https://www.wsj.com/articles/coronavirus-latest-updates-10-23-2020-11603442326
Today's Wall Street Journal front-page headline screams: "New U.S. Covid-19 Cases Top 80,000 to Reach a Single-Day Record." The less erratic 7-day average reached 71,671, which is indeed almost as high as the peak of 77,362 on July 16.
But on the following day, the graph in that same article shows that cases fell to 16,801 and deaths to only 194. Do not expect to see sensational headlines about a single-day record when cases or deaths are so inconveniently low.
Question #1: Does a two-month rise in the number of reported cases mean the virus is spreading more rapidly, or does in it mean testing is spreading more rapidly?
The complex of seven high-rise condos where I live has lately been testing all the maintenance and food workers, and a couple without symptoms tested positive and been quarantined. In the past, when tests were scarce, those workers would never have been tested and therefore never reported among Florida's "confirmed cases."
In September, my two nephews were required to be tested before being admitted to a new prep school and university out of state. Neither had a fever nor any symptoms but both tested positive and were quarantined. In the past, they would never have been tested, never guessed they had been infected, and never reported among North Carolina's "confirmed cases."
Question #2: Since the rise in cases is being compared to a similar "surge" in July, what happened to COVID-19 deaths after July?
The Wall Street Journal writers don't answer that question because they don't ask it. What they write is:
"Deaths haven’t surged in the same way, increasing 7% since the beginning of the month. However, there are indications fatalities may be rising. More than 850 deaths were reported Thursday, according to Johns Hopkins data, down from more than 1,100 a day before. Reporting varies among states, and deaths recorded on a certain day might have happened earlier."
"As of Thursday, the seven-day moving average of deaths had reached 763, the highest level in a month. That metric has outpaced the 14-day average nine of the past 14 days, suggesting fatalities are rising."
What happened "as of Thursday" is misleading, however. The seven-day moving average of deaths was 858 on October 1 and 388 on October 18, for example. The average reported on any single day does not indicate a trend, nor does the arbitrary comparison with 9 out of 14 days.
The article repeats a common claim that the rising cases (as in July) predict rising deaths with some curiously unspecific time lag: "Deaths can lag behind a surge in cases by six weeks or more."
But a lot more than six weeks have passed since July, and we're still waiting for the supposedly lagged surge in deaths. The 7-day average of daily reported new cases rose from 17,787 on July 8 to 72,010 on July 29. So why didn't COVID-19 deaths rise dramatically in August or September?
To reply that it's because hospital treatments got better is an unpersuasive evasion: Treatments didn't change that much from July to September, and not very many were hospitalized.
The Journal's own interactive graph shows the 7-day average of daily deaths rising almost simultaneously with cases, from 861 on July 8 to 1,431 by July 29. Yet deaths certainly did not keep rising over the following "6 weeks or more." On the contrary, the 7-day average of daily reported COVID-19 deaths fell to 592 by the end of August and to 194 by October 23.
One reason for the cliché that "there's a lot we don't know about COVID-19" is that many seemingly fact-based claims about newly reported cases and deaths are, in fact, not factual.