Another Look at Covid Vaccine Studies

From the beginning of the COVID-19 crisis, the risk factors for severe forms and death from – or “with” – the respiratory virus called SARS-CoV-2 had been clearly identified: Advanced age, obesity, severe chronic comorbidities (other diseases, e.g. hypertension, diabetes, cancer).

For people without any of these characteristics, the risk of dying from (or even only “in relation with”) COVID-19 is very low and close to 0.

The vaccines are supposed to prevent severe disease and death; otherwise, they – and a fortiori their expedited approvals – would be entirely pointless. 

At this point in time however, we still cannot possibly know whether they actually do. Martin Kulldorf is therefore entirely right when he demands in his recent article that the manufacturers “conduct a proper randomized clinical trial that proves that the vaccines reduce mortality.”

The design and execution of such a trial – in the high risk group (e.g. >65 years of age, plus at least one comorbidity), over a reasonable timeframe (at least 6 months), comparing overall (not only test-positive) mortality in a placebo to a verum-group – would have been (and would still be) straightforward and much less complex than the registration studies that were in fact carried out with these products. 

How the trials were conducted is clearly stated in the protocols, publications, and FDA submissions: People who developed symptoms (the lists of these symptoms changed a little from one manufacturer to another, but they were all non-specific common cold or flu symptoms) underwent PCR testing. If – and only if – the test turned out to be positive (in the Pfizer study, this was the case in merely 170 out of more than 3,400 symptomatic patients), the endpoint of “symptomatic Covid-19” was considered as having been reached.

What these studies showed was that in people presenting with common cold or flu symptoms, the SARS-CoV-2 virus was detected significantly less frequently in the vaccinated than in the placebo group. 

What was thus demonstrated was in no way a reduction in any clinically defined and distinguishable disease entity, but only in the number of positive tests for one particular virus of many which are known to cause the non-specific symptoms in question.

What was not demonstrated however, was a reduction in common cold and flu symptoms per se. Quite the contrary. 

All the observational studies which have been carried out with the Covid-19 vaccines suffer, apart from some of the well-known general biases, from exactly the same fundamental flaw:  They show a decrease in “Covid-19-related” symptom-free or symptomatic cases, hospitalizations or deaths, but they do not ask the question whether this decline in test-positive patients translates into an overall reduction of flu cases, of (atypical) pneumonias, of hospitalizations and deaths. 

However, this is the clinically truly relevant question.

It is impossible to draw any firm conclusions from the data on the vaccines’ effect on general mortality which have been published up to now. The recent Danish analysis, apparently submitted to the LANCET, is again entirely correct when it argues “for performing RCTs of mRNA and adeno-vectored vaccines … comparing long-term effects on overall mortality.” 

These RCTs (Randomized Clinical Trials) absolutely need  also to, and above all, include a Placebo group, and not just compare the vaccines with each other though. 

The apparent superiority of the DNA-vector vaccines, as reported by the Danish group, is based on very small numbers with little inherent reliability. Moreover, one needs to be extremely careful with post-hoc statistical analyses on clinical endpoints which had not been pre-defined for the trial(s) in question – this can very quickly become akin to “data dredging.” 

Overall mortality has not been an endpoint in any of the Covid vaccine trials or studies so far. Conceptually, as the Covid mortality is part of the unavoidable mortality of the general population (we are not immortal, and on average we die at our average age of death), it may be impossible to demonstrate a general mortality benefit for the Covid vaccines – even more so as they do have potentially severe side effects. 

But properly conducted clinical trials with relevant (“hard”) clinical endpoints are the only way to find out and conclude.