A short video about the central role of the eugenics movement in the development of modern statistics by Karl Pearson, Ronald Fisher, and their colleagues.
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
This is a short post on the critical distinction between the Case Fatality Rate (CFR) of a disease such as the SARS-COV-2 coronavirus thought to cause COVID-19 and the Infection Fatality Rate (IFR), also sometimes known as the actual mortality rate or lethality. This remains a source of confusion and perhaps deliberate obfuscation several months into the crisis.
The case fatality rate or CFR is the number of deathsattributed to the disease usually among those diagnosed with the disease divided by the number of diagnosed casesaccording to some diagnostic criterion, for example a “positive” RT-PCR (Reverse Transcriptase-Polymerase Chain Reaction) test.
The infection fatality rate or IFR is the number of deaths attributed to the disease divided by the actual number of people infected, which generally includes mild or asymptomatic infections that are not diagnosed. Most diseases have many mild or asymptomatic infections. This is not unusual and is the case for the coronavirus SARS-COV-2.
The CFR generally reflects those with more serious infections who seek medical attention, go to a hospital emergency room, etc. It is generally biased, usually higher than the IFR for most diseases, and also can vary a lot depending on the availability of tests and on other causes unrelated to the genuine lethality of the disease.
A disease that kills everyone who exhibits symptoms and no one who has no symptoms even though actually infected can have a case fatality rate (CFR) of 100 percent and and an infection fatality rate (IFR) of nearly 0.0 percent.
For example, an exotic disease that produces distinctive green and purple spots in those it kills — easily identifiable even without advanced tests like RT-PCR — but in fact kills only 100 people out of a United States population of 330 million even though most are infected for some reason.
Although there are a number of subtleties in the definition and computation of these numbers that I have omitted for clarity, the infection fatality rate (IFR), ideally broken down by age, medical conditions, and other risk factors, is key to evaluating the proper public health response to an outbreak of an infectious disease. Not the case fatality rate or CFR.
(C) 2020 by John F. McGowan, Ph.D.
About Me
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
Article at statnews.com by Vinay Prasad and Jeffrey S. Flier
(C) 2020 by John F. McGowan, Ph.D.
About Me
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
A key issue in the current coronavirus COVID-19 pandemic is how it compares to “seasonal flu,” “flu,” “influenza” or “pneumonia and influenza,” terms that are often used interchangeably but have different definitions or implied definitions in different contexts.
An overlapping key issue is whether false positives from the coronavirus RT-PCR test or diagnoses without the test (e.g. lung x-rays or examination of a patient by a doctor) has attributed a substantial number of conventional pneumonia deaths and heart attacks to the COVID-19 coronavirus.
Remarkably, the United States Centers for Disease Control (CDC) appears to use two different counts (or model outputs?) of annual deaths from “influenza and pneumonia” or “pneumonia and influenza” that differ by over a factor of three.
The “Deaths: Final Data for 2017” report (Page Six, Table B) lists “influenza and pneumonia” as the eighth leading cause of death with 55,672 deaths in 2017.
Influenza and pneumonia (8th Leading cause of death in Deaths: Final Data for 2017)
In contrast, the Weekly Pneumonia and Influenza (P&I) Mortality Surveillance lists over 180,000 deaths from “pneumonia and influenza” (mostly pneumonia) in 2017 in the data files on the site apparently used to generate the FluView plot displayed. The weekly surveillance number provides a much larger pool of potential false positives than the more widely quoted number of about 50,000 “flu” deaths per year.
Influenza: a study in contemporary medical politics by Peter Doshi
This is a well written but long (312 pages) dissertation. All of it is relevant to the current pandemic crisis, but it is a lot to digest. The most important and most relevant to the current pandemic section is Chapter 4: False Assumptions: a Shaky Foundation for Consensus (Pages 151-212, including tables and figures).
Key topics discussed in detail in this chapter include:
The CDC uses models to assign many deaths to influenza (the influenza viruses) even though doctors rarely diagnose influenza, rarely list influenza as a cause of death on death certificates, and most laboratory tests of samples from patients with respiratory illnesses (often called Influenza Like Illnesses or ILI) do not confirm the presence of the influenza viruses and often identify other viruses such as rhinovirus, adenovirus, various coronaviruses, etc. as present instead. These models may even assign deaths listed on death certificates as heart attacks to the total.
