Randomised controlled trials have their limits. For example, you cannot randomise healthy people to an intervention or real placebo and keep them blinded and untreated for a long time to assess long-term effectiveness or harms profile.

 These days, this principle is routinely abused, with the assumption that everything tested is “safe and effective” and that participants cannot be deprived of “safe and effective” interventions for too long. However, the principle is, in theory, ethically correct.

But this is another story tied to active “placebos”, which we will develop in a subsequent post.

Share

In post 11c, we listed the definitions of the Cominarty cases according to Protocol C4591001 (PF-07302048) Statistical Analysis Plan (SAP), part of Clinical_Study_Report_Part_2 and Appendices at pdf page 529 of the same document.

As our readers know, C 4591001 was Comirnaty's Phases 1/2/3 registration trial. Because of the vagueness of the definitions and the use but incomplete reporting of PCR positivity, we concluded that the definitions and the outcomes reported in the main body of the clinical study report were black boxes.

In our Riddle series, we have explained why incomplete reporting of PCR results is incompatible with confirming the presence of replication-competent viral particles.

The definition reported in the clinical study report and those used by EMA in its EPAR have no cutoff for the likelihood of viral load; hence, infectivity is assumed, not proved. In addition, they do not exclude co-infection with another agent.

But let’s wave a magic TTE wand and assume all these “cases” are real “cases” of active SARS-CoV-2 infection; in each sample taken, there is no contamination, coinfection, and a very high viral load.

Next, let’s look at some of the figures EMA used in its SPC to explain why and how it licensed the product. Here’s one of the tables from page 115 reporting the primary efficacy analysis:

By the way, the TTE magic wand is on sale for a special offer of 40K GBP cash; if signed by Matt Hancock, the price is 50K GBP cash.

Do you notice anything strange (in the table, not the wand)?

Look at the placebo arm. The incidence is 162/18325 “at risk” placebo recipients.

That’s 884 cases per 100,000 recipients. If we assume they followed up people for 7 weeks post dose 1, the weekly incidence is 884/100k/7  = 126/100k or 0.13% (95% CI 0.10% to 0.15%)—a rate per 100K of 130 per week. And remember that this figure was derived by torturing the data with our magic wand and smoothing out the peaks and troughs of viral circulation.

The data from the table refers to one of the periods of maximum panic worldwide: April to October 2020. 

The thing about trials is that you, the researcher and investigator, have control over the design, data collection and interpretation of the results. This is why placebo arms of trials, real ones, not active placebos, are so important.

This trial, run between spring and late autumn 2020 in multiple sites, tells us about the nature of the pandemic……..where was it? It probably had taken a summer break. It’s hot work infecting people.

EMA never commented on any of this, and neither did the other regulators we can discern. Why?

This post was written by two puzzled old geezers.