The Numbers Just Don't Add Up

California issued a revised and more restrictive stay-at-home order that closed restaurants and reduced capacity at many other businesses and activities over most of the state, effective November 21. According to medical.net,

. In the case of coronavirus, the incubation period is estimated to be 5 – 6 days; however, the exact recommended range is 1 - 14 days.

November 21 was more than three weeks ago, and we might expect many incubations since that time to emerge as full-blown cases -- but with the new restrictions, wouldn't we expect to see some effect in the number of new cases since that time?

On December 2, when it had been ten days since the order, I went to that day's data from the LA County Health Department to see what the actual effects of the newest restrictions might be (Click on the image for a larger copy). The data leads to several questions. First, although we might expect to have seen the newest restrictions begin to show a clear reduction in cases at that time, this wasn't showing up on the chart. The cases continued, and still continue, to spike.

Second, the "case death rate", the stepped black line lower down on the chart, notwithstanding the drastic increase in "cases" over the course of the year, had stayed steady at 2% since the summer. I went to the CDC site to get national numbers. As of that day, there were 13,626,022 "cases", with 269,763 deaths, which corresponds to a national case rate of 2%.

Los Angeles County had a total number of cases at 414,000, with total deaths at 7740, for a slightly lower case death rate of 1.9%.

Third, the cumulative deaths, the dark line at the very bottom of the LA County graph, simply haven't spiked. However, the news media keeps referring to a COVID "surge" in recent weeks, with an increase in "hospitalizations", but it doesn't appear that there's been a corresponding increase in the death rate.

But it might be objected that 2% of a continuing increase in cases is still bad -- but recent data seems to be inconsistent. The December 2 rate, according to data from the LA County Health Department was 5,987 new cases and 40 new deaths. This gave a case death rate of only .6 of 1%. Curious, I went to the November 27 numbers from the same source and found that on that day, there were 4544 new cases and 24 new deaths, giving a case death rate of .5 of 1%, again much lower than previous national and county case death rates. (Both December 2 and November 24 were midweek days, avoiding potential underreporting due to weekend delays.)

This morning I went to check the LA County daily stats again. Although weekend statistics may show a delay, yesterday, December 13, had 12,731 new "cases" reported in LA County, with 29 new deaths. I brought up my trusty calculator and found that this amounts to a case mortality rate of 0.2%, roughly equivalent to the seasonal flu. This is one tenth of the currently reported national case mortality rate of about 1.9%. I went to the Friday, December 11 stats and got 13,815 new cases with 50 new deaths, for a case mortality rate of 0.3%, not a big difference. So then I went to the November 4 statistics to see what things looked like midweek before the Thanksgiving lockdown. There were 1843 new cases and 22 new deaths, for a case mortality rate of 1%. This is still half the national rate, when LA County is said to be a major trouble spot.

On November 24, only three days after the new restrictions took effect, LA County experienced 3,692 new cases and 51 deaths, for a case mortality rate of 1.4% still well under the rate for the year of about 2% in both the county and the US.

So LA County has had a jumble of state and local restrictions that are about the most severe anywhere in the US -- outdoor dining is prohibited, for instance, when it's allowed in New York City -- while its statistics have consistently been below national levels, and one critical measure, the case mortality rate, has in fact been falling significantly. The stepped black liine on the chart above should in fact be stepping significantly down.

But another explanation for the lower case mortality rate since early November, and since Thanksgiving as well, could simply be the increased testing. But here's a question. If, as the CDC bigwig Aron Hall, whom I quoted the other day, says, the nationial case rate could be as much as a third of the population, then these people have been infected for quite some time -- and their infections have been mild or asymptomatic. So the perception of a declining death rate may be just a statistical artifact and an indication that the disease has never been that serious as a percentage of the population. Like, say, the seasonal flu.

Even so, if it's hard to avoid thinking that large numbers of "cases" resulting from increased testing are in fact resulting in a lower case mortality rate, shouldn't the political class and the moral enforcers be patting us on the head for doing a good job? And maybe, just maybe, easing some of those restrictions for Christmas?

Don't count on it.