). However, worth noting that the US (and Canada) have some of the lowest numbers of hospital beds per capita among the developed nations (source).
3) Italy’s high death rate (almost 10%) is clearly visible here as you can see how close their case and death curves are to each other. This is a heartbreaking tragedy. We can see similar behaviour with Spain. The reasons behind the higher death rates are beyond my expertise, but please let me know if you’ve found useful resources on the reasons why. I have to assume age-mix, health system strain and testing availability to be some of the contributing factors.
4) Italy has had the second day of declining cases (4789 vs 5560) and second day of declining deaths as well (601 vs 649). Hoping that was the peak.
5) There’s some drama in Canada’s reporting and stems from Quebec (surprise!). Canada had a surge in new cases with 621 new cases vs 199 the day before. However, this surge is coming from the province of Quebec, where they changed their reporting methodology of tests and now include what they used to call “probable” cases and were not adding in those cases. Here’s the official explanation:
“The significant increase in confirmed cases is explained by the fact that on March 22, 2020, cases tested positive by hospital laboratories are now considered confirmed. They no longer need validation by the Laboratoire de santé publique du Québec (LSPQ).”
It sounds to me like these were 400 cases that were in Quebec all along, and we are taking a one time hit. Reminds me of the 15,000 case day in China when they came clean with a bunch of new cases in one day. If it weren’t for these 400 cases, Canada would have had a flat day. However, this also explains why Canada has been tracking so below the other countries. C’est la vie.
Here’s the cases chart with Canada, where we are still tracking below but with an uptick to account for yesterday. I have not included Canada in the new chart above as our death count is too low.
6) The chart for 7-day compounded daily growth rate (CDGR) for cases is below. The critical thing from yesterday is that the US infection growth rate keeps increasing.
I’ve also added the same growth rate for deaths (as requested by readers). There seems to be an upward trend for US and Germany and downward trend for China, Italy, South Korea and Spain. I think it’s too early to tell for Canada.
A common question I get is, “When will the US peak?”. And I understand what’s behind that question — people want to know when this will be over. I can’t tell when the US will peak, but the mathematical fact is the growth rate needs to start decreasing before we can talk about a peak in sight. What the US will do over the next few weeks to control and limit the spread of this virus will surely have the most significant impact on when the peak arrives and when it all ends.
PS: Regarding the changes to the chart: As more people have started finding these daily updates useful, I’ve gotten some exceptional feedback from some brilliant people.
One of the top comments I’ve received has been whether the surge in new cases in the US is simply because of increased testing. Followed up with the suggestion to normalize the data with the number of tests conducted or to add death count (and ICU admissions) to provide a more complete and balanced view of how a country is doing in her fight against Covid-19.
Unfortunately, I was not able to find a reliable and complete data source for tests conducted. Three different sources (1,2,3) I discovered were all out of date, incomplete, and none were daily, which would not work with a daily tracking chart. I was also surprised to find out that China stopped reporting that metric in late February, which would leave a pretty large gap in the analysis. Any suggestions on how to incorporate this data, please DM me — I’d love to find a way.
However, death counts are consistently reported, which I was able to add to the chart. The idea behind using death counts is that death counts don’t care if the victim was tested or not. It is a more tangible measure of how the virus is progressing through a country. One problem is that, due to insufficient testing or other procedural reasons, the death cause may not be recorded as Covid. There were concerns about this in Japan due to the low number of tests they’ve done.
But the argument goes, if the surge in infections in the US was simply due to increased testing, we shouldn’t have as many deaths or the death line should be close to flat? The chart seems to show it’s not flat and that it’s on par with how China was tracking and might even exceed it.
I’m going to look for a downloadable data source for ICU rates, which I think will help as well. Please let me know if you know of any sources.
Raw Data: European Centre for Disease Prevention and Control
