COVID-19 Through A Different LensReading time: 8 minutes
COVID-19 continues to impact the world and I have become increasingly absorbed in the data surrounding it. Recently, Mike wrote a blog piece about the projected impact of COVID-19. This week I’ll take a look at the data through a few different lenses to see what it might be telling us.
The previous blog was published on Friday, 3/20/20. At that point in time there were 272,167 international confirmed cases. As of today (03/30/20) there are 735,560 confirmed cases according to Johns Hopkins. That is a 170% worldwide growth rate over 10 days, or a daily growth rate of about 17%.
In the United States, we’ve seen quite a bit of fluctuation in the growth rate over the past weeks, with an average daily growth rate of 31%. The graph below is likely one most of you are familiar with – the total number of confirmed cases in the US, which is rapidly increasing.
However, a raw number of cases can only tell us so much. We can assume that the actions that the government is taking have a delayed effect and won’t be immediately visible in the data. Later in the post, I will be comparing the US’s numbers against that of other countries from two different perspectives to see how their approach to addressing the pandemic impacts its rate of growth.
The two alternative views we’ll use are a per capita view (number of cases per 1 million residents) and a view that demonstrates the growth rate since the day a country hit 100 confirmed cases. This will normalize the data to make for better comparisons with countries that have different population sizes and who are further along in dealing with the crisis. Here is how the data looks from those two perspectives for the United States:
While the number of total cases is increasing, the trend in daily growth rate faces downwards. Studying the data from these different perspectives can tell us a lot about what measures are effective and how best to approach the pandemic. First, let’s look at an approach I assume most of you reading this are familiar with.
Flattening the Curve through Social Distancing
Social distancing and community isolation are measures that can have a very real impact on the infection rate. Put simply, there are a finite number of resources available (hospital beds, ventilators, medical staff, etc.) and the infection rate indicates growth in a very short term beyond these resources.
To examine this further, let’s take a look at one of the countries hit hardest by this pandemic, Italy, and how social distancing impacted the growth rate of COVID-19 there.
Comparing the Data: Italy vs the US
As of March 25th, Italy had the highest number of active coronavirus cases with 57,521. Italy reached 100 confirmed cases back on February 23rd, 2020. The United States reached 100 cases on March 2nd, just over one week later. If we look at the three weeks following those start dates, Italy averaged a daily increase in confirmed cases of 28% while the United States is averaging a daily increase of 34% (data is from 2/23/20-3/14/20 for Italy and 3/2/20-3/22/20 for the US). While the US had fewer cases until 3/27/20, it continues to show a much higher rate of growth, implying that we will far exceed Italy’s case count.
Despite continually increasing numbers, we have seen Italy start to minimalize the rate of growth. Over the past seven days Italy has reduced the growth rate to 7.44%.
In contrast, the US has seen an average growth rate during the past seven days higher than 23%.
For another means of comparison, let’s now see how the total number of cases look through a per-capita lens, meaning the number of confirmed cases per 1 million residents:
While the US growth rate is consistently higher, by looking at the cases per 1 million residents we can see that the US rate reflects a much smaller percentage of the population. However, since the US saw its 100th case more than a week after Italy, we can assume that there will be a delay reflected in the data as well. If we normalize the timeline to look at the number of days since the 100th confirmed case, this might give us a better picture of how the trends intersect.
While we see quite a bit of variation day-to-day, we can hope that the US trend continues similar to what Italy is now seeing. However, since the US has given individual states the authority to determine how to handle the situation, we are likely to see quite a bit of variation day to day in the US.
In Italy, we’ve seen the impact that a sudden spike of patients admitted to the hospital with COVID-19 symptoms has had. The big question is how best to handle this situation so that hospitals in the US are not similarly overwhelmed. Italy implemented strict social distancing and self-quarantine procedures, and this appears to have controlled their rate of growth.
In addition to discussing “flattening the curve” and “social distancing”, I would like to suggest we also start saying “we need more tests”. I realize it isn’t as catchy a phrase, but I will explain why it’s important.
The Impact of Testing
Some countries have taken a dramatic approach to testing their population. Studies from both Iceland and Italy have found that a large percentage of those confirmed to have coronavirus show no symptoms. Iceland, who arguably has tested the largest percentage of their population, have found that 50% of carriers show no symptoms. A study in Italy of the town Vo Euganeo, found similar results. The individuals that are asymptomatic or only have mild symptoms now appear to be a larger contributing factor to the spread of the virus than previously thought.
What this means is that countries and regions that took an aggressive approach to testing have seen a significant reduction in confirmed cases. That may be why on March 16th, the World Health Organization tweeted:
“We have a simple message for all countries:
Test every suspected #COVID19 case.
— World Health Organization (WHO) (@WHO) March 16, 2020
Since new research suggests that a high percentage of individuals are either asymptomatic or have mild symptoms, it can be assumed that a higher rate of testing might help to contain the spread of the virus. The data seem to suggest that as well.
The country that has been, arguably, most aggressive when it comes to testing, especially per capita, is South Korea. Let’s take another look at the growth rate charts, but include South Korea in the view.
As you can see, South Korea has largely leveled its daily growth rate.
So how have these growth rates impacted these countries? The story can change depending on how you look at the data. For instance, as of today, the US has 130,000 more total cases of Coronavirus than South Korea. However, the US population is over seven times larger than South Korea. If we look at the cases per 1 million residents, we see a different picture.
On March 25th, the US reached the same level of infection based on population size as South Korea. South Korea has been praised for how they have controlled the situation, and it’s a good sign for the US that our level of infection is not seven times higher than South Korea’s. But again, the data can be misleading. If we adjust the timelines to look at the days since the 100th confirmed case, we see yet another story. The US reached a higher infection rate than South Korea, 14 days earlier in the spread of the virus:
Let’s look at a comparison of the US, Italy, and South Korea through the three different lenses. First, let’s look at the raw numbers, looking only at total cases and the date each country arrived at that number.
Now, let’s take a more controlled look at those numbers. Here is the same data, per million residents, since each country reached 100 confirmed cases:
Next, let’s look at the daily growth rate since each country hit 100 confirmed cases:
The downward trends in Italy and South Korea’s growth rate seem to indicate that social distancing and mass testing help contain the spread of COVID-19. As we examine this information, though, it’s important to be wary and avoid drawing conclusions that are too strong. This situation is rapidly developing, and the lack of mass testing in most countries means we don’t have a completely accurate idea of how many active cases there truly are.
In order to provide different perspectives on COVID-19 data, I put together the dashboard below. I will continue to update this with new data as soon as it is available.