The models differ from researcher and publication to researcher and publication, have changed dramatically over the years, notably a jump from 20,000 estimated “flu” deaths in 2002 to a widely quoted estimate of 36,000 in 2003.
The evidence that flu vaccines work is weak and contradictory.
Conclusion
There is a remarkable lack of key measurements in the current coronavirus COVID-19 pandemic. These include the actual mortality rate (aka infection fatality rate) broken down by age, sex, race, pre-existing medical conditions, ambient temperature, sunlight levels, pollution levels, and other risk factors. The false positive and false negative rates of the tests for the disease, both the tests for an active infection such as the RT-PCR tests and tests for past infection such as the antibody tests. The methods and rates of transmission for the disease. Aerosol transmission probably occurs at least at a low level and is virtually unstoppable.
The confusing language and numbers on pneumonia and influenza on the CDC web site and in various official reports and documents seem to be primarily for marketing the flu vaccines rather than enabling informed decisions by patients and doctors or supporting external scientific research into the influenza viruses or other diseases.
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
Santa Clara County’s Remarkably Low COVID-19 Numbers
Santa Clara County, California has remarkably low COVID-19 death numbers despite extremely close ties to China and the many missteps in the response to the pandemic. I discuss this in detail and the possible reasons for the low numbers.
This is a written version with slides and notes (a PDF file). It is usually faster to read the written version than watch the video. The video may be clearer and more detailed on a few points. References are included in the written version.
A key question about pneumonia and influenza (including the COVID-19 Sars-Cov-2 coronavirus disease) is the role of weather and temperature on the incidence, severity, and mortality rates from the diseases. Pneumonia and influenza cases and deaths are seasonal, peaking in the winter when sunlight levels and temperatures are lower. The curve for pneumonia and influenza deaths is roughly sinusoidal, which would be expected from something connected to sunlight levels. This is not what one would naively expect from children spreading the disease during the school year; we would expect an abrupt step up in the fall when kids return to school and a step down in the spring when school closes.
A previous post listed a number of resources on the possible direct effects of sunlight through vitamin D production and direct destruction of viruses and bacteria by sunlight, especially the ultraviolet component.
Seasonal Variation in Deaths from Pneumonia and Influenza (2014-2020)
There is a fair amount of research on the role of weather and temperature in the incidence, severity and mortality of pneumonia and influenza. Generally the research seems to support a positive correlation between lower temperatures and also rapid changes in temperature with higher incidence, severity and mortality from pneumonia and influenza as well as other causes of death, notably coronary disease. It also suggests extreme heat “heat waves” is positively correlated with pneumonia. Below are several scientific and scholarly articles; the interested reader can find more articles and details at PubMed (enter “pneumonia and weather” for example in the search box).
Note that correlation does not prove causation.
Scientific and Scholarly Articles
The Lancet
Volume 315, Issue 8183, 28 June 1980, Pages 1405-1408
Journal home page for The Lancet
Occasional Survey
THE WEATHER AND DEATHS FROM PNEUMONIA
Author links open overlay panelG.M.Bull
https://doi.org/10.1016/S0140-6736(80)92666-5
Age Ageing. 1978 Nov;7(4):210-24.
Environment, temperature and death rates.
Bull GM, Morton J.
Abstract
Analysis of recorded monthly deaths in England and Wales shows a close association of death rates with external temperature in most diseases other than the cancers. Analysis of daily deaths in England and Wales and in New York shows the following relationships between temperature and deaths from myocardial infarction, strokes and pneumonia. Between -10 degrees and +20 degrees C mimimum temperature there is a nearly linear fall in deaths as the temperature rises. Above 20 degrees C deaths rise steeply as the temperature rises and below -10 degrees C rise steeply as temperature falls. These associations of deaths with temperature are much stronger in the elderly than in younger subjects. Detailed analysis of the daily deaths in England and Wales from myocardial infarction, strokes and pneumonia show that short-term (1–2 days) temperature changes have little effect on death rates but medium-term (7–10 days) and longer-term (three or more weeks) changes associated with very significant changes in death rates. The three diseases vary in the time relations between temperature change and change in death rates. In all three there is an interval between the change in temperature and death and this is shortest in the case of myocardial infarction (1–2 days before death), longest in the case of pneumonia (about a week before death) and intermediate in the case of strokes (about 3–4 days before death). At low temperatures death rates increase as the duration of temperature change increases, while at high temperatures (but below +20 degrees C) death rates decrease as the period of temperature change is longer. The implications of these findings are discussed and it is postulated that there is probably causal relationship between temperature change and deaths from a wide variety of diseases. A proximal link in the chain is probably a failure of autonomic control of body temperature in the elderly leading to a change in body temperature and some humoral change which in turn leads to death. It is not appropriate to concentrate on hypothermia as the relationship between temperature and death is seen at all temperatures.
