[g1313]

I think for many (most) people, statistics have limited ability to affect behavior, and people rely on their lived experience. A 20% increase is a lot, but because of the small baseline rate it still means that the chance that the average person is not close to anyone who has died from COVID is pretty high. 

I think many Americans believe some other people will die, and those deaths are OK.  If deaths approach 250,000 we would be saying ~1 in 1,000 adults would die from COVID19.   I guess that is an acceptable death rate to many Americans.

I wouldn't put it the way you put it.  There's a temporal aspect to the deaths.

The 1 in 1000 is applied to the full population.  Since we see age, comorbidity (+ blood type), ethnic, and social associations with chances of getting the disease and then chances of dying from it, your death chances may vary.

Los Angeles had 965 reported COVID-19 deaths and 32,403 cases (out of 407,463 tested) from May 15 to June 15.  Annualized, that's 11,370 in 9.5M, or ~1.2 per 1000.  More traditionally, that's 120 per 100K per year.  LA's normal, raw all-cause death rate is 615.7 per 100K.  So, death rate appears to be up 19.5%.   That's a of immense proportions jump, IMO. 

To calculate a CFR, I will use the new cases from the month a week earlier (May 8 - June 8, during which 437,860 tests were given, 4.6% of the 9.5M population, 7.2% positive rate)
965/31670 = 3.047%
(Why a week?  NYC and Italy deaths peaked a week after new cases peaked.  I had thought it should be more delayed.)

You can argue all you want about how the IFR should be lower, but just applying that CFR to the 9.5M LACo (minus Pasadena, LB) population gives a raw number of 289,469 (about 25 years worth of 11,370 dying per year).

To be clear, the tests only detect current illness (last 2 weeks or so, if that).  The 400K negative tests and the 9M untested people could have all had previous asymptomatic cases and (potentially) never get COVID-19 again.

[g1313]

I think for many (most) people, statistics have limited ability to affect behavior, and people rely on their lived experience. A 20% increase is a lot, but because of the small baseline rate it still means that the chance that the average person is not close to anyone who has died from COVID is pretty high. 

I think many Americans believe some other people will die, and those deaths are OK.  If deaths approach 250,000 we would be saying ~1 in 1,000 adults would die from COVID19.   I guess that is an acceptable death rate to many Americans.

I wouldn't put it the way you put it.  There's a temporal aspect to the deaths.

The 1 in 1000 is applied to the full population.  Since we see age, comorbidity (+ blood type), ethnic, and social associations with chances of getting the disease and then chances of dying from it, your death chances may vary.

Los Angeles had 965 reported COVID-19 deaths and 32,403 cases (out of 407,463 tested) from May 15 to June 15.  Annualized, that's 11,370 in 9.5M, or ~1.2 per 1000.  More traditionally, that's 120 per 100K per year.  LA's normal, raw all-cause death rate is 615.7 per 100K.  So, death rate appears to be up 19.5%.   That's a of immense proportions jump, IMO. 

To calculate a CFR, I will use the new cases from the month a week earlier (May 8 - June 8, during which 437,860 tests were given, 4.6% of the 9.5M population, 7.2% positive rate)
965/31670 = 3.047%
(Why a week?  NYC and Italy deaths peaked a week after new cases peaked.  I had thought it should be more delayed.)

You can argue all you want about how the IFR should be lower, but just applying that CFR to the 9.5M LACo (minus Pasadena, LB) population gives a raw number of 289,469 (about 25 years worth of 11,370 dying per year).

To be clear, the tests only detect current illness (last 2 weeks or so, if that).  The 400K negative tests and the 9M untested people could have all had previous asymptomatic cases and (potentially) never get COVID-19 again.