A New View of COVID-19 Case Volumes in the US

First, many thanks to those of you who have read, liked, commented and expressed interest in my prior posts on the topic. I am totally amazed, disappointed and disturbed by how something so prevalent (according to National Geographic article on 15 April 2020, the number of individual viruses on the planet is 100 million times more than stars in the universe) but, normally so innocuous (only about 200 cause disease in humans), can totally disrupt our lives, economy, politics, society – everything. Selfishly, I am at least glad the topic has some relevant, current, and seemingly accurate data available to create a welcome diversion as I have adjusted to retirement from American Express – and until football season gets going again. 

So, in the interest of full disclosure, I am providing below an up-to-date view of the growth rate charts I produced previously, and will then go on to show a new view of case volumes that probably better help in understanding the spread of the virus at this point of the pandemic. As is clear in the charts below, all curves of growth rates based on total confirmed cases appear to converge somewhere in the 60-75 day range whether we look at countries across the globe or geographic regions in the US.  

• Data for country-level charts from the “full_data.csv” file produced by the “Our World in Data” organization updated thru 7 June 2020 

• Data for  US charts from the “us-counties.csv” file published in the NYTimes repository on GitHub updated thru 6 June 2020

Changing the scale of the y-axis does bring out some differences but has the drawback of hiding the beginning of the story which can still provide important context for volume levels at this point. For instance, the chart below has the same data as the US chart above but, the range for the y-axis has been truncated to 100-200%. Each region defined here is now between 85-96 days since crossing 100 confirmed cases (recall, that is how day 0 is set) – and you could draw the conclusion that the WEST (comprised of CA-OR-WA-AZ-NV-NM-CO-UT) is doing the worst at containing the virus at this time and the NE7 region (NY-NJ-CT-DE-MA-PA-RI) is doing the best. But, the fact that a number of those states enacted strong social distancing measures early (CA and WA especially) and never had the extreme spike in cases that maxed out hospital capacity as was seen in NY and NJ means the higher growth rate at this time is probably “manageable.” And, I believe the concept of “manageable volumes” is something we are going to have to learn to live with for the next few months or more.    

 We cannot stay in our homes forever. We cannot force businesses and broad swaths of the economic activity to stay shuttered forever. We need to accept the fact that the virus will be with us until a vaccine is successful, safe, available, and widely administered. That is, we are going to have to learn to live with this disease continuing to spread, continuing to make people sick, and continuing to make some require intensive care in a hospital. So, we need to make sure that we control the volumes so the healthcare system can deliver the best care possible to those patients when they need it. 

Otherwise, some people will die – not for lack of capability, but for lack of capacity. That would be a tragedy.  

I now believe the below charts can provide a better view of how to track whether we are effectively controlling against that outcome, while we try to return to some version of normal life and work. The first one provides a simple 7-day moving average on the total number of newly confirmed cases in each of 5 regions in the US – the two described above plus the G-lakes (MN-OH-WI-IL-IN-MI), the SOUTH (AL-FL-GA-MS-SC-TN-AK-MO-LA-TX), and the REST. This chart clearly shows how extreme the spike the Northeast experienced back in mid-April reached (days 30-40 on the x-axis), when there were more than 120k new cases confirmed each week. I think it would be hard to argue against the statement that “New York City has one of the best concentrated set of hospitals in the country and probably the world” but, when you learn that there were refrigerated trailers being repurposed from food delivery to temporary deceased body storage, you know that healthcare system must be at the absolute edge of that world-class capacity.  

Second, the chart also makes clear how well the new case volume is under control in recent weeks in this region and, in fact, that the NE7 region has the lowest new case volume of any region in the last week or so (represented by the last few data points on each of the lines).

Third, the chart also shows how steady the case volume has been in the other 4 regions over the full period that the virus has been dominating life in the US. Yes, there is growth in 7-day case volumes in both the South and West over the last 2 weeks – the second chart which plots the growth rates from the first chart makes this even clearer (above 0 means weekly volume is growing, below means declining). But, both of these regions have shown that they each have successfully managed case volumes in the 20-40k range for the last 45 days or so – and, I think that is a volume we will need to live with for a much longer period – until a vaccine rolls out, or the virus runs its course to herd immunity.

Admittedly, these regions are defined somewhat broadly and there could be localities where hotspots and capacity stresses emerge. In fact, there have been news reports regarding capacity challenges in Arizona this past week. But, the reports also include information about how the hospital systems in Arizona are working with systems in California that currently have the capacity to provide recommended treatment. That is the point of this type of view – to understand if capacity for effective treatment remains available within a reasonable distance, we can still manage volumes at this level of infection. Yes, effective care may not be available as close as we have become accustomed to over the past few decades – but, it is accessible and that is a reasonable trade-off to make to not keep our communities in a lockdown. Zero new infections is not a reasonable goal at this time.     

The net of all these charts is we are going to need to learn to live with this thing – the virus, the math, and the trade-offs – as we return to work again, socialize with friends again, see our full families again, support our local businesses again, explore new opportunities with our children again, watch (bet?) on live sports again, and just get away from these blue screens again. We can track various statistics like the ones shown here and probably a few others (e.g. - positive test ratios, antibody present rates, mortality) to ensure that we are in control of the infection rate, are able to react quickly to local hotspots, and are delivering the best care possible to those infected – and get back to life, work, and building a better future for all once again.