================================================= Age Ageing. 1975 Feb;4(1):19-31. Seasonal and short-term relationships of temperature with deaths from myocardial and cerebral infarction. Bull GM, Morton J. Abstract
In subjects over 60, changes in temperature lasting two or more days are associated with highly significant changes in death rates from myocardial infarction and cerbral vascular accidents. In both cases, the lower the temperature the higher the death rate and vice versa. Moreover the temperatures one to four days prior to the clinical onset of infarction are more relevant than that on the day of onset, a fact which may have a bearing on prophylaxis. In the case of strokes, a high temperature on the day of onset is also associated with an increase in deaths on that day. The relevance of these findings to possible mechanisms and prophylaxis is discussed.
Lancet. 1979 May 5;1(8123):982-3. No abstract available.
PMID:
87658
J Intern Med. 1991 Dec;230(6):479-85.
High coronary mortality in cold regions of Sweden.
Gyllerup S(1), Lanke J, Lindholm LH, Scherstén B.
Author information:
(1)Health Sciences Centre, Lund University, Dalby, Sweden.
The hypothesis that cold climate is associated with high coronary mortality in
Sweden is tested. Cold exposure was calculated in each of the 284 municipalities
of Sweden. There was a significant association between cold exposure and coronary
mortality in both sexes in all age groups. The strongest association was found in
men aged 40-64 years (coefficient of determination k = 0.39). The decile of men
aged 40-64 years who lived in the coldest municipalities had a 40% excess
mortality. A significant association was also found between cold exposure and
mortality from cerebrovascular diseases. We conclude that there is a strong
regional association between cold exposure and high coronary mortality.
DOI: 10.1111/j.1365-2796.1991.tb00478.x
PMID: 1748856 [Indexed for MEDLINE]
Author information:
(1)Husensjö Group Practice, Helsingborg, Sweden.
In many European countries there is a tendency towards higher coronary mortality
in the northern parts of the country. Furthermore the highest coronary mortality
rates are found in the colder parts of Europe. We studied the regional variation
in coronary mortality in the 284 Swedish municipalities during a ten-year period
and the relation to the cold exposure in each municipality during the same time
period.METHODS: Mortality rates for each municipality were acquired from the
death certificates and indirectly standardised against the country. Temperature
readings from measurements 5 times a day during daytime were used to form a cold
index. We also compensated for wind chill by using Siples wind chill index.
Multiple regression models were used. Second degree polynomials were used for the
explanatory variables.
RESULTS: There was a strong relation between the cold exposure in a municipality
and coronary mortality. The cold index alone could explain 39% of the regional
variation in coronary mortality. In a multiple regression model, cold index was
the strongest explanatory variable. The coronary mortality in the coldest decile
of the population was 40% higher than in the country as a whole.
CONCLUSIONS: There is a strong regional association between cold exposure and
coronary mortality in Sweden. However, in this type of study, it is not possible
to determine whether this association is a causal one or not.
Cold climate is an important factor in explaining regional differences in
coronary mortality even if serum cholesterol and other established risk factors
are taken into account.
Gyllerup S(1), Lanke J, Lindholm LH, Schersten B.
Author information:
(1)Health Sciences Centre, Lund University, Dalby Sweden.
Earlier studies have shown a strong regional association between cold climate and
coronary mortality in Sweden and that coronary mortality is more strongly
associated with cold climate than with other explanatory factors such as drinking
water hardness, socioeconomic factors, tobacco and sales of butter. To examine
the joint impact of these factors and to investigate regional differences in
serum cholesterol and their relation to cold climate and coronary mortality,
regression analyses were performed with 259 municipalities in Sweden as units.
