More on the Covid-19 spread

This is an update of last week’s post on the spread of Covid-19 infections and deaths in the US, California, and Monterey County. I’ve introduced a few refinements to the graphs.

The first graph is of the growth of the daily counts of confirmed cases and deaths. It is almost certain that the number of confirmed cases dramatically undercounts the actual number of Covid-19 cases, perhaps by as much as a factor of 10 or 20.

 

The spike in last weeks death data deserves explanation. It was caused by New York’s changing of their criteria for Covid-19 deaths from confirmed-by-testing to probable. The good news is that the number of new cases is fairly flat in the last two or three weeks. The bad news is that it remains high, as does the number of deaths that occur every day.

The above graph plots the daily percentage increase in both cases and deaths. The number of cases is currently growing at somewhat less than 5% per day. At that rate, the number of cases doubles every two weeks.

Let’s turn to the California numbers. I have previously used the Los Angeles Times data, which is the most accurate that I know of. Today, I’ll show you the California Department of Public Health numbers. The reason I’ve switched to the CDPH data is that it contains historical and current information not obtainable from the LA Times site. In the interest in keeping comparisons relevant, I’m using only the CDPH numbers in this post.

Here are the confirmed cases data, with both daily increases and the cumulative counts.

The case increases have approximately leveled off recently.

Here is the same presentation for deaths:

The number of daily deaths continues to rise.

Next is the percentage daily increase in the cumulative number of confirmed cases.

The orange line is the 7-day  moving average. Here’s the same plot for deaths:

You can’t have a confirmed case without a completed Covid-19 test, and the availability of those tests has been limited, so it’s worthwhile looking at the number of completed tests over time:

There is an anomaly. The number of tests completed rose slowly until April 4, when it took a big jump. Since then, it has remained more or less constant at on the order of 10,000 tests per day.

If we take a look at the numbers of total tests, which includes those not completed, we can get a better idea about was was happening.

A large backlog developed in late March, which was largely cleared on April 4. But looking at the percentage of tests that were positive shows something else:

Looking at the cumulative tests, we can see that the percent positive climbed dramatically before April 4, then fell to its present value of slightly over 10%.

We can gain more insight by looking at the daily results:

In the week or so before April 4, the majority of the completed tests were positive. Hardly any of the 4/4 backlog clearance tests were positive. My guess is that in the time immediately prior to 4/4, tests on the sickest patients were prioritized, leaving uncompleted tests that were less likely to be positive. I have no explanation for the spikes on 4/6 and 4/15.

For Monterey County, I had been using the Carmel Pine Cone’s numbers, which are derived from the Monterey County Health Department’s data. I am now using the County’s data directly. It’s kind of a pain to do so, since the County has a habit of revising the historical data every day or so, but, because they do that, the Pine Cone’s numbers don’t properly represent the history.

Here are the cumulative cases, with a 5.5% line posted for reference:

And here are the daily percentage increases in the cumulative cases.

You can see that about 10 days ago, it looked like the rate of increase was falling fairly consistently, but last week was a bad week, especially in Salinas, and that drove the rate up again. Don’t read too much into the last two or three days — if history is a guide, those numbers will be revised upwards.