by Damian Rafal & coll.
BACKGROUND: What do the data presented in the CDC tables „Deaths involving coronavirus” mean? The one objective information is: „xxx thousands of people have died and being probably infected with Covid-19”. But how many of these people would for sure still live if not Covid-19? The aim of this paper is to present the math-logic method that makes possible to reveal the real number of lethal Covid-19 victims of in the U.S.
METHODS: The ideas for solutions are fully original, mathematical – logical; there were used constructed by us estimators. The calculated data are usually slightly rounded, because the method presentation is the main aim of the article.
FINDINGS: Only up to about 10% of those reported as Covid-19 victims, in the US in 2020, died from Covid-19 complicity and all the rest would have died in the same (or close to identical) time anyway (also without Covid-19) because their deaths resulted only from the normal age-structure of deaths in the United States, creating the average expected age of death in the given year.
INTERPRETATION: The official number of Covid-19 victims is in a vast majority “the double counting” of those who would die whatsoever in the same time even without Covid-19. The ‘ex post’ analysis is necessary to discover the real number of deaths due to Covid-19.
It seems there is no correct essay analyzing the real Covid-19 mortality to find. What do the data presented in the CDC tables „Deaths involving coronavirus” mean? The one objective information is: „xxx thousands of people have died and being probably infected with Covid-19”. But how many of these people would for sure still live if not Covid-19? The main summary reason of deaths is “aging” =advancing age and all diseases (conditions) the frequency and deadly effects of which are very strongly correlated with it (what means, with the overall weakness of the organism); those conditions sources are in the body itself or a condition progress needs much time and advancing age. Next, there are deaths caused by fully external causes like different injuries. Infections have burdening actions (deadly effects are strongly correlated with the overall weakness of the organism /age). Infant mortality is another quite important group of causes of death. The key point to remember is that the number of chronic conditions and life expectancy are strongly correlated too. The aim of this paper is to show how to calculate the real number of Covid-19 lethal victims.
/Any potential influence of misdiagnoses (e.g. Covid-19 instead of the flu) is not considered in this analysis./
The ideas for solutions are fully original, mathematical – logical. Also a number of logic guesses had to be resolved. At first it was calculated what the average age of death should be in the same year in a similar group (to the one assumed to be killed by Covid-19) if nobody was infected. Then, the average further life expectancy for the people from the whole “deaths involving Covid-19” group, if they were still alive, was calculated. The calculations widely used the CDC, NSC and other institutions’ databases, and also Life Table. There were used constructed by us estimators. To understand the procedures of calculations and what the consequences are a reader must follow the resolving and explanations given below. The obtained data are further delicately rounded. In general, the method is in some places slightly simplified to chase calculations, because not all most detailed data were available and because the idea presentation is the main goal of this article, however it still cannot meaningfully influence the final result of this analysis.
Detailed Procedure & Results
The average age of those who officially died from Covid-19
No ready data can be found, apart from the median age of 78 in early data . Could we use the known age ranges and then plot it on a life table? Let’s check it with the average (in a year) age of death in the society. The growing population (with a big role of immigration of mobile people) made it surprising much lower than life expectancy (78.5, the World Bank 2018). A number of deaths “speeds up” in an age range, so if we use the age ranges and calculate, based on ‘numbers of lives’ , median* ages of those dying in the age ranges and then multiply each median by a subgroup’s volume [3-p.25] (when there are differences then calculate separately for men and women) then we receive the following result:
Medians* for the age ranges: <1, 1-4, 5-14, 15-24, 25-34, 35-44, 45-54, 55-64, 65-74, 75-84, 85+ :
M: 0.5, 2.4, 11.4, 21.3, 30.40, 40.50, 51, 60.75, 70.65, 80.55, 91
W: 0.5, 2.5, 10.5, 21.2, 30.85, 40.65, 51, 60.80, 70.85, 80.85, 92,05
[10.46 + (4.9 + 4.09) + (36.32 + 24.27) + (463.23 + 170.09) + (1257.07 + 549.96) + (2164.16 + 1201.21)
- (5041.2 + 3138.85) + (13846.44 + 8938.33) + (22625.45 + 16671.86) + (28603.31 + 26917.15) + (30773.01
- 49300.23)] /2854.838 (total number of deaths, in thousands) = 74.17 (years)
But on the site wonder.cdc.gov –Underlying Cause of Death we can get data grouped by ‘Single-Year Ages’ plus choose 2019 and then calculate it to receive 73.78 years. 73.78 is not 74.17 (for 2019) by a part because the subgroup-medians are higher than averages. …For the purpose of this analysis we take the average theoretical (assuming Covid-19 absence) age of death, in the society in the year 2020, to be 74.0 [it strongly increased from 2018 to 2019 -by 0.5 year (aging of the society), so a next increase should slow down and make up to about 74.0 years.] . …It is advised to get the most precise estimation/anticipation by a detailed demographic with death rates study.