Mortality from acute myocardial infarction in men aged 40-64 during 1975-1984 was
used as the dependent variable. A cold index was calculated, this index and the
above mentioned factors were used as explanatory variables. The main results
were: Cold index was the strongest factor when introduced into a multiple
regression model. Four other strong factors had to be used to obtain the same
explanatory strength as cold index did alone, and even when introduced as the
last factor, cold index increased the coefficient of determination substantially.
In a subsample of 37 municipalities, serum cholesterol was not significantly
associated with coronary mortality. However, there was a significant correlation
between cold index and serum cholesterol.
DOI: 10.1177/003693309303800604
PMID: 8146634 [Indexed for MEDLINE]
Am J Epidemiol. 2016 Oct 15;184(8):555-569. Epub 2016 Oct 6.
Pneumonia Hospitalization Risk in the Elderly Attributable to Cold and Hot
Temperatures in Hong Kong, China.
Qiu H, Sun S, Tang R, Chan KP, Tian L.
The growth of pathogens potentially relevant to respiratory tract infection may
be triggered by changes in ambient temperature. Few studies have examined the
association between ambient temperature and pneumonia incidence, and no studies
have focused on the susceptible elderly population. We aimed to examine the
short-term association between ambient temperature and geriatric pneumonia and to
assess the disease burden attributable to cold and hot temperatures in Hong Kong,
China. Daily time-series data on emergency hospital admissions for geriatric
pneumonia, mean temperature, relative humidity, and air pollution concentrations
between January 2005 and December 2012 were collected. Distributed-lag nonlinear
modeling integrated in quasi-Poisson regression was used to examine the
exposure-lag-response relationship between temperature and pneumonia
hospitalization. Measures of the risk attributable to nonoptimal temperature were
calculated to summarize the disease burden. Subgroup analyses were conducted to
examine the sex difference. We observed significant nonlinear and delayed
associations of both cold and hot temperatures with pneumonia in the elderly,
with cold temperatures having stronger effect estimates. Among the 10.7% of
temperature-related pneumonia hospitalizations, 8.7% and 2.0% were attributed to
cold and hot temperatures, respectively. Most of the temperature-related burden
for pneumonia hospitalizations in Hong Kong was attributable to cold
temperatures, and elderly men had greater susceptibility.
Influenza Other Respir Viruses. 2016 Jul;10(4):310-3. doi: 10.1111/irv.12369.
Epub 2016 May 17.
Cold, dry air is associated with influenza and pneumonia mortality in Auckland,
New Zealand.
Davis RE(1), Dougherty E(1), McArthur C(2), Huang QS(3), Baker MG(4).
Author information:
(1)Department of Environmental Sciences, University of Virginia, Charlottesville,
VA, USA.
(2)Auckland City Hospital, Auckland, New Zealand.
(3)Institute of Environmental Science and Research, Wellington, New Zealand.
(4)University of Otago-Wellington, Wellington, New Zealand.
The relationship between weather and influenza and pneumonia mortality was
examined retrospectively using daily data from 1980 to 2009 in Auckland, New
Zealand, a humid, subtropical location. Mortality events, defined when mortality
exceeded 0·95 standard deviation above the mean, followed periods of anomalously
cold air (ta.m. = -4·1, P < 0·01; tp.m. = -4·2, P < 0·01) and/or anomalously dry
air (ta.m. = -4·1, P < 0·01; tp.m. = -3·8, P < 0·01) by up to 19 days. These
results suggest that respiratory infection is enhanced during unusually cold
conditions and during conditions with unusually low humidity, even in a
subtropical location where humidity is typically high.
Environ Res. 2019 Feb;169:139-146. doi: 10.1016/j.envres.2018.10.031. Epub 2018
Oct 30.
Impacts of cold weather on emergency hospital admission in Texas, 2004-2013.
Chen TH(1), Du XL(1), Chan W(2), Zhang K(3).
Author information:
(1)Department of Epidemiology, Human Genetics and Environmental Sciences, School
of Public Health, The University of Texas Health Science Center at Houston,
Houston, TX 77030, USA.
(2)Department of Biostatistics and Data Science, School of Public Health, The
University of Texas Health Science Center at Houston, Houston, TX, USA.
