COVID-19: The Latest Nothing-Burger Anti-Vax Nonsense

by Raywat Deonandan, PhD
Epidemiologist & Associate Professor
University of Ottawa
(I add my credentials to these COVID-19 blog posts in case they get shared. I want readers to know that my opinion is supposedly an educated and informed one)

As I write this, I have a sick child at home to whom I am giving solo care, have pulled multiple all-nighters this week, and have an enormous document due tomorrow. And yet I’m taking the time to write this brief blog post because some journalists have tagged me in an ongoing discussion about the latest COVID-19 vaccine nothing-burger “scandal”. Please understand if my patience seems strained, as I am running on fumes today.

In short, some are claiming that a Pfizer executive “leaked” the supposed “bombshell” revelation that their COVID-19 vaccine was never tested for its effectiveness against transmission:

Forgive my profanity, but the number of times some stupid shit like this interrupts my day is just too damned high. Dear journalists: ten minutes of Googling or a phone call with anyone capable of reading scientific clinical trial results would clear this up.

Luckily, BBC News science reporter Rachel Schraer has already created a very good Twitter thread responding to the ridiculousness of these claims. I encourage you to read it thoroughly.

For my part, I’ve already explained on this blog, multiple times, how vaccine efficacy against transmission can be computed and why it is rarely done.  (You might want to revisit this post, this post, and this post.) But let me summarize nonetheless.

The hierarchy of things these vaccines can do, from easiest to hardest is:

(5) prevent death,
(4) prevent hospitalization,
(3) prevent serious disease,
(2) prevent symptomatic disease, and
(1) prevent actual infection.

Somewhere around (0), (1) and (2) is “prevent transmission.”  It’s possibly the hardest thing to do.  It’s definitely the hardest thing to measure.

The famous COVID-19 clinical trials of 2020 had as their main endpoint symptomatic disease.  That is, when a trial participant got COVID symptoms, they then got a PCR test. And if they tested positive for COVID, that that was logged. At the end of the trial, they summed up the number of people in the vaccine group who got logged and compared that to the number of people in the placebo group who got logged. And the difference was staggering.

Pfizer’s clinical trial reported results after 170 cases were detected using this approach. Amazingly, 95% of those cases were in the placebo group. That’s how they got that very impressive vaccine efficacy number: by comparing symptomatic cases between the vaccine and placebo groups. I explained all of this in a blog post from late 2020.

Let me repeat something important. People only got tested if they started showing symptoms of COVID. That’s why the endpoint is symptomatic disease and not infection per se. But we all know that asymptomatic infection is a thing. These trials could not detect asymptomatic infection because if you had no symptoms, you didn’t get tested!

Think about it. The trials enrolled tens of thousands of people over several months. To detect all infection, including asymptomatic infection, you’d have to give a PCR test to every participant every day of the trial. For 10,000 people that’s 300,000 tests per month. Back in 2020, PCR testing capacity was scarce. It would be have been impossibly expensive and intrusive to be doing that level of detection, and compliance would have likely been dismal, given the discomfort of PCR testing.

So the trials were unable to solidly detect asymptomatic infection.  What about transmission?

Well, what is transmission?  It’s when an infected person passes infection on to someone else.  Whether or not transmission happens depends on a number of factors: was the first person vaccinated? (That’s the key question here.)  Was the second person –the one who the first one infected- vaccinated?  How close did they associate?  (This is important because if you’re comparing vaccinated people against unvaccinated, you want to make sure they have equal opportunity to infect others.) That means those living in crowded homes are more likely to transmit than those living alone, for example.

In epidemiology, the best construct for measuring transmission is something called a “secondary attack rate” (SAR), which is the incidence of infection among the household contacts of people who become infected.

To measure SAR, you need to first identify everyone who became infected, including asymptomatic people.  (We’ve already established that that was impossible to do in these trials.) Second, you need to test all of the household contacts of those infected people…. and test them repeatedly over several days. Ideally, not just their household contacts, but everyone with whom they come into contact in the course of their lives.

You begin to see how impossible it is to perfectly compute vaccine efficacy against transmission.

All of this was transparent in the clinical trials documentation. To anyone who read the actual reports and knew how to read scientific papers, it was clear that there is no “bombshell revelation” here. This is just how these clinical trials were structured. And you know what? >90% efficacy against symptomatic disease is pretty damned awesome. If you reduce the chances of symptomatic disease, you also reduce the chances of transmission.

What am I saying?  I’m saying that even though the trials did not overtly measure transmission, there was a strong argument to make that their results nevertheless implied significant reduction in transmission. Indeed, that is what was observed in the community after the vaccines went global. Until the Delta variant reared its head, the vaccine was doing a remarkable job in curtailing transmission. Then Omicron came along and blew that to smithereens. But up until that point, the mRNA jabs did a splendid job in slowing transmission, despite the fact that the original trials were not set up to specifically measure that outcome. In mid 2021, we were so close to smothering this pandemic via the awesome power of these vaccines. Damn you, variants!

Okay, so how does one measure vaccine efficacy against transmission?  Well, after the vaccine is deployed to a large community, rigorous contact tracing of infected individuals can identify which close contacts were infected (i.e., compute secondary attack rates), then work backwards to see which index cases were vaccinated and which were not. That was done a few times. For example, this Dutch study computed an efficacy against transmission of 71% (pre-Delta) by comparing secondary attack rates.

And there are some more modelling-based hand-waving techniques one can employ, perhaps by measuring viral load, as this paper suggests.

Those obsessed with this non-story will point to an FDA press release from Dec 11, 2021, which included the nugget, “nor is there evidence that the vaccine prevents transmission of SARS-CoV-2 from person to person.” But it’s important to note that the beginning of the sentence is, “At this time, data are not available to make a determination…”

Those data came later from community studies, as I’ve described above.


So let me summarize:

(1) Pfizer was always transparent that they were not measuring transmission. The primary endpoint was symptomatic disease.

(2) This was obvious to any scientist who read the clinical trials. I myself pointed this out in multiple media interviews at the time. Non-experts were conflating “symptomatic disease” with “any infection” and “transmission”. These are different constructs requiring different tools to measure them.

(3) Even so, there was strong evidence right up until the arrival of Omicron that the vaccines were doing a good job of curtailing transmission.

(4) There is no scandal, leak, or bombshell revelation here.  Just a bunch of people who don’t know how to read science papers properly.


Sorry if I sound a little intemperate, impatient, and catty today. I’m off now to force my screaming toddler to take his medication and to down another huge cup of coffee. (Me, not him.)