The theoretical average age of those forming the “deaths involving Covid-19” group we estimate in a similar way, starting with shares of age-subgroups (the CDC data – early January 2021) plotted on a life table, to receive 76.6 years (based on the medians), the result only by about 0.05 year higher than if to be based on the earlier data of May 2020 . To receive a better approximation it should be next multiplied by 73.78/74.17 to receive 76.2 years. But the shares of the oldest subgroup (85+) look as follows: 31.05% (the “deaths involving Covid-19” group) vs. 31.65% (the normal one in 2019, after deducting deaths due to injuries) . From the comparisons we can see that shares of most of subgroups in the 0-64 age range are lower in the “deaths involving Covid-19” group. But those percentages that are “missing” in the lower age subgroups are not proportionally distributed in older (65++) subgroups; the oldest subgroup (85+) has not increased but a bit decreased share(!) = that subgroup is by about 1.6% underweighted so has very probably a decreased its average age of death. The estimated negative revision for the whole group is 0.1 year; it is based on the assumption that the difference between the theoretical average age of death in that subgroup  and the age of 85 diminishes proportionally to the diminishing share of the subgroup, so in fact the 0.1 value should be the maximal one. But there can also be unidentified and hidden disturbances in another older subgroup so to be very careful we subjectively deduct 0.1 more to receive the final 76.0 years. …However it would be a help if the needed data was just officially given.
How many of the US “died from Covid-19” had in real their date of death accelerated
a) At the beginning we must calculate what the average age of a decedent would be in a close to identical group* (like the one assumed to be killed by Covid-19) but if nobody was infected and so without Covid-19 related deaths (*CTINI). The already taken, assuming Covid-19 absence, average age of death in the US in the year 2020 (AD) is 74.0 years. But this value needs to be revised upwards due to some factors. Deadly injuries shorten a person’s life and their impact is unique because are not derivatives of already ‘not far from deadly’ health status! Any death due to, for example, a mechanical accident excludes assuming the Covid-19 causative participation, so the average age of death for our group must exclude the impact of injuries in their broad meaning. In the CDC.gov data named „Leading Causes of Deaths” and see there are some groups of causes not directly dependent on aging of the organism.
-Accidents (unintentional injuries): 167127 cases in 2018 /(data for 2019 not available then yet)
-Intentional self-harm (suicides): 48344
Going deeper into it (data for 2018, imported in January 2021 from the website: https://injuryfacts.nsc.org), we can see there are some sub-categories concerning ‘Accidents’, with different age structures of their victims.
-„Poisoning’ 19.9 per 100,000 (deaths per 100,000 population)
-‘Motor-vehicle crashes’ 12.4 per 100,000
-‘Falls’ 11.2 per 100,000 (before the site revised it to 12.0 in February 2021)
-‘Choking’ 1.6 per 100,000
-‘Drowning’ 1.1 per 100,000
-‘Fires/smoke’ 0.9 per 100,000
-‘Mechanical suffocation’ 0.4 per 100,000
We calculate the negative contribution of ‘Poisoning’- (P) to the average age of death (AD) in the following way. The share of all ‘accidental’ deaths in the structure of US deaths is 0.0589 and the share of the ‘Poisoning’ category in ‘accidental’ deaths is 0.37 (0.0589 x 0.37 = 0.0218). We calculate using the following constructed by us estimator [43.5 = the average age of a victim of lethal poisoning (estimate)]:
(1 – 0.0218) x (AD + P) + 0.0218 x 43.5 = AD
0.9782 x 74 +0.9782 x P + 0.9483 = 74
72.3868 + 0.9782 x P = 73.0517
P = 0.6649 /0.9782 = 0.68
The ‘Poisoning’ category by about 0.70 y. has its negative impact on the average age of death in the US. The calculations of the influence of the less important categories in the US: ‘Suicides’, ‘Moto-vehicle crashes’ and ‘Assaults’ give for our group: 0.45, 0.40 and 0.25 year respectively. ‘Drowning’, Choking’, ‘Fires’/‘Smoke’ and ‘Mechanical suffocation’ are all trifles and add up together to the additional 0.10 year. There is one important category with the average age of a victim meaningfully higher than the average age of death in the society = ‘Falls’. We estimated (Injuryfacts.nsc.org or ) the average ‘Falls’ victim age to be 80.0 years. The share of all ‘accidental’ deaths in the structure of the US deaths is 0.0589 and the share of the ‘Falls’ category in all ‘accidental’ deaths is 0.22. So again: 0.0589 x 0.22 = 0.013.
(1 – 0.013) x (AD + F) + 0.013 x 80.0 = AD
0.987 x 74 + 0.987 x F + 1.04 = 74
73.038 + 0.987 x F = 72.96
F = – 0.078 /0.987 = – 0.079
There are additional minor causes of ‘preventable injuries’ (Accidents) with their total share 9% (Injuryfacts), but their age structures are not given there, so we take 0.10 year as its influence on the average age of death (it can be verified by another source [3–p.40]).
There are also deaths due to injury-like preventable medical errors like drug events, mistakes during operations and postoperative events. But they do not let to make any meaningful revision of the average age of death for our group, because those deaths should not concern otherwise of a standard health status people but those who are mostly old and in a worse state and so seek for intensive care. Additionally, opinions about their number are extremely different one from the other. There could be even up to about 50 thousands of such deaths yearly (=unofficially, so many times bigger value), but there are also opinions that the official data are most objective and there are only 5 thousands of all deadly ‘complications of medical and surgical care’ yearly [3-p.40], so of an injury-like type there should probably be 1-2 thousands. We estimate (subjectively) the revision needed due to this factor as from Zero to 0.20 year, and we take 0.1 for the further analysis.