(3)Department of Epidemiology, Human Genetics and Environmental Sciences, School
of Public Health, The University of Texas Health Science Center at Houston,
Houston, TX 77030, USA; Southwest Center for Occupational and Environmental
Health, School of Public Health, The University of Texas Health Science Center at
Houston, Houston, TX, USA. Electronic address: kai.zhang@uth.tmc.edu.
Cold weather has been identified as a major cause of weather-related deaths in
the U.S. Although the effects of cold weather on mortality has been investigated
extensively, studies on how cold weather affects hospital admissions are limited
particularly in the Southern United States. This study aimed to examine impacts
of cold weather on emergency hospital admissions (EHA) in 12 major Texas
metropolitan statistical areas (MSAs) for the 10-year period, 2004-2013. A
two-stage approach was employed to examine the associations between cold weather
and EHA. First, the cold effects on each MSA were estimated using distributed lag
non-linear models (DLNM). Then a random effects meta-analysis was applied to
estimate pooled effects across all 12 MSAs. Percent increase in risk and
corresponding 95% confidence intervals (CIs) were estimated as with a 1 °C (°C)
decrease in temperature below a MSA-specific threshold for cold effects.
Age-stratified and cause-specific EHA were modeled separately. The majority of
the 12 Texas MSAs were associated with an increased risk in EHA ranging from 0.1%
to 3.8% with a 1 ⁰C decrease below cold thresholds. The pooled effect estimate
was 1.6% (95% CI: 0.9%, 2.2%) increase in all-cause EHA risk with 1 ⁰C decrease
in temperature. Cold wave effects were also observed in most eastern and southern
Texas MSAs. Effects of cold on all-cause EHA were highest in the very elderly
(2.4%, 95% CI: 1.2%, 3.6%). Pooled estimates for cause-specific EHA association
were strongest in pneumonia (3.3%, 95% CI: 2.8%, 3.9%), followed by chronic
obstructive pulmonary disease (3.3%, 95% CI: 2.1%, 4.5%) and respiratory diseases
(2.8%, 95% CI: 1.9%, 3.7%). Cold weather generally increases EHA risk
significantly in Texas, especially in respiratory diseases, and cold effects
estimates increased by elderly population (aged over 75 years). Our findings
provide insight into better intervention strategy to reduce adverse health
effects of cold weather among targeted vulnerable populations.
Environ Res. 2014 Jul;132:334-41. doi: 10.1016/j.envres.2014.04.021. Epub 2014
May 14.
Impact of temperature on childhood pneumonia estimated from satellite remote
sensing.
Xu Z(1), Liu Y(2), Ma Z(2), Li S(3), Hu W(1), Tong S(4).
Author information:
(1)School of Public Health and Social Work & Institute of Health and Biomedical
Innovation, Queensland University of Technology, Brisbane, QLD, Australia.
(2)Rollins School of Public Health, Emory University, Atlanta, GA, United States.
(3)School of Public Health, Shanghai Jiaotong University School of Medicine,
Shanghai, China.
(4)School of Public Health and Social Work & Institute of Health and Biomedical
Innovation, Queensland University of Technology, Brisbane, QLD, Australia.
Electronic address: s.tong@qut.edu.au.
The effect of temperature on childhood pneumonia in subtropical regions is
largely unknown so far. This study examined the impact of temperature on
childhood pneumonia in Brisbane, Australia. A quasi-Poisson generalized linear
model combined with a distributed lag non-linear model was used to quantify the
main effect of temperature on emergency department visits (EDVs) for childhood
pneumonia in Brisbane from 2001 to 2010. The model residuals were checked to
identify added effects due to heat waves or cold spells. Both high and low
temperatures were associated with an increase in EDVs for childhood pneumonia.
Children aged 2-5 years, and female children were particularly vulnerable to the
impacts of heat and cold, and Indigenous children were sensitive to heat. Heat
waves and cold spells had significant added effects on childhood pneumonia, and
the magnitude of these effects increased with intensity and duration. There were
changes over time in both the main and added effects of temperature on childhood
pneumonia. Children, especially those female and Indigenous, should be
particularly protected from extreme temperatures. Future development of early
warning systems should take the change over time in the impact of temperature on
children’s health into account.
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
Weekly Pneumonia and Influenza Death Numbers for 2020 and 2019 Compared
The updated plot is from a LibreOffice spreadsheet which can be downloaded at the link below. LibreOffice is a free, open-source alternative to Microsoft Office. It is available for Microsoft Windows, Mac OS X, and most flavors of Unix. It can be downloaded here.
