Another forecast gets bitten by the Tuesday effect? The Texas people were sure that we had passed the peak in one-day death rates. But today was the biggest one-day death rate, at least according to this tabulation.
The 3DDRR only went up a bit, to 1.17. Outside NY, it is at 1.21
The main point of tracking the 3DDRR is to get an idea of what the trend in infections was a few weeks earlier. And I don’t see any point at which you can argue that “Aha, this was when the lockdowns got going, and you can see that a few weeks later the spread rate started to plummet.”
Lockdowns started to become widespread around March 20. So we would expect the big decline in the increase in deaths to begin somewhere between April 5 and April 15. But if you look at the chart, the big decline in the death rate was taking place from around March 26 through April 6, and subsequently the declines have been more gradual.
Perhaps the lockdowns failed to dramatically reduce the overall spread rate. But I think that a more likely scenario is that they did slow the spread rate–among the population that is least likely to die from the disease. The overall death rate remained high, because we have not figured out how to protect the elderly, particularly in nursing homes.
If my hypothesis is correct, then a weekly series of random-sample tests in the population would show a sharp decline in the spread rate, but a demographic breakdown of deaths by week would show an increase in the proportion of deaths among the elderly. I know we don’t have the former data. Are the latter data available?
The CDC provisional death counts page is a good place to start, though the way they report the numbers is a bit confusing.
Table 2 breaks it down by age, and Table 4 by place (e.g., home, nursing home, etc… ) However, if someone got sick at a nursing home, and died in the hospital, I think that’s still one of the “healthcare setting” categories, which frustrates the effort.
What we need is deaths by the actual date of death. Deaths by date of announcement is close to meaningless, but that’s what we are getting. States are starting to go back and count probable deaths, and they haven’t been doing it in a uniform way.
The CDC should be collecting all of this data, normalizing it, and making it public.
I am becoming more convinced that heterogeneity is extremely important and also little understood. Very likely, the first two be infected were those most active socially — the so-called “super spreaders.” As that group was exhausted (reached herd immunity?), contagion rates were certain to fall. Similarly, those most likely to die conditional on infection were among the early to succumb. Thus, on average, death conditional on infection should be lower than a month ago among all demographic groups.
Here in Massachusetts, long-term care facilities are accounting for about 30% of confirmed cases and 55-60% of deaths.
On a population basis, cases are distributed evenly across everybody except very young (few) and over 80, who have far more. Out of slightly under 2000 dead so far, 95% are over 60.
Some people are arguing the reason that the expected post-lockdown decline did not occur is because infections had actually spread widely, much earlier, perhaps in December: https://www.thesavvystreet.com/the-pandemic-cardinal-numbers-no-one-talks-about/ The recent testing of older blood samples that has found deaths attributable to COVID in California in early February would seem to support this theory. It increasingly seems as if we are going one step forward, two steps back, with what we can’t assume much about what we know about this disease.
With regards to nursing homes, On 3/25, NY Health Dept directed that nursing homes re-admit seniors from the hospital, even if they were confirmed to have the virus.
Note: I’m not saying there’s a better option – where else do they go, special cv19-only homes? – but it seems a risky move.
On 4/20, the NY health commissioner, Howard Zucker, was asked about it, and said, “if you are positive, you should be admitted back to a nursing home. The necessary precautions will be taken to protect other residents there.”
Um, maybe?
where else do they go, special cv19-only homes?
Maye that’s what the Javits Center in New York City and the Boston Convention and Exhibition Center should have been turned into.