There are still factors that will noticeably revise upwards the average age of death for our group (CTINI), but these factors are associated mainly with the lowest age ranges. We can look at the ‘actuarial life table’  to see that the lowest age ranges factors are in a vast majority “consumed” in the 0-1 age range. The negative impact of infant mortality (birth defects, low birth weight, term birth complications and the rest of the causes) on life expectancy is 0.56 year. As it could be expected, the weight of this age sub-group in the “deaths involving Covid-19” group is close to none (over 70 times less than in all deaths in the society  = 0.01% vs. 0.73%. But from the second value we must subtract classical injuries -mainly ‘mechanical suffocation’ cases (Injuryfacts), not to repeat it. It gives: 0.73 – 0.04 = 0.69%
[1 – (0.0069 – 0.0001)] x (AD + I) + (0.0069 – 0.0001) x 0.5 = AD
0.9932 x 74 + 0.9932 x I + 0.0034 = 74
73.4968 + 0.9932 x I = 73.9966
I = 0.4998 /0.9932 = 0.5032
Any further upward adjusting of the expected average age of death for our group is necessary if the deaths-age-structure of the “deaths involving Covid-19” (DIC) group is further disrupted by deficits of lower age-subgroups’ shares when compared to the normal shares after subtracting deaths due to ‘injuries’ -Those people, who die at age of a lower age range, create this age range negative impact on the average age of death in the year; that impact diminishes with diminishing % of people dying at age of this age range. …We compared shares of age subgroups of the DIC group (early January 2021) with its normal shares in all deaths in the society . Then we corrected the second values by deducting all deaths due to ‘injuries’ (Injuryfacts and ). Next, we calculated the preliminary values by which the average age of death in our group should additionally be revised upwards.
the 01-14 subgroup shares: 0.02% vs. 0.32% (0.20% after the correction)
[9.17 /2854.84 and (9.17 – 3.82) /2615.80]
the 15-24 subgroup shares: 0.16% vs. 1.04% (0.27% after the correction)
[29.77 /2854.84 and (29.77 – 22.58) /2615.80]
the 25-34 subgroup shares: 0.72% vs. 2.07% (0.81% after the correction)
[59.18 /2854.84 and (59.18 – 38.12) /2615.80]
the 35-44 subgroup shares: 1.92% vs. 2.91% (1.83% after the correction)
[82.99 /2854.84 and (82.99 – 35.19) /2615.80]
the 45-54 subgroup shares: 4.97% vs. 5.62% (4.85% after the correction)
[160.39 /2854.84 and (160.39 – 33.58) /2615.80]
the 55-64 subgroup shares: 12.20% vs. 13.13% (13.00% after the correction)
[374.94 /2854.84 and (374.94 – 34.69) /2615.80]
the 65-74 subgroup shares: 21.70% vs. 19.46% (20.38% after the correction)
the 75-84 subgroup shares: 27.24% vs. 24.10% (25.48% after the correction)
There are considerable share-differences in the 01-14 and 15-24 subgroups, but next happen only delicate ones. Deaths due to congenital anomalies have 5%-share in the 1–19 age range and conditions (mainly cancer and heart diseases) play a very small role .
Those preliminary revisions due to deficits of subgroups’ shares are calculated in the following way (the example for the 15-24 subgroup):
[1 – (0.0027 – 0.0016)] x (75.75 + S) + (0.0027 – 0.0016) x 20.85 = 75.75
0.9989 x (75.75 + S) + 0.0229 = 75.75
75.6667 + 0.9989 x S = 75.7271
S = 0.0604 /0.9989 = 0.0605
/75.75 = the average age of death (in 2019) after deducting the negative impact of ‘injuries’; 20.85 = the theoretical average age of death in the subgroup (= the median minus 0.4)/
If some percentages are missing in lower subgroups then it means they are next distributed in higher-age subgroups. For example the total deficit in the 0-24 age-range is: 0.72 + 0.18 + 0.11 = 1.01%, so the next share of 0.72% (the 25-34 subgroup) we should compare not with 0.81% on the right side, but with: 0.81 /(1 – 0.0101) = 0.82% …and further we compare 1.92% not with 1.83% on the right side, but with 1.85%, etc.
The sum of preliminary revisions (PR) due to changed shares within the 1 – 64 age-range is:
0.1256 + 0.0605 + 0.0455 – 0.0245 – 0.017 + 0.1478 = 0.3379
The total sum of PR (for the 0 – 64 age-range) additionally includes the revision due to ‘infant mortality’ (which is by a big part similar to ‘injuries’):
0.5032 + 0.3379 = 0.8411
…What about the 65++ age-range is explained on the next page.
