Where Do They Find These Scary Statistics? Part II: Gross Estimation–Diphtheria Statistics Defy Reality
Top graph from page 423 of The Questionable Contribution of Medical Measures to the Decline of Mortality in the United States in the Twentieth Century.
Lower graph from page 208 of Trends in Diphtheria Mortality.
Series Links: Part I, Part III
From an article by Geoffrey Edsall–Immunization of Adults Against Diptheria and Tetanus–presented in 1951 and published in 1952:
…generally agreed the relative susceptibility of adults to diphtheria is related to the steady decrease in the incidence of the disease, a decrease which in this country has proceeded almost without interruption for the past eighty years, and which has occurred in states with no extensive immunization programs as well as in those with long established programs. (emphases added)
Eighty years would put the beginning of the decrease back to 1870.
In my previous entry in this series we saw an amazing statistic. According to various news stories (Healthfinder, Infectious Disease News, and Time) two medical journal papers claim that vaccination saves 33,000 lives a year. Time Magazine offers up this enthusiastic summary:
The authors cite a 2001 paper [actually published in 2005], estimating that every year seven of the 12 routine childhood vaccinations given in the U.S. prevent 14 million disease cases and 33,000 deaths. The JAMA authors believe their own death-rate reduction figures may in fact underestimate the true benefits of vaccination. The numbers don’t account for chronic disease averted because of a prevented infection: Hepatitis B, for example, is a major risk factor for liver cancer. (emphasis added)
However, the study they are describing also says this (under Main Outcome Measures on page 2155):
Estimates of the percent reductions from baseline to recent were made without adjustment for factors that could affect vaccine-preventable disease morbidity, mortality, or reporting.
Translation? “We paid no attention to changes in medical care, living standards, sanitation, or nutrition which clearly decreased the morbidity and mortality from infectious diseases. Neither did we analyse why states which didn’t vaccinate saw a reduction in diphtheria incidence and deaths similar to those which did.”
To find the original sources for the numbers mentioned in the news reports, we’re going to look at the diphtheria cases and deaths statistics from the two studies, side by side.
The two papers cited in the news articles are: Historical Comparisons of Morbidity and Mortality for Vaccine-Preventable Diseases in the United States, which was published in the Journal of the American Medical Association in 2007. Author Affiliations: National Center for Immunization and Respiratory Diseases (entity formerly known as the “National Immunization Program”); Centers for Disease Control and Prevention. The other paper is Economic Evaluation of the 7-Vaccine Routine Childhood Immunization Schedule in the United States, published in the Archives of Pediatrics & Adolescent Medicine in 2005. Author Affiliations: National Immunization Program, Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services.
Page 2156 in the 2007 study shows pre and post-vaccine stats for each disease. For diphtheria the prevaccine numbers are:
- average cases/year 1936-1945–21,053
- average deaths/year same period–1,822
- peak cases, 1938–30,508
- peak deaths, 1936–3,065
In the 2005 study, page 1140, Table 4 shows:
- projected cases and deaths from diphtheria if we stopped vaccinating: 247,214 cases, 24,721 deaths.
The difference between 1,822 deaths and 24,721 deaths is enormous! So how did the authors of the 2005 study come up with their numbers?
They started with a hypothetical birth cohort of 3,803,295 children.
For comparison, because the diphtheria cases are based on children 5 to 9 years old, I looked up the births in the U.S. from 1931 to 1940 and averaged the data, which gave a yearly birth cohort of 2,431,500 between 1936 and 1945.
This birth cohort is only one-third less than the hypothetical birth cohort of 3,803,295 from the 2005 article. A smaller birth cohort number could explain part of the discrepancy between the 2005 study’s diphtheria statistics, compared with the real diphtheria data in the 2007 study, but not so much that 90% of the diphtheria cases “projected” in the 2005 study could disappear.
Table 1 in the 2005 study on page 1137, gives a diphtheria rate of:
- 600 cases per 100,000. But they need:
- 6000 cases per 100,000 to reach their total of 247,214 cases from their hypothetical birth cohort.
