Row

Row

Deaths as % of 55 and older population is loosley correlated to 55 and older population density. Other variables likely to be important are use of mass transit; # of international passengers; prevalence of multi-generational households, and distribution of pre-conditions like coronary, diabetes, and asthma issues (these in turn are likely correlated with socio-economic variables and regional differences in diet etc.)

Deaths, Implied Cases, and Age Distribution

Cumulative Covid deaths as of 10-16-2020: 218,233

Current IHME Univ of Washington Model mean estimate for projected cumulative total deaths: 389,087. Estimate range: 358,386 - 415,084.

Normal annual US total deaths all causes is 2,866,909 (i.e., 7,855 per day) Source: US Census Pop Clock. In 2018, 2.62% of the population aged 55 and older died. If the mean Univ of Washington forecast for Covid deaths proves correct and those deaths are added to 2018’s numbers, 3.01% of the 55 and older population would have died (assuming no offset from lower accidents, other infectious diseases, workplace accidents etc. from the shelter in place policies).

CDC weekly death data through October 3rd show total deaths from all causes has increased in 2020 (see graph below) – and may increase more as 2020 data is incomplete due to lags in reporting. Through Week 39 of 2020, there have been 2,371,107 deaths from all causes. That is 247,534 more than the same period in 2019 or 12% more. It is 246,931 (12% more) than the same period in 2018 (a bad flu year).

Implied number of cases from which currently reported total deaths came using estimates from a preliminary CDC paper: 7,564,575. Early CDC estimates are 13 days from Covid confirmation to death; Univ. WA estimates indicate shorter period of 6 days from hospitization to death. The reported number of cases from 13 days ago was: 7,341,303, reprted number 6 days ago: 7,676,683.

Using today’s reported cases (but that number will grow) of 8,012,739 and the CDC paper relationship between cases and deaths would imply a future cumulative number of deaths of 231,162. These relationships suggest that the actual number of Covid cases has been significantly higher than the number confirmed by tests and/or the actual death rates and/or time between Covid confirmation and death have been different than preliminary CDC estimates.

Implied number of cases consistent with CDC paper for IHME Univ. Washington estimate for projected total deaths: 13,486,876

Implied number of cases from Kinsa thermometer data is approximately 7,944,000 if you assume maximum 2.67% of excess influenza-like diseases are attributable to Covid (Kinsa data is missing for many sparsely populated counties however).

2018 Annual US death by age from all causes from CDC WONDER. Covid death/case distribution by age based on CDC paper.

CasesCDCpaper = Covid cases by age reported in the preliminary CDC paper referenced above.

midDeathRateCDCpaper is the midpoint of the upper and lower death rates estimated in the CDC paper.

2018AnnAC = US 2018 data on all deaths from all causes by age group from CDC WONDER.

CurrRepDeathEst = Estimated age distribution of currently reported Covid deaths calculated based on midDeathRateCDCpaper and CasesCDCpaper distributions.

CurrAnnzedPct2018 = CurrRepDeathEst annualized (365/days elapsed in 2020) as a percent of 2018AnnAC

uWAprojMean = Mean projection from University of Washington model cited above for ultimate number of US Covid deaths distributed by age group based on the midDeathRateCDCpaper and CasesCDCpaper distributions..

uWAPct2018 = uWAprojMean as a percent of 2018AnnAc (2018 total deaths from all causes by age group)

currImpCases = CurrRepDeathEst/midDeathRateCDCpapaer. This is the inferred population of cases consistent with reported deaths to date and the mid point of the death rate range reported in the CDC paper. Value for 19 years and younger age group inferred from CasesCDCpaper.

uWaImpCases = uWAprojMean/midDeathRateCDCpapaer. This is the inferred population of cases consistent with reported deaths to date and the mid point of the death rate range reported in the CDC paper. Value for 19 years and younger age group inferred from CasesCDCpaper.

Row

Row

Call:
lm(formula = stpw5$pctPosPop ~ stpw5$pctTestPop)

Residuals:
       Min         1Q     Median         3Q        Max 
-0.0198562 -0.0071948  0.0001672  0.0066784  0.0170486 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)    
(Intercept)      0.021763   0.003520   6.183 1.13e-07 ***
stpw5$pctTestPop 0.004102   0.009801   0.418    0.677    
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.009042 on 50 degrees of freedom
Multiple R-squared:  0.00349,   Adjusted R-squared:  -0.01644 
F-statistic: 0.1751 on 1 and 50 DF,  p-value: 0.6774

Call: rlm(formula = stpw5$pctPosPop ~ stpw5$pctTestPop)
Residuals:
       Min         1Q     Median         3Q        Max 
-2.006e-02 -7.350e-03  1.761e-05  6.466e-03  1.686e-02 

Coefficients:
                 Value  Std. Error t value
(Intercept)      0.0220 0.0036     6.1562 
stpw5$pctTestPop 0.0038 0.0100     0.3843 

Residual standard error: 0.01051 on 50 degrees of freedom