But what matters is the sum of final revisions! Shares of lower age-subgroups in deaths can be a bit lower in the “deaths involving Covid-19” (DIC) group than in the society also due to medical staffs’ discretion. But the factor of genuine ‘Covid-19 deaths’, lowering the average age of death in the DIC group, is superimposed what produces the downwards (age) directed pressure, so we should eliminate the effect of this factor on the revision for the CTINI group next. -If not that decreased average age of death then the deficits of shares of lower age-subgroups could be yet bigger and so the revisions should be yet higher. To calculate the pure impact (the sum of final revisions) of deficits of age-subgroups on the CTINI group we must adjust the average age of death in the DIC group to the higher value of: ’75.95 plus the sum of final revisions itself’, quite a good approximation here is ’75.95 plus the preliminary sum of revisions’ (75.95 = 74 +1.95; 1.95 -impact of injuries). Initial shares of the very lowest subgroups are expected/assumed (with the adjusting) to keep proportions with shares of the following subgroups of the 01-64 age range. Those shares we recalculate/diminish in the way: S x 76.00 /(75.95 + 0.84). Theoretically, we should recalculate this way all shares of the 0-84 age range, but the last and open summary range (85++), which means “the rest” = ‘100% minus the sum of all previous shares’, receives an increased adjusted share and its average decedent-age goes up as well. …It gives the following shares:
0.02%, 0.16%, 0.71%, 1.90%, 4.92%, 12.07% (within the 01 – 64 age range).
Then when calculating the sum of “final” revisions we must also make the correction (a simplified way is: x 75.75/75.95) due to the fact that we take (for 2020) the increased average age of death, by 0.2 year.
The total sum of recalculated revisions, based on the 0-64 age range, is 0.88 year. Next, we finally take 0.85 because of the 65+ age range (explanations are given on the next page*). [0.85 is the highest one of possible values.]
/*The deficits finish before the 65+ age range and so the increased shares of the older subgroups could be purely the result of genuine Covid-19 deaths as well. If they were not, then continuing revisions (starting with calculating the sum of preliminary revisions), to the very end, gives the final result = 0.80. But if the increased shares (within the 65+ range) were purely the result of genuine Covid-19 deaths then to cease that overweight (21.70% vs. 20.77% and 27.24% vs. 25.97% – the shares on the right side are recalculated as explained on the previous page) we should diminish these shares more than proportionally: ‘S x 76 /(75.95 + 0.84 + X)’ = some of the shares of the lower subgroups would not diminish and some would slightly increase [with the two-step-adjusting: first the shares should be proportionally increased (x 1.0221) and only then the adjusting to 75.95+0.84 ought to be executed], so the 0.84 value would change only very delicately. |In such a variant there should be first adjusted the average age of death in the DIC group due to moving a little of deaths from the 65-84 age range, but the change would be vsymbolic, because the closer to 76 years an average age of a subgroup is the smaller an adjusting value (towards Zero; with the same %-difference) can be.
/Injury-like preventable medical errors are not included in the above calculations because their precise age-structure is unknown and their potential effect is very small and unsure./
[If we had a huge subgroup of decedents (in the given year) and its age-structure was very similar (close to identical) to that of all deaths (after eliminating all injuries and infant mortality) and the subgroup was (before all deaths) of the standard age-state-of-health then no new killing factor could be common exclusively in that subgroup; any new and common killer should have considerably diminished the average age of death! There cannot be any higher adjusting (than calculated above), because saying that ‘a 70-year old man’ would otherwise live yet longer would already mean that Covid-19 “chooses” to kill stronger/strongest ones.]
…Thus, the total value of the upwards adjusting for our group (CTINI) is:
(0.70 + 0.45 + 0.40 + 0.25 + 0.10 – 0.10 + 0.10)* + 0.10** + 0.85 = 2.90 year
/*the summary value above is revised up by 0.05 because finally available data for 2019 show the number of injury-deaths increased by some %; **assessed subjectively/
So: 74.0 + 2.9 = 76.9 years
However, there is also one factor that in turn could force the average age of death (for the CTINI group) to be adjusted downwards. This is the state of health factor.
The share of people without a chronic condition drops to its minimum at age 75, and next, at age 85 this share is the same (not falling more), according to the Canadian data (CIHI.ca 2011). There are studies [6,7] according to which people who do not abuse alcohol +do not smoke +are physically active +eat healthy live on average 9-10 years longer than the US average is, being free, in a majority, of chronic conditions. A similar effect was signaled in other developed countries [8,9]. …The CDC revised (noted in very early April 2021) the average number of underlying conditions to 4.0 for 94% with conditions in the ‘deaths involving Covid-19 group (the number of conditions and life expectancy are strongly correlated -please read the Discussion part!). However our previous assumptions must be revised. -The small % of condition-free ones in the ‘deaths involving Covid-19’ group could have been, by a very big part, the result of taking now into account conditions like obesity and hypertension which are extremely common in the U.S. [10,11,12]. Additionally, in fact, people with 0 conditions have life expectancy, on average, by little bigger than people with 1 condition . …But what should the average number of conditions be in the U.S. society-cluster with the same age-structure (the comparative group)? That number could not be noticeably smaller (and so the average health status could not be better) in the “deaths involving Covid-19” group than in the comparative group because that group is really huge and Covid-19 cannot selectively infect and then kill persons with lower numbers of conditions (= healthier). However this number could be bigger due to killing by Covid-19 weaker ones of already infected persons. If a person has a few conditions (of the Chronic Condition Warehouse list) then what matters much for life expectancy is the pure number of conditions .