I made a mistake copying the column of pneumonia and influenza deaths from early 2019 (weeks 1-13 of 2019 and weeks 50-52 of 2020) with LibreOffice Calc (the spreadsheet). The spreadsheet copied the macros in the cells instead of the number values. These macros were then applied to the columns to the left of the copied column, giving incorrect values that exaggerated the excess of deaths in comparable weeks last year (2019).
The actual excess is 1,841 more deaths in 2019, not about 6,000.
I was using the spreadsheet to make the results more accessible to a general audience. The copying error was consistent with the results from the more in-depth Python data analysis. In retrospect I should have checked the spreadsheet numbers more carefully.
Discussion
This does not change the conclusion that there is no sign of COVID-19 in the numbers until March 14, 2020 and a weak rise consistent with normal fluctuations in the weekly numbers in the final two weeks (March 14-28, 2020, weeks 12 and 13). It does reduce the size of the discrepancy between the two years. It remains possible that all the about 1,600 COVID-19 deaths reported as of March 28, 2020 could be conventional pneumonia and influenza deaths labeled as COVID-19 due to false positive RT-PCR tests and other misdiagnoses.
As I have discussed, there are strong reasons to doubt the CDC numbers. The most egregious I have found so far is the remarkable difference between the about 55,000 deaths from “influenza and pneumonia” in the leading causes of death tables (Table B, Page Six) and the about 188,000 deaths from “pneumonia and influenza” in the NCHSData14.csv file and other NCHSData<Week Number>.csv files.
An educated guess is that the 55,000 deaths from “influenza and pneumonia” is the output of a model the CDC uses to estimate the number of deaths directly or indirectly caused by “influenza viruses.” In the weekly pneumonia and influenza death numbers, the vast majority of deaths are listed as pneumonia and not the separate “influenza” category. Thus about 130,000 deaths appear to have been assigned to other categories in the final deaths for 2017 report, possibly “chronic lower respiratory diseases” which is the fourth (4th) leading cause of death. This is however a theory and CDC should carefully clarify what they are doing.
Accordingly, it is difficult to know what pre-processing or modeling/estimation may have been applied to the weekly pneumonia and influenza death numbers, although the commentary on the CDC web site implies these numbers are counts of death certificates and the causes of death on death certificates reported to the CDC by state and local authorities.
I am looking through the NCHSData<Week Number>.csv files to see how complete they may actually be. The FluView web page contains a table that seems to imply that all weeks except the very last week in the file are complete or almost complete. They use the label “> 100%” where > is presumably “greater than”. Of course, 100 percent usually means complete.
CDC “Percent Complete” Table (Misleading language at best)
There are many possible reasons for COVID-19 deaths not showing up in the weekly pneumonia and influenza death numbers before March 14, 2020 despite the Chinese coverup in December and early January, the US testing fiasco, the 430,000 visitors to the United States from China since the coronavirus surfaced, and the many asymptomatic carriers now being detected. These different possible reasons have different, even opposite in some cases, implications for public health policy.
Possible reasons include:
Despite the many problems above, the public health authorities have been remarkably successful in identifying nearly all COVID-19 deaths up to March 14, 2020. This seems too good to be true, but cannot be excluded.
The infection fatality rate (aka actual mortality rate) of the COVID-19 coronavirus is much less than early numbers such as 3.4 percent from the World Health Organization (WHO) or the 0.9-1 percent used by various authorities. Iceland, South Korean, Denmark and German data suggest about 0.5 percent mortality rate – which still could be higher than real rate.
Many COVID-19 deaths are due to aggressive treatment of the disease, e.g. intubation, rather than the disease alone.
The weekly pneumonia and influenza death numbers are substantially incomplete, due to normal delays or due to unusual delays associated with the crisis.
There has been a compensating drop in non-COVID pneumonia and influenza deaths due to shelter-in-place and taking it easy. Elderly and susceptible persons may have taken precautions in January and February due to the publicity, even before the shutdown in mid March.
Something else
Some combination of some or all of the above!
