Injecting evidence into the vaccine spin

[Wandering Eyes]

Simon Thornley, Gerhard Sundborn.

12 April 2021

We are usually supporters of vaccines, and our children are all immunised.

Yet we have deep unease that the least tested vaccine in living memory, for a virus posing little risk to most people, has been purchased at great cost, and is being commended to New Zealanders as “safe and effective”.

When Covid Plan B opposed the elimination strategy a year ago, it was not because of doubts over vaccines. We said the strategy was too costly to pursue while waiting for a vaccine. We said that most solutions were likely to be unwarranted by the small danger posed by SARS-CoV-2.

The early development of vaccine-like products is a triumph of science.

Plan B had worried that vaccines would take years to pass the usual safety and effectiveness tests. We hadn’t counted on the panic being so strong that usual stages would be curtailed.

The speed, haste, and enthusiasm with which covid-19 vaccines have been thrust upon us has made us pause and examine the evidence.

New Zealand has purchased enough of the Pfizer mRNA vaccine for everyone in the country. Several other supply arrangements exist in case one falls through.

We do not know the cost of the vaccines, as it has been hidden from us. Sources overseas indicate the cost to governments of the Pfizer/BioNTech vaccine as US$19.50/dose – or NZ$27.63 at current exchange rates. So for 10 million doses, NZ may have paid NZ$276 million.

That doesn’t include the costs of distribution, cold-chain, quality control and administration of the jab.

Is this vaccine worth the cost? What is the evidence for the vaccine’s benefit and are there any potential downsides?

A primary factor in whether a vaccine is warranted for a particular disease in pre-covid times has rested on three principal factors. These are, among others:

• Benefits for high-risk individuals compared to the whole population.
• Healthcare system and societal costs incurred by vaccination programmes compared to treating disease.
• Life years and quality adjusted life years gained because of vaccines.

The World Health Organization has released criteria which focus on clinical consequences, such as safety and efficacy, but also encompass other areas such as public health benefits such as the reduction of infection rate.

The document mentions the ‘preferred’ need for at least 70% vaccine efficacy, with consistent results in the elderly. 50% efficacy is considered minimal. The endpoint may be a combination of disease, severe disease, and or shedding or transmission. The document says 6 months of protection may be acceptable, but one year is preferable.

The WHO discussion of vaccine safety is vague. It says a ‘highly favourable benefit to risk profile’ is ideal, but the benefits outweighing the safety risks is ‘acceptable’.

A footnote mentions that the ‘potential for enhanced disease’ should be considered. Clearly, authors of the document are cognisant of the potential for antibody dependent enhancement, a phenomenon present for other viruses such as dengue, Zika, Ebola and other coronaviruses. This occurs when high antibody concentrations are initially effective against a pathogen, but waning immunity and lower concentrations paradoxically enhance the severity of infection.

The true cost of the epidemic is hard to pin down, and we’re yet to see a formal cost-benefit analysis of the effects of the vaccine, since the long-term effects of the product are unknown.

The main choice facing New Zealanders now is whether to be vaccinated with the Pfizer one, so we’ll focus on this. The main evidence relating to this vaccine is outlined in the New England Journal of Medicine. The trial randomly allocated ~40,000 participants, roughly half each into vaccinated and placebo groups and followed them for two months to see whether they developed covid-19 infection. The study excluded people under the age of 16, pregnant women and people with a history of either covid-19 or any immunocompromising condition. The primary outcome was defined by a positive PCR test, with at least one symptom of infection.

The headline results were 169/20,172 covid-19 events in the placebo group, compared to just 9/19,965 in the vaccine group, giving an efficacy of 94.6% after two months. The most reported side effects from the vaccine were transient, such as fatigue, headache, and muscle pain. The two-month trial did not address viral transmission nor long-term safety and effect.

The trial has come under scrutiny for unexplained incongruities. Further documentation has revealed that the rate of ‘suspected, but not confirmed’ covid-19 were similar between the two groups: with 1,594 cases in the vaccinated and 1,816 in the placebo.

There was also an imbalance in exclusions due to unexplained ‘protocol deviations’. In the vaccinated group 311 were excluded, compared to only 60 in the placebo. The chances of being excluded from the trial were therefore (311/21,720)/ (60/21,728) = 5.2 times higher in the vaccinated group, a ratio which is extremely unlikely to be due to chance (P < 0.001). This finding is buried in papers only made available to US regulators, rather than highlighted in the trial results. The selective exclusion of individuals in trials is an area commonly exploited by drug manufacturers to exaggerate claims. The best evidence conventionally comes from including all randomised subjects (‘intention-to-treat’), whether they deviate from the trial protocol or not. Requests for scrutiny of the trial data, to understand what factors lead to these exclusions, have been ignored.

Since the primary outcome of the trial is related to mild covid-19 events, we know little about whether the vaccine prevents deaths from the virus. Others have indicated that trials will not be sufficiently powered to detect differences in need for hospital treatment, let alone death from covid-19, since the number of subjects and resources required for such a trial would be prohibitive. We cannot assume that prevention of infection translates to fewer deaths. For influenza vaccinations, for example, even though they reduce infection, that has not translated into lower mortality after widespread uptake.

The vaccine has been studied in a healthy population who are largely unaffected by covid-19 hospitalisation or death, even with subjects drawn from supposedly ‘hard-hit’ regions, including the US, Brazil, South Africa, Germany, and Turkey. This trial evidence supports reductions in mild covid-19 infections only. We simply don’t know whether the vaccine will prevent what really matters: hospital and intensive care admissions and deaths.

The trial confirms that for people of working age, the risk of fatality from covid-19 is extremely low. So low in fact, that there will be too few deaths to run a trial without spending an extraordinary amount of money on a very large one. The conclusion by health authorities not to run a trial on vaccine effectiveness to prevent death means they themselves conclude that the fatality risk of the virus, and hence need for a vaccine, is overblown.

New Zealanders are being offered a covid-19 shot with the inaccurate assurance that it is “safe and effective”. From the evidence reviewed here, this message is disingenuous.

The long-term benefits and harms from covid-19 vaccines are unknown since they have only been recently used in humans. This is acknowledged in Medsafe’s 58 conditions for the emergency use. They require early alerts to company reports about the product’s safety and possible benefits. If the government’s own officials are sceptical and demand transparency, we should as well.

View the full article