A potentially increased, by a limited value, average number of CCW-conditions (e.g. 3.5 vs. 3.0 …if a bit lowered 3.0 was the norm for the comparative group) would have less negative effect on life expectancy to having the same initial and then increased number by everyone (the theoretical 3.5, because practically a person cannot have 3.5 conditions, but 3 or 4), in the proportion 8.5 to 10 here (rough estimate on the basis of ‘Table 1’ and ‘Table 2’ ). We could process and extrapolate data from the tables, with the help of a life table, to estimate what time to deduct (A) from an average age of death for a statistical 59-year-old individual if he then lived with the conditions (3.5) for 24 years, and the time to deduct from an average age of death for a statistical 67-year-old individual, if he then lived with the number of conditions for 19 years, is 1.1 year. According to the British data guideline  the crude %-increase in multimorbid patients is stable from age 55 till 85 years and next the %-increase quickly slows down. At the same time, in the U.S., the prevalence of 2+, 3+ and 4+ chronic conditions (not fully of CCW but the key proportions are similar) in a group of an average age of 55 (45-65) is already approximately: 77%, 62% and 47%, respectively, of that at age 65+ according to another guideline . We could concentrate on 4+ conditioned, because the lower summary values are always much ahead and 3 is rather neutral here, on the fact that actuarial life expectancy at age 65 is 19.4 years , and initially on the average age of a potential victim of Covid-19, to correct the result for our group (CTINI) next (D).
0.47 + [4 /(84.4 – 55)] x (1 – 0.47) = 0.542 …so:
0.542 x [(76 – 59) /24] x (A=1.45) + 0.53 x [(76 – 55) /29.4] x 0.5 x [(10.5* /9.0) x (9.0 /19)] x 1.10 =
0.5567 + 0.115 = 0.6717 (year) …D1 = 0.6717 x 0.85
/D1 –the approximate value by which the exemplary increased number of CCW-conditions should have diminished the average age of death (to 76 y.) in the “deaths involving Covid-19” group/
/*10.5 is the distance between 76 and 65.5; 65.5 = (76 – 55) x 0.5 + 55/
However the so far result for our group (CTINI) is 76.9 years what means our group would have delicately more time for chronic conditions to show their negative effect [76.9 > 76 + 0.67] and so the preliminary result should be then slightly closer to 0.7 year. …Any preliminary result should be multiplied by the mentioned 0.85, so we would finally receive about 0.575.
/The above 0.575 year value is the result from our example. The estimator is based on some minor assumptions and on 21 conditions of the 2008-CCW list so in fact describes the negative influence on the average age of death of increasing the average number of conditions from >3.0 to >3.5, by >0.5 (because there are more conditions than of the 2008-CCW list)./
b) Since people from the “deaths involving Covid-19” group were allegedly killed by Covid-19 (accelerated deaths), it means that without its ‘intervention’ these people should still live. Thus, we calculate the average further life expectancy for the people from the whole DIC group if they were alive. We plot their age-of-death structure plus shares of women and men on the ‘actuarial Life Table’ . Initially we base on median values from age-subgroups and then taking into account weights of those age-subgroups we calculate the initial result for the whole group. The very careful calculations give the result of 12.25 year but it must be then udjusted up because the average age of death, for the DIC group, was adjusted down from 76.6 to 76.0 -what gives 12.65 year; and has also to be revised upwards because our group consists of those who could not die (if to be included into the group) because of fully external causes. For each mentioned category we must calculate the still existing, after forming by the deceased the “deaths involving Covid-19” group, potential length of life diminishing effect (X). For example, there are many people in that group at age 45-75 which could otherwise be important in number victims of lethal ‘Poisoning’. The calculation is the sum of the partial ones (Xn) for different age ranges (including 75++ too):
1.0 x [78.5 + Xn x (Sn /SN) /(Cn /CN)] – 0.0218 x (Pn /PN) x LE = 78.5
Xn x (Sn /SN) /(Cn /CN) = 0.0218 x (Pn /PN) x LE
Xn = LE x 0.0218 x (Pn /PN) x (Cn /CN) /(Sn /SN)
Xn -the potential length of life diminishing effect for an ‘n’ age range in the “deaths involving Covid-19” (DIC) group
Pn -the number of Poisoning victims in an ‘n’ age range, PN -the number of all Poisoning victims (in the year)
Cn -the number of people in an ‘n’ age range of the DIC group; CN – the whole DIC group size.