According to the 2007 study’s formula, using the actual birth cohorts, the projected prevaccine average number of cases per year between 1936 and 1945, should have been 145,890, but the 2007 paper, using actual case numbers, states that there was a yearly average of only 21,053 cases. According to the CDC, by 1945, the rate was 15 cases per 100,000! It is important to note that this number is looking at the entire population of the U.S., not just cases in children. Taking the known yearly average of 21,053 cases against the known averaged birth cohort of 2,431,500, shows the rate in children was 866 cases per 100,000. Compare this to the ‘hypothetical’ rate of 6,000 cases per 100,000 postulated in the 2005 paper.
No matter how you slice it, the two medical studies, 2007 and 2005, which were quoted in the news stories as supporting each other, are in deep disagreement when it comes to diphtheria rates.
The 2005 study details 24,721 “projected diphtheria deaths” to provide roughly 2/3 of the “33,000 deaths” saved by vaccinating, while the 2007 study gives an average of only 1,822 actual deaths. Therefore, if the real diphtheria totals were unscientifically inflated by 90% to reach the predictions in the 2005 study, the oft quoted 33,000 statistical mountain has already shrunk to a much smaller statistical hill. And we still have six more diseases to evaluate!
The data presented in the 2007 study clearly cannot support 33,000 deaths from vaccine preventable diseases, therefore the people who wrote the news articles were given false information about the results of the published scientific research on the topic.
Of course they could have actually read the medical papers and looked at the numbers, but that is a lot to expect.
So, where did authors of the 2005 study find their statistics? The data sources are very vague:
The data used in this analysis were compiled from a variety of sources: the published literature, including surveillance data, study data, and expert consensus; several large, computerized data sets; and CDC unpublished data. When it was necessary to make estimates about the incidence of disease and complications from multiple publications, results from existing meta-analyses were used as the estimates if possible. Otherwise conservative assumptions were made, thereby intentionally underestimating the benefits of vaccination. (emphases added)
One source mentioned for the diphtheria statistics is the National Health Interview Survey. They also cite this study (note 5): Economic Evaluation of the use of Diphtheria, Tetanus, and Acellular Pertussis Vaccine or Diptheria, Tetanus, and Whole-cell Pertussis Vaccine in the United States, 1997, published in 2001. The statistics for diphtheria are really similar to the stats in the 2005 study. Their imaginary birth cohort is 4.1 million and they estimated 276,750 diphtheria cases resulting in 27,675 deaths. Table 1 on page 800 gives the case rate for diphtheria: 6,750 per 100,000, which is fairly close to the 6,000 in the 2005 article. The note for this number explains:
Estimated disease burden without vaccination is based on the average number of cases reported during the prevaccine era divided by the cohort of 4.1 million children and multiplied by 100,000.
Obviously, their prevaccine era doesn’t correspond with the prevaccine era statistics given in the 2007 paper. But when was the prevaccine era? The authors don’t tell us. In their Materials and Methods section, under Estimating Disease Burden Without Vaccination the authors do say this:
When no published data were available, the conclusions of an expert panel conducted by Batelle Inc on February 22, 1993, were used (Table 1). The panel members were Kathryn Edwards, MD, Scott Halperin, MD, Eric Hewlett, MD, and Edward A. Mortimer, Jr., MD. Based on review of of diphtheria incidence rates in the prevaccine era and after the introduction of diphtheria toxoid, the expert panel estimated that in the absence of vaccination, the age-specific incidence of diphtheria would be similar to that observed during the prevaccine period. (emphases added)
On what basis did this “expert” panel ignore the major society changes and public health changes which were responsible for a precipitous decline, or the antibiotic development which occurred after the vaccine was first used? On page 802 is a statement admitting that they ignored these key points:
It is possible that our model over estimates some of the economic benefits obtained when vaccinating against diphtheria, tetanus, and pertussis. When calculating benefits, we used the consensus from an expert panel that the disease burden without vaccination would be equal to that of the prevaccination era before the introduction of mass immunizations. In developed countries, however, the incidence of many infectious diseases, including diphtheria, tetanus, and pertussis, were declining before the introduction of universal immunization. (Emphases added)
Still no dates, no numbers of cases, no time trends. This is essentially the unverifiable opinion of a panel of vaccinologists, some of whom have been paid by the vaccine manufacturers.
