by John Jones, JD, PhD, Vaxxter Contributor
Part I: TB vaccine study in the New England Journal of Medicine
The New England Journal of Medicine publishes its weekly online version for free. Frequently they publish codswallop touting the safety and efficacy of some new vaccine. On December 19, 2019, they detailed the results of a trial with a new tuberculosis vaccine.
A team of 27 scientists, MDs and PhDs, lead by Dereck Tait, experimented on adults (ages 18-50) from three countries: South Africa, Zambia, and Kenya. They injected them with what they call M72/AS01E and … surprise, they claim that this product from GlaxoSmithKline (GSK) works better than a placebo to reduce the incidence of “first case definition” tuberculosis.
For the Editors at the NEJM, such a declaration was good enough but … I actually read the study.
When I read the full report, the cross-references and clarifying definitions, then looked at baseline infection rates, and considered the general health of the populations whence the samples were taken, their claims are falsified. Funny how those who are not paid by GSK (or Aeras, a non-profit funded in part by the Bill and Melinda Gates Foundation) can determine that this latest report about a shiny new vaccine is just more of the same: another experiment where victims were injected with a toxic stew that did nothing to promote health or protect them from infection.
About TB in Southern Africa
According to the article, the drug trial lasted nearly three years, using a population of “HIV negative [sic] adults, all with [Mycobacterium] tuberculosis infection [sic] … who had no signs or symptoms of tuberculosis disease and who had a sputum specimen that was negative for M. tuberculosis” (Tait et al. 2019).
Putting aside the problematic claims of HIV, we see that Tait et al. (2019) tried to select nominally healthy adults. But what of the claims about TB infection in the test subjects? According to the authors, the subjects were infected but had no signs or symptoms of disease.
By comparison, when TB rates were higher in the United States (in the era of routine smallpox vaccination), TB infections were usually found in people who lacked indoor plumbing. The standard practice was to confirm TB via microscopic analysis of sputum (see State v. Snow, 324 S.W.2d 532 (Ark. 1959)). Without the infection being confirmed by the sputum analysis, how can seemingly healthy people have a TB infection?
It’s the Vaccine, Stupid!
Nearly every government in the world, recommends and or administers BCG vaccine to newborns or 1 to 2-month-old infants. According to BCG World Atlas, Kenya and South Africa started injecting babies in the early 1970s. Depending on the year, BCG administration rates for infants in Zambia fall between 95% and 100%. What effect might such a practice have on public health and TB rates, especially when people are malnourished, have poor hygiene, and have limited access to potable water?
I am not surprised to read of high TB rates in countries where people lack toilets and soap. According to UNICEF, in Zambia, only 31% of the population uses basic sanitation services (49% urban areas, 19% rural areas); only 14% of the population has access to a hand-washing facility with soap and water (26% urban, 5% rural). As for Zambian schools: 21% do not have potable water, 34% do not have flush toilets, and 46% lack hygiene services. In such an environment, zymotic diseases will run rampant.
BCG vaccines are live vaccines derived from Mycobacterium bovis, a slow-growing aerobic bacterium and the causative agent of tuberculosis in cattle. It is related to Mycobacterium tuberculosis, the bacterium associated with tuberculosis in humans. M. bovis can jump the species barrier and cause tuberculosis-like infection in humans.
The organism used in the BCG vaccine was cultivated and developed over a period of 13 years (from 1908 to 1921) by two researchers, Calmette and Guerin, in Lille, France (Grange et al. 1983). Not only does the mycobacteria stay active once injected, in a small percentage of people, it becomes actively virulent (Grange et al. 1983: 133-134). So the most likely reason that anyone in Africa, or other parts of the world, would test positive for TB is that they are injected with the organism that can cause it!
Did the M72/AS01E vaccine ‘work’?
[Note: Tait et al. (2019) started with a total of 6,903 subjects, but ultimately only used 3,289 subjects in the final analysis].
Tait et al. (2019) report their results thusly:
(a) 3,573 received at least one dose of M72/AS01E or placebo [sic], and 3,330 received both planned doses. … 13 of the 1,626 participants in the M72/AS01E group, as compared with 26 of the 1,663 participants in the placebo [sic] group, had cases of tuberculosis that met the first-case definition (incidence, 0.3 vs. 0.6 cases per 100 person-years).
(b) The vaccine efficacy at month 36 was 49.7% (90% confidence interval [CI], 12.1 to 71.2; 95% CI, 2.1 to 74.2).
(c) Among participants in the M72/AS01E group, the concentrations of M72-specific antibodies … increased after the first dose and were sustained throughout the follow-up period.
(d) Serious adverse events, potential immune-mediated diseases, and deaths occurred with similar frequencies in the two groups.
Let me unpack these.
Note, Tait et al. (2019) admit that some people, in what they call the placebo group, received one dose of the independent variable (M72/AS01E). So this study has no true control group. The lack of any control, against which to compare the treatment, is made evident in the last comment about serious adverse events (SAEs).