Sn, SN -the same as above (C) but in the whole society
LE -life expectancy (average) of a victim from an ‘n’ age range or at least life expectancy at the average age
78.5 –the life expectancy in the US according to the World Bank (2018)
/The imputed 0.0218 value in the above estimator assumes that the percent of people dying in the given year from the injury (‘Poisoning’ in this example) is similar to the percent of people dying, at all, from the injury. The difference would be very small and the results received below are so little so this assumption cannot influence the final result of the analysis in any meaningful way, whereas saves our time./
We must repeat the same kind of calculations with all of the mentioned (in the –a- part) categories. After that, the results concerning different categories must be summed up all together. All needed data, concerning age ranges of victims of different types of injury, are in tables and charts of https://injuryfacts.nsc.org. The calculations gave the following final values (the same order like in the -a- part) to sum up:
0.25 + 0.2 + 0.15 + 0.05 + 0.05 + 0.05 + 0.05 + 0.05** = 0.85
If there was any ‘D value’ (please go back to the end of the -a- part) we would have to calculate:
(76.9 – D) /76.9 = R
(76.0 + 12.65 + 0.85) x R – 76.0 = ALE
/ALE – adjusted life expectancy, assuming keeping the health-proportion/
/ If ‘D’ meant increasing the average age of death then we would have to add it to 77.0/
Because the whole “deaths involving Covid-19” group is huge (size =363 thousands) the health status of persons soon forming the DIC group (just before they got infected), could have been, on average, only:
a) very similar to …or
b) worse than that of the comparative group
/The comparative group is the U.S. society-cluster with the same (as of the DIC group) age-structure and of the standard, with this structure, health status./
…Let’s assume (for a while) the D value to be Zero (it means we assume, for a while, the health status of persons soon forming the “deaths involving Covid-19” to be, on average, identical to that of the comparative group), so we take the value of 12.65 + 0.85 = 13.50 years for the further analysis. …But why, for example, for the age of 76 a still alive person should live, on average, for over 11 more years (‘life table’)? Because some people die being (much) younger, and any person aged 76 is the one who is lucky to still live. Those who died being younger lower the average ‘length of life’ and the still living will increase it. The average ‘length of life’ and the average ‘summary length of life expected at a given age’ are equal only at birth.
c) What are the conclusions so far and what next?
-If 100% of people died due to “aging” (the CTINI group), the average age of death should be about 76.9 years. Some people had their worst health status even at age 90 or more while at the same time some people had their worst possible health status at age 60 or less. …Covid-19 should accelerate deaths of people.
-At the same time, if Covid-19 killed all persons of the “deaths involving Covid-19” group then it means that without the virus ‘intervention’ all of them should be still alive, for the next 13.50 years on average! It also means that each individual genuine Covid-19 related death shortened its victim personal life, on average, by 13.50 years*
/*with our temporary assumption of the health status of persons soon forming the DIC group (just before they got infected) to be, on average, identical (or just very similar) to that of the comparative group/
/It is nonsensical to believe that Covid-19 selectively kills, at typical for people to die ages (= with a similar to normal age-structure of deaths), only titans of health who would otherwise live to an average age of 89.50 years !/
-Persons from the ‘deaths involving Covid-19’ group died at an average age of about 76.0 not of 76.9, so there is the 0.9-year loophole (if to take exactly 76.0 and 76.9) caused probably by lethal effects of Covid-19, which is a life-shortening factor.
/We do not take into account potential influence of misdiagnoses, e.g. Covid-19 instead of the flu).
The average expected approximate contribution of a single individual genuine Covid-19 related death to the size of this gap is as follows:
[13.50 – (LEWII – 76.9)*] x 1 /N
/’N’ is the size of the entire group = 363,000/
/LEWII -life expectancy without impacts of injuries and infant mortality/
/*Each genuine Covid-19 death shortened its victim life, on average, by 13.50 years (in a variant with a standard health status), but not versus the average age of death in the given year for the CTINI group, but versus the average expected age of death of a person at all (so not in a given year); the first value is lower due to current demographic processes, e.g. immigration of younger mobile people./
We can count negative impacts of different injuries on the average at-birth life expectancy with the estimator (the example for ‘Poisoning’):
1.0 x (78.5 + P) – 0.0218 x LE = 78.5
78.5 + P – 0.0218 x 37.54 = 78.5
P = 0.0218 x 37.54 = 0.8184
/LE -life expectancy at age 43.5 years (43.5 = the average age of a victim of lethal poisoning) /
All the results are (the order the same like in the -a- part):
0.80 + 0.5 + 0.45 + 0.3 + 0.2 + 0.1 + 0.15 + 0.15** = 2.65
The negative impact of infant mortality on the average at-birth life expectancy is 0.54 year . Without injuries (not to repeat them): 0.56 x 69 /73 = 0.53. …So:
[13.50 – (LEWII – 76.9)] x 1 /N = [13.5 – (81.7 – 76.9)] x 1 /N = 8.70 x 1 /N
The total Covid-19 contribution to the size of the gap cannot be more than the gap itself is. Let’s count exactly:
C x 8.7 /N = 0.9
/‘C’ is the number of real/genuine Covid-19 related deaths/
C = 0.9 x (N /8.7) = 37.55
C /N = 0.9 /8.7 = 0.1034 (= 10.34%)
/’C/N’ –the share of real Covid-19 related deaths in the “deaths involving Covid-19” group in the U.S./
So, if the health status of persons soon forming the “deaths involving Covid-19” group was, on average, identical to that of the comparative group then only about 10% of those of the official DIC group died from Covid-19 complicity and all the rest were already in their terminal states and would have died in the same (or close to identical) time anyway, also without the Covid-19 infection, because their deaths resulted only from the normal age structure of deaths in the United States and from causes/conditions already existing before Covid-19, creating the actual average age of death. [It is however possible, strictly conditionally, that some of the rest (about 90%) had their deaths earlier by a number of days (or maybe even weeks); but this number must be (on average) very low if not to result in a noticeable overstating influence, already incorporated in the just calculated number of deaths, visible in the yearly statistics. This number is the maximal one (>37 thousands here) with the assumption of those 90% not having their deaths earlier at all –it can be explained more on request (= the factor of “skips” of some death-dates through the border between the years, from 2021 to 2020, producing a hidden, although rather limited, overestimate of the result).]
