That only way to get a true count of COVID deaths

[edhopper]

is to look at total deaths year over year for March and April, handicap for the difference in causes like car accidents and cancer and see how many more deaths there were.

For example New Jersey had 9,000 more deaths in April this year than last. But reported COVID deaths were 7,000.
One could assume that those added deaths were also COVID.

I think that the current toll of 80,000 is an undercount.

https://nj1015.com/more-than-twice-as-many-nj-residents-died-in-april-2020-than-last-year/

[Igel]

1. There is no way.

We can adjust the stats to try to eliminate noise and get a good average, but there's no way to get an average death toll in the absence of COVID for one month.

Adjustments for car crashes have to be balanced against deaths due to strokes and other conditions that weren't presented to the ER for treatment. Or just strokes and heart attacks while infected with COVID. I have a very mild case , ignore it and do work that gives me a heart attack because I'm 69 and my arteries are clogged, I'm dead from a heart attack ... but the COVID test says otherwise.

You're 80 and die of a respiratory ailment, not COVID-tested, you're a COVID death. 60% of those tested in NY, 95% of those tested in TX, with screening for having COVID symptoms, are negative. Clinical diagnosis in the US is tough; in China, they had mobile scanners for looking less obvious symptoms. Even at DU people are saying they had a fever, cough, etc., back in January--when most coughs were flu, 99% or more--and are convinced that it had to be COVID.

What's happening is that people are choosing average versus last year's monthly toll, on a month by month basis. Last year 7012 people died in April, but the average for the last 3 years is 6750, go with 6750. For March, last year 6250 died, and the average is 6800, go with 6250. In other words, pick the past that gives you the largest net present death toll. Honest? Only in the sense that it achieves the greater truth of saying how bad things are.

But there's no way of fixing the current month that isn't a guess. Minus car crashes (lock down), increase deaths from doing nothing or overexerting yourself, decrease deaths from the flu, increase deaths from fear of the ER and no elective surgeries, increase deaths from depression and social isolation.

[edhopper]

2. Maybe not for an exact count

but it seems you describe why we can get a very good estimate by taking in all those factors.

[dalton99a]

3. Kick

[edhopper]

4. Thanks