Daily Fluctuations

A good friend of mine is a finance guy. One of the best pieces of advice he gave me when I asked about long term investment planning (with my very meager public health income) was “Don’t look at the daily rates. Look at trends over time.” Daily rates go up and down, we need to see long term trends.

The same is true for an outbreak of a respiratory disease with a 14 day incubation period. And especially true for a pandemic when lab results are taking 10–14 days to report.

It’s easy to be hopeful about the recent decline in cases. Those shorter blue bars over the past few days look promising.

But look at the red line, which averages just 3 days worth of cases. Yes, it’s lower today, but it fluctuates every few days. Even if we remove that anomaly of more than 5000 cases, that average is still substantially higher than our current systems can manage. In May we had consistently less than 100 cases per day, often less than 50. In July we’ve had more than 500 per day, most days.

This is why we look at it over time. Here’s our 21 day average. Note the red line is not going down. The average is still near 1000 cases per day over the past three weeks.

It’s plateauing, but at a rate higher than we can manage.

These rates for hospitalizations are a little skewed as the denominator is the number of active cases on any given day. On July 20, that number abruptly dropped from almost 17,000 active cases down to less than 6,000 active cases. 11,000 people didn’t suddenly recover. Probably many backlogged cases got reclassified. Since July 29, I’m using the total number of cases minus the recovered cases for the number of “active” cases, which may still be inaccurate, but it’s what we have. So, given that, our rate of hospitalization has decreased. Great.

The number of hospitalizations are also decreasing. They are still well over where we were in May, though. And the number of staffed beds does not depend on a wobbly denominator.

Bottom line, we can be optimistic that San Antonio is moving in the right direction. New case numbers are declining, but it would be foolhardy to look at a few days or even a week. Like the stock market, the numbers go up and down every day. We need to look at the trend over time.

(And if these graphs don’t look a whole lot different than last week’s, that’s because the trends haven’t changed. The data are up to date as of 8/2. The pattern is pretty much the same)

As usual, all data can be found at https://covid19.sanantonio.gov/Home