Tait et al. (2019) found that SAEs occurred with similar frequencies in the two groups? But placebos are not suppose to create any serious adverse events. Apparently the editors at the NEJM saw nothing strange about the use of harmful and potentially deadly placebos. And Tait et al. (2019) were supposedly reporting on “safety”! This is what passes for “state of the art” vaccine science.
And what of efficacy? In his review of cities and nations in Europe, from the early 1700s to the 1880s, White (1885) noted that the standard used to judge the efficacy of smallpox vaccination was the general death rate.
But vaccination did not correlate with lower death rates in the 19th century (cf. White). Little wonder then that today, the primary standard used to determine if a vaccine is effective [sic] is the level of antibodies in the blood called a titer.
According to Claire-Anne Siegrist (2018), vaccination theory holds that the primary mechanism by which a vaccine mediates protection is in stimulating the production of antibodies. (See Chapter 2 in Plotkin’s Vaccines). Hence, at a minimum, in any paper published in the NEJM, should report the details of an increased titer level. And more importantly, Tait et al. (2019) only reported that the concentrations of M72-specific antibodies increased and remained elevated throughout the 3-year follow-up period. The authors did not explain the amount of increase and had no definition for the level of protection that the antibody levels conferred to their test subjects.
In section 5 of their synopsis of this very trial, the WHO writes:
“The exact mechanism of action of M72/AS01E is not known. Previous studies have showed that this vaccine induces … activation of interferon-gamma producing CD4+ T cells, and the production of antibodies. WHO strongly urges basic research to [understand] how this vaccine works … to inform general understanding on the mechanisms of protection against tuberculosis.”
Well, if we cannot know how it works, or if titers per se improve health or protect from disease, perhaps we should focus on the simple question asked by critics of vaccination in the 19th century: “What are illness and death rates?” For some bizarre reason – unexplained by the authors – the death rates were the same in the double-vaccinated and the placebo (“maybe once vaccinated”) groups.
Now we come to the TB disease rates reported. Looking at their two groups, Tait et al. (2019) report a difference in the TB disease rate (i.e., those cases meeting the standardized criteria for TB disease). At first glance, the disparity of 0.3 vs 0.6 per 100 person-years looks impressive. It might even be statistically significant, but …
Comparing Two Proportions: statistics 101
The concepts of Z-scores and t-tests are rudimentary in statistics. In my undergraduate courses, I regularly taught students how to compare groups. Chapter 9 of Healy’s Statistics (2009) shows us how to calculate and interpret a z-score to compare two proportions. Taking the data from Table 2 of the Tait et al. study, I followed Healy’s steps. The results are not surprising.
Measuring the numbers of TB cases in person-years, Tait et al. report that for their two-dose group, there were 13 cases of TB for 4427.61 person-years (which equals a rate of 0.002936 per 100 person-years). For the so-called placebo group, 26 versus 4463.07, the rate is 0.005826.
Superficially these rates seem distinct. However, the only way to determine if the difference is not due to random chance or another unaccounted for variable is to calculate the z-score for the difference in these proportions. Tait et al. did not produce such figures. But I did (and so can you).
The formula for a z-score of a difference of proportion is this:
z = (P1-P2) / standard deviation of pooled proportion
Since we know P1 and P2 (the proportions of cases of TB in each test group), in order to construct a z-score for the difference between two proportions, we need the following information: (a) the pooled proportion; and the (b) standard deviation of that pooled proportion.
The pooled proportion is simply the total number of TB cases divided by the total number of person-years: 39 / 8890.68, which is 0.0044. With this number, we apply the formula to calculate the standard deviation. (See Healy, Chapter 9). In this instance, the standard deviation of the pooled proportion is 0.022.
Putting the numbers into the formula we get:
z = [(13/4427.62 – (26/4462.07)] / 0.022 = -0.1287
NOTE: A z-score which ranges between -1.96 and +1.96 is NOT statistically significant.
With a z-score so close to zero, no statistical certainty can be ascribed to this vaccine trial.
Every medical doctor and PhD who worked on this study knows how to calculate a z-score and evaluate if these proportions were statistically distinct. Their paymasters at GSK know the importance of proving that a drug trial generates a statistically significant result. Without finding a statistically significant difference in the disease rates of those who got the shots vs the placebo group, they all know that M72/AS01E is worthless.
Share this article with your friends. Help us grow.
John C. Jones received his law degree (2001) and his Ph.D. is in political science (2003) from the University of Iowa. He has over 15 years of research and writing (both academic and journalistic) in fields of public policy and law, criminal and Constitutional law, and philosophy of science and medicine. His additional areas of expertise and specialized knowledge include applied statistics, etymology, political communications/public relations, litigation and court procedure. He has a particular interest in the science and history of vaccines.
Photo Credit: ID 26651723 © Crazy80frog | Dreamstime.com