To give an example of the “quality” of the notes, the source for the diphtheria statistics (note #17) is an obscure, hard to obtain book from 1941: Frost WH. Papers of Wade Hampton Frost, MD: A Contribution of Epidemiological Method. Macy KF, ed. New York, NY: Commonwealth Fund; 1941. Why is this considered a useful source? Why aren’t all the prevaccine years, cases and deaths in graph form in the experts’ analyses in the first place? Especially as some of the information is even available (in two contradictory versions, no less) on the CDC web-site!
For example; here is an excerpt from the CDC’s info for the general public:
In the 1920’s, diphtheria was a major cause of illness and death for children in the U.S. In 1921, a total of 206,000 cases and 15,520 deaths were reported. With vaccine development in 1923, new cases of diphtheria began to fall in the U.S., until in 2001 only two cases were reported. (emphases added)
It is a hallmark of vaccine literature, that the development of any vaccine results in a significant decline in new cases of the disease, even before the vaccine is widely used. Diphtheria illustrates this phenomena quite vividly. It is also a signature mark of the CDC’s presentation of data, that instead of covering an extended period of 10 or 20 or 30 years, or even all the available data, the CDC cherry picks the cases and deaths of a single year to display in uninformative isolation. To leave out all the rest is an unscientific omission, as can be seen by the two graphs, each illustrating declines occurring over six decade periods, shown at the beginning of this blog entry.
CDC’s Pink Book states (page 6 in the linked chapter) :
In the 1920s in the United States, 100,000–200,000 cases of diphtheria (140–150 cases per 100,000 population) and 13,000–15,000 deaths were reported each year. In 1921, a total of 206,000 cases and 15,520 deaths were reported. The number of cases gradually declined to about 19,000 cases in 1945 (15 per 100,000 population). A more rapid decrease began with the widespread use of toxoid in the late 1940s. (emphasis added)
The CDC ignores the massive statistical decline from 1870 to 1920 shown in the graphs at the top, and categorizes the decrease from 200,000 cases to 19,000 cases in 25 years as “gradual”. The CDC then mentions that the widespread use of the toxoid in the late 1940s resulted in a “more rapid” decrease. More rapid than what?
The news stories linked above were based on a press release, probably from the CDC. A thoughtful reader will be left with these questions: ““What was the purpose of publishing false inflated statistics?”, and, “How often are we told to be grateful for lives saved on the basis of unscientific exaggerations by orders of magnitude?”
An obvious area for further statistical fudging is the projected death rate of 10%. We were lucky enough to track down a medical paper from 1917 which is chock full of statistics showing declining death rates from diphtheria due to treatment with antitoxin. On page 451 in Table 4, the author lists statistics for children with diphtheria treated with antitoxin from 1895 to 1915 in Newark, NJ. The death rate had dropped below 10% by 1899 and stayed below 10% from then on. The lowest rate, reached in 1915, was 4.4%. However, according to the 2005 article, unvaccinated children in the 21st century would be dying from diphtheria at more than twice the rate seen in 1915! Has medical care really deteriorated that badly in the United States?
Would diphtheria have gone the way of scarlet fever, gradually becoming a mild, easily treated illness? Of the original 33,000 deaths, we can safely eliminate at least 24,721 dying from diphtheria. If you assume that the prevaccine living conditions of 1920 apply to children in 2001, and that medical practice has gone backwards to 1895, some sort of “stuck in a time warp” figure can be mocked up. Those are huge assumptions.
Our analysis will continue to go down the list of diseases used to create the estimated 33,000 deaths. The next entry will look at hepatitis B, where the vaccine is hypothetically preventing 232,001 cases and 3,427 deaths. Stay tuned!