Post #1398: Still looks like COVID-19 hospitalizations are flattening, despite rising case counts

Posted on January 12, 2022


I know you’re probably still hearing that things are wildly out of control in terms of COVID-19 hospitalizations.

All I can do is show you what’s on the data file the Feds use to track that.

The data source for this analysis is what the US CDC refers to as the Unified Hospital Dataset.  The actual file I’m using can be seen and downloaded at this U.S. DHHS link.

This the update through file date 1/12/2022, but I’m only showing the data through 1/11/2022 because the last day of data tends to be incomplete.  (Best guess, about 8% too low, but it seems to vary from day to day, so it’s best just to leave the last day off.)

Recall that this file has a strong weekly pattern of under-reporting and catch-up, and that I performed some sort of statistical whatsis to even that out as best I could.

Here are the raw and adjusted data.

Source:  Calculated fro unified hospital dataset cited above.  Data through 1/11/2022, file date 1/12/2022

Just by way of explanation, in the raw data (top), today is the day that the numbers should more-than-bounce-back, to make up for the under-reporting in the preceding dip in the line.  Note that they only came up level to the previous high rate.  And so, after adjustment, my inference is that daily new hospitalizations actually fell on that final day, after netting out the expected variation in reporting.

Nationally, we’re getting to the point where COVID-19 patients occupy as large a fraction of U.S. hospital beds and ICU beds as they did at the peak of last year’s winter wave.  Like so, through 1/12/2022 this time, because this is is simple proportion of what’s reported (modest incompleteness of the file doesn’t matter much for these percentages).

Source:  Calculated from unified hospital dataset cited above.  Data through 1/12/2022, file date 1/12/2022

But you have to keep in mind that last year’s winter wave peaked in a few states at a time.  What we saw as the share of all U.S. beds was really an extremely high caseload in some states, divided by all the beds in the U.S.  By contrast, this year’s Omicron wave has involved all the states at once.  It’s spread across all the states.  You can achieve the same fraction of all U.S. beds occupied with lower proportions in each state.

Let me do some round-numbers quantification of that.  Let me count the number of states, over time, where COVID-19 cases filled 30% of more of ICU beds, 40% or more of ICU beds, and 50% or more of ICU beds.  And I’ll color it as a “stoplight” chart:  Green is OK, Yellow is warning, and Red is a problem.

Source:  Calculated from unified hospital dataset cited above.  Data through 1/12/2022, file date 1/12/2022

Notice what’s missing from the Omicron wave?  No red, almost no yellow.  Sure, the COVID-19 share of U.S. ICU beds is as high as it has ever been.   But that’s because Omicron is everywhere, all at once.  What you’re not seeing (yet) is a small subset of states having their hospital systems maxed out.  Which is precisely what you saw in prior waves.

So it’s an odd thing, in that the national numbers look as bad as they ever have.  But it’s really not at all the same level of distress.  And that’s because all hospital bed shortages are local.  In prior waves, we achieved those high national numbers with ultra-high numbers in a few states at a time.  This time, we’re achieving the same national numbers by spreading the cases far more evenly across states.  And thus (so far) avoiding the crises we saw at the last two peaks.