Post #1217: COVID-19 trend to 8/19/2021, and the shifting ratio of deaths to hospitalizations.

Posted on August 20, 2021

Well, we haven’t peaked yet, but it looks like we’re close.  The U.S. is now at 44.5 new cases / 100K / day, up from 43.5 yesterday.  Seven-day growth rate for new cases per 100K per day is 14%.

Data source for this and other graphs of new case counts:  Calculated from The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 8/20/2021, from”  The NY Times U.S. tracking page may be found at

At any rate, my simple curve-following projection still appears to be on track.  The .gif below shows my projection from a week ago, and then the actual data through 8/19/2021.

I guess if you look at the full complexity of all states, for the entire pandemic, you can’t quite see a peak yet.

If you narrow it down to just the six states that account for the majority of U.S. cases, by contrast, it looks like it’s peaking.

Too few deaths.

As the Delta wave has grown, we’ve have plenty of cases and hospitalizations, but few deaths, compared to prior waves.

Plenty of people have noticed this, and I’ve been thinking of taking a look at it for some time.


It’s tricky to analyze the deaths data because there’s about a two-week lag between the count of new cases and new hospitalizations (which are nearly coincident in timing) and deaths.  That said, the CDC recently placed all of the relevant data onto a single graphing tool on its COVID data tracker website.  With that, I think I can finally illustrate the issue simply and convincingly.

Below is the last year of COVID-19 case, hospitalization, and deaths data from CDC.  This animated .gif shows new COVID-19 cases, then flips between new hospitalizations and deaths.  Note that the hospitalizations line is way up, but the deaths line is not.   The ratio of hospitalizations to new cases is far higher now than in prior waves.  By contrast, the (concurrent) ratio of deaths to new cases is far lower now than in prior waves.

Source:  CDC COVID data tracker.

I find it particularly puzzling that we’re seeing this combination of more hospitalizations and fewer deaths.  In the past waves, a relatively constant proportion of hospitalized individuals died.  That made sense in some hierarchy of case severity (asymptomatic, outpatient only, inpatient, death).  Those severely ill enough to die were more-or-less a constant fraction of those severely ill enough to be hospitalized.

But now, apparently, a larger fraction of those infected are sick enough to require hospitalization, but not sick enough to die?  That’s odd.

I don’t think there’s been any material change in treatment options available.

And, on an apples-to-apples comparison, research suggested that Delta generates significantly higher risk of hospitalization and death compared to the prior (Alpha) variant.

Based on that, a change in the population being infected must be the cause.  I’ve seen a lot of possibilities tossed out, but nothing that was linked to any data-driven analysis.  Basically, just arguments based on plausibility, either pointing to the shifting age distribution (younger), or pointing to the high vaccination rate.

I’m not going to resolve that here.  For now, I think it’s enough to show just how pronounced this shift has been.  Far more hospitalizations, far fewer deaths, per new COVID-19 case.