This was originally a Facebook post of mine from 2021-6-11, and is archived here as a curiosity. Minor changes to formatting, as well as basic copy-edits, may have been made in the transition from Facebook post to web format.
Despite some after-effects of the labor day weekend affecting the stats, things improved steadily this week as far as Covid is concerned.
We can start with the source of data least impacted by holidays -- the daily hospitalizations. Hospitals run right through holidays and weekends, because heart attacks don't take days off, and consequently they form their own little universe, unaffected -- at least as far as data reporting goes -- by the events of the outside world.
This week, we averaged 2167 new hospital admissions with covid each day, compared to 2511 the week before, down by about 14%. This rate of fall is consistent with the excellent progress we've been making each week since mid-April.
Now, when a major holiday weekend occurs, as we've seen over and over with this virus, cases and deaths are undercounted that week, and overcounted the next, causing both to turn up relative to their previous trajectory. It's delayed reporting. That's happened here too.
On May 27th, the seven-day rolling average of cases stood at about 21,400. Over the next seven days, it was 14,200, and now runs about 13,800. This extra-steep drop, followed by almost no drop, is probably a Memorial day effect -- the 14,200 was probably an undercount; the 13,800 may be an overcount. Likewise with deaths, which stood at 425 May 27th, 331 June 3rd, and 352 now. If, a week from now, we see both deaths and cases dropping in the old way again, then that will confirm that things are continuing to move in the right direction.
The latest numbers from outbreak.info show about 14% of recent sequences taken in the US are coming from the B.1.617.2 strain, originally found in India. Those numbers are continuing to grow fast, and suggest that about five weeks from now, this more transmissible strain will make up the majority of US cases. But no matter how fast it spreads, 64% of adults in the US, and 87% of adults above 65 years old, have already been vaccinated. And the percentage of people with some immunity to the coronavirus is higher than that -- there's also the large number of people who have already had the virus who must be accounted for.
So even if, in mid to late July, we start seeing some increase in covid cases, it will only be a minority of Americans who are susceptible.
Sources
For most basic statistics, CDC COVID Data Tracker.
For the prevalence of B.1.617.2, see outbreak.info.
This page is released under the CC0 1.0 license.