Conclusion
There is a remarkable lack of key measurements in the current coronavirus COVID-19 pandemic. These include the actual mortality rate (aka infection fatality rate) broken down by age, sex, race, pre-existing medical conditions, ambient temperature, sunlight levels, pollution levels, and other risk factors. The false positive and false negative rates of the tests for the disease, both the tests for an active infection such as the RT-PCR tests and tests for past infection such as the antibody tests. The methods and rates of transmission for the disease. Aerosol transmission probably occurs at least at a low level and is virtually unstoppable.
It is important to collect this data and measure these key parameters as quickly as possible in an open, “transparent” manner with multiple independent teams, not all funded or controlled by the CDC, as soon as possible to make good decisions based on knowledge and data, rather than fear, ignorance, and the primal fight or flight response.
(C) 2020 by John F. McGowan, Ph.D.
About Me
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
In coverage of the coronavirus COVID-19 pandemic, one often sees a value for the infection fatality rate (also known as the actual mortality rate, which is different from the “case fatality rate”) of 0.1 percent, meaning one in 1000 people infected by the “flu” dies. Infected includes people who are asymptomatic, have mild cases — anyone who is actually infected even if never detected. It is often explicitly or implicitly argued that if the infection fatality rate of COVID-19 is only 0.1 percent as suggested by a recent study by Stanford researchers (https://www.medrxiv.org/content/10.1101/2020.04.14.20062463v1.full.pdf) we can relax and go back to work. Unfortunately, it is probably not that simple.
We can use our prevalence estimates to approximate the infection fatality rate from COVID-19 in Santa Clara County. As of April 10, 2020, 50 people have died of COVID-19 in the County, with an average increase of 6% daily in the number of deaths. If our estimates of48,000-81,000 infections represent the cumulative total on April 1, and we project deaths to April 22 (a 3 week lag from time of infection to death22), we estimate about 100 deaths in the county. A hundred deaths out of 48,000-81,000 infections corresponds to an infection fatality rate of 0.12-0.2%. If antibodies take longer than 3 days to appear, if the average duration from case identification to death is less than 3 weeks, or if the epidemic wave has peaked and growth in deaths is less than 6% daily, then the infection fatality rate would be lower. These straightforward estimations of infection fatality rate fail to account for age structure and changing treatment approaches to COVID-19. Nevertheless,our prevalence estimates can be used to update existing fatality rates given the large upwards revision of under-ascertainment
COVID-19 Antibody Seroprevalence in Santa Clara County, California by Bendavid et al
It is thought the vast majority of adults get at least two symptomatic “colds” or “flus” in common usage (The CDC claims adults get 2-3 “common colds” per year and children more on their web site which matches common experience.). These are caused by a wide variety of viruses and bacteria and sometimes chemical toxins. These include the rhinovirus, various coronaviruses other that the “novel” SARS-COV-2 coronavirus, and many others including a category of viruses known as “influenza” or “influenza viruses”.
With a total US population of about 330 million, we can estimate at least 660 million individual cases and separate infections of these “cold” or “flu” organisms (either viruses or bacteria) each year. This gives a naive effective infection fatality rate averaged over the population and different diseases of:
188,000 divided by 660 million is: 0.028 percent (0.00028484848484848485)
55,000 divided by 660 million is: 0.008 percent (0.00008333333333)
This is of course much less than 0.1 percent (one in 1000).
What gives?
In common English usage, the terms “cold” and “flu” are often used interchangeably. The use of the terms “flu” and “influenza” to describe respiratory illnesses that vary in incidence seasonally predates the discovery of the influenza viruses, a category of viruses that can cause these symptoms. Influenza is Italian, from the Latin “influentia,” for “influence,” referring to the baleful influence of the stars that the ancients blamed for the disease.
The CDC hopelessly blurs the distinctions, if any, between “common cold”, “cold”, “flu”, “influenza”, “influenza like illness,” “influenza associated,” “pneumonia,” and other terms in its promotional and “scientific” materials.
Influenza as in the influenza viruses is rarely listed as a cause of death on death certificates. The weekly “pneumonia and influenza” death numbers from the National Center for Health Statistics (NCHS) only list about 8,000 deaths from influenza in 2017. The CDC cites several different reasons for claiming there is massive underdiagnosis and underreporting of influenza (THE VIRUS) deaths, dating back to at least 2005 and persisting despite the CDC’s extensive educational efforts.
