Suspicious timing of Pfizer vaccine announcement

On November 9, Pfizer announced that its COVID-19 vaccine candidate demonstrated efficacy of over 90% based on the mix of 94 cases between the placebo and vaccine groups in its Phase 3 clinical trial.  This was extremely exciting news for the country less than a week after the presidential election.  Some of us had wondered about the timing of the announcement, given that the results far exceeded the primary efficacy endpoint of only 50% (with appropriate statistical confidence intervals) set by the FDA for a successful vaccine, and Dr. Fauci's concurrence that 50% would be acceptable. 

That same day, Pfizer CEO Albert Bourla denied that politics had anything to do with the timing of the release:

In an interview with CNN Chief Medical Correspondent Sanjay Gupta, Bourla said he first learned about the results of the trial Sunday and he has yet to see the data, throwing cold water on the idea that there was any political motivation behind releasing the news after voters in the United States chose their candidate for president.

Pfizer took several interim looks at the data during its vaccine trial, but it did not have enough infections in its test group to gain any meaningful data — until Sunday, Gupta reported.

As it turns out, the Pfizer announcement itself discloses that earlier interim analyses were planned, but decisions were made by Pfizer and the FDA to drop those analyses and reporting of results:

After discussion with the FDA, the companies recently elected to drop the 32-case interim analysis and conduct the first interim analysis at a minimum of 62 cases. Upon the conclusion of those discussions, the evaluable case count reached 94 and the DMC  [independent Data Monitoring Committee] performed its first analysis on all cases.

Nine days later, on November 18, Pfizer issued another announcement to report the conclusion of the Phase 3 study.  It reported that the final data indicated a vaccine efficacy of 95%, based on 170 confirmed cases of which 162 were observed in the placebo group and only 8 in the vaccine group.

So let's look at what the worst-case efficacy would have been if results had been reported earlier — at the revised 62-case analysis point agreed to by Pfizer and the FDA.  (This minimum efficacy calculation and the one for the 32-case point below assume that there were equal numbers of participants in the placebo and the vaccine groups as specified in the study protocol.)  Even if we assume that every one of the 8 final vaccine group cases had been observed at the 62-case point in the trial, there would have been 54 cases in the placebo group and 8 in the vaccine group, for an efficacy of 85%. 

But what would the efficacy have been had Pfizer and the FDA not "recently" agreed to scrap the 32-case interim analysis specified in the study protocol?  Again, assuming the very worst case, that every one of the final number of 8 cases in the vaccine group had been observed at the 32-case point, we would have had 24 cases in the placebo group and 8 in the vaccine group, which equates to a 67% efficacy, still easily exceeding the 50% efficacy success threshold that had been established by the FDA and Dr. Fauci.

I do not know the date that the 32-case point was reached, but I would be willing to wager that it was before the election, perhaps several weeks before.  Given that there were never any serious safety concerns reported throughout the trial, what could the reason be for burying great interim news based on analysis points specified in the study protocol, until after the election?

Photo credit: CoolcaesarCC BY-SA 4.0.