What to make of the vaccine pledge

Investors have schedules: they want to make a return of at least x percent within y years. Science does not follow schedules. That's why vaccine manufacturers got into trouble when the U.S. government keeps seting deadlines for when a coronavirus vaccine will come to market. The anticipated arrival date of late October defies belief. It implies that the U.S. government is poised to approve a vaccine that has been observed for basically one month. (Anyone who joined the Moderna trial at the beginning is just getting the second dose right now.)

This week, we learned that the top pharmas signed a "pledge" to assure the public (link). "We believe this pledge will help ensure public confidence in the rigorous scientific and regulatory process by which COVID-19 vaccines are evaluated and may ultimately be approved."

But those are empty words as they have not committed to any specific action. For example, the information submitted to ClinicalTrials.gov by Moderna, predicted as one of the first to cross the finish line, is as sparse today as it was in late July when I first looked at it (link). The document continues to omit the most crucial pieces of information needed to judge its scientific rigor.

First, the submitted documentation does not explain how the scientists will establish whether someone is infected. It appears that they are relying on self-reported symptoms, rather than diagnostic testing, to determine infection. I say this based on what I read at Moderna's recruiting website (link), noting that this is a marketing website, not a scientific website. If true, that's not what one would describe as rigorous science.

Second, it does not state when the scientists plan on conducting analyses. If they analyze the data after one month, the best they can say is that the vaccine provides immunity for one month (and only for some proportion of people, as is the case with any medicine).

For this pledge to have any value, the pharma CEOs must commit to releasing more details of the clinical trials. What I listed above are not fancy requirements; they are rudimentary and fundamental to the practice of pre-registering clinical trials. It is shocking that those documents omit the information; it is insulting after signing a pledge to use rigorous science.

***

See this previous post to learn about the dangers of not pre-registering the analysis plan, and selecting some short observation period based on initial outcomes.

PS. The Oxford/AstraZeneca trial is catching attention as they have paused due to a severe side effect. I must say that the design document they submitted to ClinicalTrials.gov addresses the two questions listed above. They plan on doing an analysis after 6 months, and they will confirm infections using a PCR diagnostic test (although it sounds like they will count asymptomatic infections as not infections - the devil is always in the details).

Comments

[Ken]

All you need https://www.modernatx.com/cove-study

One point about this, is that it is not unexpected that they have a through protocol. It is a requirement for ethics approval for human research, and also the FDA require conduct of clinical trials to meet a certain standard.There would have been a fair amount of thought went into the choice of the primary outcome. They obviously feel that if someone has a respiratory illness with certain characteristics it is much more likely to be COVID-19 and that is a better option than using a PCR criteria. I have my doubts but they are more likely to be correct.