/It can be added that the average (long-term) age of a supposed flu victims is by some years lower than of Covid-19./
But what would the result be if the health status of persons soon forming the “deaths involving Covid-19” group was meaningfully worse (the average number of conditions was bigger) than that of the comparative group, with the same age-structure? If the average number of conditions was higher by >0.5 (of the CCW list) then the average age of death in the CTINI group, according to the example in the end of the -a- part, would not be 76.9 but about (76.9 – 0.575) = 76.325 years. Then ‘the loophole’ could be only 0.325 year.
(76.9 – 0.575) /76.9 = R = 0.9925
(76.0 + 12.65 + 0.85) x 0.9925 – 76.0 = ALE = 12.83
12.83 – (81.7 x 0.9925 – 76.325) = 8.0677
C = 0.325 x (N /8.05) = 14.66
C /N = 0.325 /8.05 = 0.0404 (= 4.04%)
So if the state of health of persons soon forming the DIC group* was meaningfully worse than that of the comparative group then with their* very high average age of death (76 years) the share of real Covid-19 related deaths would be only yet lower! …But the second variant is almost impossible, because those 4% should be the main contributor for the whole DIC group having its average number of underlying conditions increased, so persons of that 4% should have, on average, even 15 conditions. …Thus we think that the health status of both groups should be more like similar one to the other and the upper result should be much closer to the reality. …From the work of DuGoff EH et al. we know that if a person has a few conditions of the CCW list then what matters much for life expectancy is their pure number; the leading causes of chronic disease death give some differences in life expectancy at age 67, but the differences considerably diminish with morbidity and/or with increasing age .
The ‘intrinsic loop’
Some of patients with other diseases are not provided with immediate help because access to treatment for the diseases that most contribute to deaths (cardiology, oncology and lung diseases) has worsened with the pandemic in many countries. Some of hospital clinics have been closed due to revealed Covid-19 outbreaks. There are also people who are afraid of going to a specialist or to the hospital because of their apprehension of becoming Covid-19 infected there (panic). Covering the face with a mask enables the creation of a dangerous concentration of microorganisms and a statistical mask user probably do not change it often enough to limit that problem; besides, masks decrease O2- and increase CO2- concentrations under it. Staying at home means limited physical activity what is negative for overall health. When a number of people die because of these reasons earlier that they otherwise would, they reduce the average age of death which should be used in the analysis. These factors role will be growing over time.
Influenza and Pneumonia
The flu reported numbers of cases, even up to 90-95%, diminished in the world in the year 2020. That fact was already visible in the very beginning of the Covid-19 appearance . Maybe a number of the flu cases were treated as Covid-19 in that year due to limited reliability of the tests, or maybe there is another explanation.
Comparative joint counting of Covid-19, influenza and pneumonia-without-Covid-19 lethal cases is necessary because when looking at the CDC data: “Deaths involving coronavirus disease” we can see that virtually all cases of “Deaths involving Covid-19 and Pneumonia” are further claimed to be Covid-19 lethal victims. Also, in the UK when influenza, pneumonia and Covid-19 were on a Medical Certificate Cause of Death (MCCD) together, without a postmortem, then almost 96% of these deaths were counted as Covid-19 deaths, according to the analysis .
In the U.S., over 62%, about 48%, 34%, 23% and 15% of persons aged 67+ have, respectively, 3+, 4+, 5+, 6+ and seven or more conditions of the CCW list (of included in 2008), but only >2% have ten or more . Some useful info adds the rand.org study: “Multiple Chronic Conditions in the United States” also . But the prevalence of 2+, 3+ and 4+ chronic conditions is approximately: 2.4 times, five times and close to ten times, respectively, greater at age 65+ than at age 20-44; at the same time, when comparing to a group of 45-64 years of age, this prevalence is approximately 1.3, 1.6 and 2.1 times, respectively, greater at age over 65 .
The number of chronic conditions and life expectancy are very strongly correlated; the average number of chronic conditions would have to be ≥ 10.0 (!) to diminish life expectancy to 80 years for a still alive 75-year-old US woman, what means shortening the remaining life to five years; at the same time a 75-year-old woman with “only” 5.0 chronic conditions should live, on average, to the age of 87, what is by one year shorter than the average for a 75-year-old woman in the US . The marginal decline in life expectancy increases with an additional chronic condition when the number of conditions is low, but this decline starts with low values -first conditions sum up to a much less effect on life expectancy than the next conditions do; at the same time selected conditions give differences in life expectancy at age 67, but the differences considerably diminish with morbidity and/or with increasing age ! The clear relationship between the number of comorbidities and life expectancy has been discovered also by other authors .
