Forget total cases, keep a sharp eye on hospitalization data

by | Mar 29, 2020


The total number of coronavirus infections is getting far more attention than it should right now, when the real number that matters is the total number of COVID-19 hospitalizations. The two are linked, of course, because more infections means more hospitalizations, but not all infections are equally serious, while every hospitalized patient needs significant medical care and will occupy one of a limited number of hospital beds.

And it’s that total number of unoccupied hospital beds that matters most. It’s the entire reason we’re trying to “flatten the curve” in the first place. By flattening the curve of infections, we also flatten the curve of patients needing significant medical care, which is a limited resource.

On Friday, Governor Ron DeSantis revealed that Florida currently has 19,000 unoccupied beds, and that he and the state’s emergency managers are watching that situation closely. The total number of beds and the total number of COVID-19 hospitalizations are measurable and relevant statistics.

As of Sunday evening, Florida had 633 hospitalizations due to coronavirus. Those two data points: 633 patients vs. 19,000 beds, gives us a clear picture of where things stand.

The complete opposite is true with total cases because there’s no way to know the relevancy or accuracy of the infection data. There are widespread reports of “asymptomatic carriers” of coronavirus who show no symptoms but have the capability to infect others. Then there is an even larger group who exhibit only mild symptoms – some of which get tested, but many more do not. The bottom line is that without testing the entire population, the total number of positively identified coronavirus cases isn’t a datapoint that we can put to good use.

That’s another reason hospitalizations are a superior metric by which to gauge the situation: it’s an accurate and easily measured data point that isn’t inflated by increased testing or underreported because of mild symptoms. Patients with serious symptoms are the most likely to be tested in the first place.

Here’s a snapshot of where things currently stand. Based on previous data, the daily admissions trendline, which is more conservative than total admissions, indicates we should expect a little over 100 new hospitalizations tomorrow, while the total admissions trendline suggests we’ll get closer to 150 new patients.

Because coronavirus has shown the ability to spread exponentially, meaning the total cases and hospitalizations doubles after a set time interval, we can very quickly overwhelm our health care system. In the past 11 days alone, the chart above shows the total cases (red line) has already doubled twice, and could double a third time tomorrow if we get more than 170 new hospitalizations. At 633 current cases, our hospitalization number can only double 5 more times before exceeding our 19,000 hospital bed limit:

600 x2 = 1200

1200 x 2 = 2400

2400 x 2 = 4800

4800 x 2 = 9600

9600 x 2 = 19,200 

On the bright side, the past two days of admissions have come in significantly below the trendline, and even total cases dipped below the trend on Sunday. Every day we slow the growth and flatten that curve, we take pressure off the state’s health care system.

Eventually, the hospitalization curve will flatten out and ultimately head back downward.

One more interesting data point: the hospitalization rates based on total hospital patients versus total number of cases. That number has steadily trended down over the past week, from 22% to just over 12% of positive cases on Sunday:

It’s too early to tell, but the chart above could be an indicator that increased testing is starting to identify more mild cases and is no longer only being done on the most serious cases. It could also mean that we’re simply finding coronavirus patients earlier, and some of those tested, while not currently hospitalized, could become so later. But regardless, if we continue to see a hospitalization rate that is lower than the 20% number of patients widely reported to have serious symptoms from COVID-19, then we’re in much better shape than previously thought.

We’ll keep watching the hospitalization data, and we’ll continue to report once we have new or relevant information.

4 Comments

  1. $64,000 Question

    Is the hospitalization number current hospitalizations or total hospitalizations?
    Are patients that are discharged or deceased subtracted from the number?

    Thank You In Advance!

  2. Brian Burgess

    Good question. I’ll verify this with FDOH and report in the next story.

  3. WBLMI

    Any idea of how many recovered?

    Wouldn’t Total cases – fatalities – those recovered equal a more meaningful number than ‘total cases’?

    I agree. Forget total cases.

  4. Allan Jacob

    Is there is site where I can follow the daily tally of admissions by county?

%d bloggers like this: