Thread by @NYHammond

The NYS Health Dept's July 2020 report on COVID in nursing homes was already discredited for understating the number of resident deaths.

But it includes another error I hadn't noticed or heard about before, and it's kind of a big deal.🧵empirecenter.org/publications/l…

As you can in the two screen shots below, the text of the report says the *number of staff* reporting symptoms peaked on 3/16/20.

But the accompanying chart says something different -- that what peaked on 3/16 was the *number of homes* reporting a 1st symptomatic staff member.

A separate chart shows that about 2X more staff infections were reported in April than in March. It doesn't give an exact date for the peak, but it's seems likely that it's in April. empirecenter.org/wp-content/upl…

Why is this big deal? Because the report's central argument is based on the timing of three trends -- in staff infections, residents deaths, and transfers of COVID-positive patients from hospitals.

Because the peak in COVID-positive transfers (4/14) came after the peak in resident deaths (4/8), DOH asserts that transfers *could not be* the explanation for resident deaths.

On the other hand, what it describes as the peak in staff illness (3/16) comes 23 days before the peak in residents deaths (4/8) -- which it notes is consistent with how long it typically takes for COVID to become fatal.

But Figure 4 does not actually show that the number of staff infections peaked on 3/16. That's the peak of something else entirely: the number of facilities reporting a first infection.

Meanwhile, the report does not share enough data to know precisely when staff infections really did peak, but does make clear that they probably peaked in April and possibly after the peak of resident deaths on April 8.

By the report's own logic, then, this would seem to *rule out* staff infections as an explanation for resident deaths.

That's an absurd conclusion -- but so, too, is the report's argument for ruling out hospital transfers.

One of the @empirecenter's 62 FOIL requests for pandemic data seeks the details that would clarify the precise peak of nursing home staff illness.

DOH recently delayed responding because it was conducting a "diligent search" for the data. empirecenter.org/publications/e…

The DOH also low-balled the number of resident deaths, which was bad faith, but did affect its conclusions.

Basing an argument on misrepresented evidence, while withholding what the evidence actually says, is the bigger deal.

Read more here: empirecenter.org/publications/l…