…It is now said that Covid-19 can cause even acute strokes and acute myocardial infarctions himself [e.g. 20,21]. What concerns different possibly mortal chronic conditions, most of them have a very similar advanced average age of a decedent. That age is meaningfully lower only due to the conditions like: HIV (<60 y.), malignant neoplasm of cervix uteri (60 y.), obesity and chronic liver diseases (>60 y.) [3-p.39-40].
The limited average number of conditions in the “deaths involving Covid-19” group deny the theory that this group was of a meaningfully worse than standard health status. Additionally, further selective increase in the average number of chronic conditions could only diminish the share of real Covid-19 deaths in the officially announced ‘Covid-19 related deaths’ because ‘the loophole’ would collapse faster than potential life expectancy!
If a person dies with Covid-19 in no way it means that Covid-19 kills him. Assuming in advance that a person is a victim of Covid-19 only because has died with Covid-19 or with a positive PCR test result is silly and worthless, almost as silly as saying that if someone wearing glasses died then wearing glasses killed him.
a) The ‘ex post’ analysis is necessary to discover the real number of deaths due to Covid-19.
b) The official number of ‘Covid-19 deaths’ is mainly the result of the double counting of those who would die whatsoever in the same (or close to identical) time, also without the Covid-19 infection, because Covid-19 started to accompany people already being in their terminal states …or where there were not any Covid-19 but only positive PCR test results. So in the US in the year 2020 there were not 363,000 ‘deaths involving Covid-19’ but only up to 40,000 of that. The rest of the deaths should be treated as wrongly attributed to Covid-19.
c) The genuine Covid-19 mortality is very close to zero (if to be based on antibody tests).
d) Hardly any excessive deaths year-over-year are due to Covid-19. Main reasons of excessive deaths most likely are:
-the worsened access to treatment for diseases other than Covid-19
-some of patients’ fear of going to a specialist or to the hospital (panic)
-‘deaths of despair’.
e) It can be supposed that one of important reasons of the official numbers of ‘Covid-19 deaths’ being hugely overestimated is including those who have had only a positive PCR test result (even 2 months prior to the death, like in the US or in the UK).
f) Comparative joint counting of Covid-19 + influenza + pneumonia-without-Covid-19 lethal cases is necessary.
g) The appearing of Covid-19 in 2020 (taking into account the simultaneous strong disappearance of the flu) made no considerable increase in the number of yearly deaths due to respiratory viral infections (when compared to previous years). So at the same time it means that limiting the Covid-19-occurrence should not have any considerable diminishing effect on the number of yearly deaths (+ the flu should recover then).
h) The analysis for the year 2020 is most informative. Later analyses can, in fact, incorporate also some health troubles produced by vaccines, and even effects of mutations of Covid-19 –if forced by the vaccines.
1) If construction of any estimation is unclear to a reader it will be explained in details on request. Only the estimator from the end of the -a- part is a bit simplified and the rest of them is maximally accurate.
2) The new CDC document appeared . We do not use it because it does not look much credible. The number of “Covid-19 related deaths” was revised up to 378 th. (compared to the data from very early January 2021 -363 th.), but the estimated average age of a decedent went up by almost 0.5 year. It would be only possible if those added 15 th. died at age almost 90 on average. Also, the CDC gives a much increased number of all deaths in the society (2020 vs. 2019) and the decreased average age of death. If the data were real it would mean that the ‘intrinsic loop’ and ‘deaths of despair’ killed additional >450 thousands of people, in 2020. But we think that preliminary data are by a big part the result of simple adding in an unwise way.
Cunningham RM, Walton MA, Carter PM: The Major Causes of Death in Children and Adolescents in the United States. N Engl J Med 2018; 379: 2468-2475
Li Yanping et al.: Impact of Healthy Lifestyle Factors on Life Expectancies in the US Population. Circulation 2018; 138:345–355
Li Yanping et al. : Healthy lifestyle and life expectancy free of cancer, cardiovascular disease, and type 2 diabetes: prospective cohort study. BMJ 2020; 368: 16669
O’Doherty MG et al.: Effect of major lifestyle risk factors, independent and jointly on life expectancy with and without cardiovascular disease. Eur J Epidemiol 2016; 31: 455-468
Nusselder WJ, Franco OH et al.: Living healthier for longer. BMC Public Health 2009; 9: 487
DuGoff EH, Canudas-Romo V et al.: Multiple chronic conditions and life expectancy. Medical Care 2014, 52/8: 688-94
Zhu Yajing et al. : Characteristics, service use and mortality of clusters of multimorbid patients in England: a population-based study. BMC Medicine 2020, 18, nr: 78
King DE, Xiang J, Pilkerton CS: Multimorbidity Trends in United States Adults. J Am Board Fam Med 2018, 31(4): 503-513
Nunes BP, Flores TR et al. : Multimorbidity and mortality in older adults: A systematic review and meta-analysis. Arch Gerontol Geriatr 2016; 67: 130-138