The CDC uses a mysterious model to estimate about 55,000 annual deaths from influenza (THE VIRUS). Presumably the number of deaths from “influenza and pneumonia” in the leading causes of death is this number or something closely related — but this is not clear. Incidentally, in this age of the Internet and pervasive computing, the CDC could publish the actual source code for their model in a free open-source language such as Python on their web site for all to see and review.
CDC Graphic About Influenza Model (There is a HUGE correction for under-detection)
Part of this model is an estimate of how many “colds” are caused by an influenza virus. Presumably this number is about 55 million to get the widely quoted 0.1 percent (one in 1000) infection fatality rate for influenza. This is an example of the CDC’s estimates from https://www.cdc.gov/flu/about/burden/index.html:
Thus, the CDC estimates about 55 million of the annual over 660 million “cold” cases in the United States is caused by “influenza disease” or “influenza” or “flu,” presumably meaning cases caused by influenza viruses. This is probably less than ten percent of all “colds.” The CDC also estimates about 55,000 deaths from influenza viruses. This presumably gives the about 0.1 percent (one in 1000) number widely quoted in the media.
Everyone should understand that a 0.1 percent (one in 1000) infection fatality rate is much higher than the effective infection fatality rate of all the diseases that cause deaths attributed to “pneumonia and influenza” and that also typically cause two “common colds” or “flus” in healthy adults each year.
Even accepting the CDC’s estimates of the prevalence of illness due to influenza viruses (THE VIRUS), less than ten percent of all “common colds,” if the coronavirus COVID-19 spreads more easily than the influenza viruses, it may be able to kill more people than the influenza viruses with the same infection fatality rate (e.g. one in 1000, 0.1 percent). We also need to know how the SARS-COV-2 coronavirus spreads and how quickly.
If everyone in the United States were infected with the COVID-19 coronavirus, a 0.1 percent infection fatality rate (one in 1,000) would probably mean somewhat less than 330,000 additional deaths on top of the roughly 188,000 deaths from “pneumonia and influenza” (or is it 55,000 from “influenza and pneumonia”). There would be some overlap between COVID-19 coronavirus deaths and deaths of susceptible, mostly elderly persons that would have happened anyway due to conventional non-COVID diseases including the influenza viruses.
Conclusion
There is a remarkable lack of key measurements in the current coronavirus COVID-19 pandemic. These include the actual mortality rate (aka infection fatality rate) broken down by age, sex, race, pre-existing medical conditions, ambient temperature, sunlight levels, pollution levels, and other risk factors. The false positive and false negative rates of the tests for the disease, both the tests for an active infection such as the RT-PCR tests and tests for past infection such as the antibody tests. The methods and rates of transmission for the disease. Aerosol transmission probably occurs at least at a low level and is virtually unstoppable.
The CDC and the National Security bioweapons defense programs should have been set up to quickly and efficiently collect these key data and parameters as soon as a possible outbreak or attack was detected, independent of warnings and information provided by a potential adversary such as China or from the World Health Organization (WHO).
The confusing language and numbers on pneumonia and influenza on the CDC web site and in various official reports and documents seem to be primarily for marketing the flu vaccines rather than enabling informed decisions by patients and doctors or supporting external scientific research into the influenza viruses or other diseases.
(C) 2020 by John F. McGowan, Ph.D.
About Me
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
This is a video showing the CDC Cold vs Flu web page on April 16, 2020 shortly after I published my “Uncounted COVID Deaths? The CDC’s Contradictory Pneumonia and Influenza Death Numbers” where I discussed the contradictory language and claims on the CDC’s Cold vs Flu web page. The video was recorded to support further my discussion in the Uncounted COVID article/presentation and because I think it likely the web page will change as the CDC fields hard questions about its Influenza and Pneumonia web pages, reports, and other documentation.
Astonishingly the CDC gives two radically different numbers of deaths from pneumonia and influenza: about 55,000 “influenza and pneumonia” deaths in the leading causes of death table in the “Final Deaths” report for 2017, the latest year available, and about 188,000 in data on weekly “pneumonia and influenza” deaths, over THREE TIMES the leading causes of death number.
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing gesture recognition for touch devices, video compression and speech recognition technologies. He has extensive experience developing software in C, C++, MATLAB, Python, Visual Basic and many other programming languages. He has been a Visiting Scholar at HP Labs developing computer vision algorithms and software for mobile devices. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech).
