« Cybersecurity mystery 2 | Main | The triumph of the press release: more on the Merck Covid19 pill »


Feed You can follow this conversation by subscribing to the comment feed for this post.


61% is definitely not “the vast majority of Americans”. I think for a majority to be “vast” you need much more than 6 out of 10. This would not even be enough to pass a bill in the senate! :)


Cris: Since they excluded anyone with one or more doses, the current number is 72% according to OurWorldinData, and this number is monotonically increasing over time. I still think any of these studies should include vaccinated people, otherwise the result is not useful.


The problem with including vaccinated is that you need a larger sample size, as their risk is somewhere like 1/5 to 1/10. The problem is that I will bet that if Merck had found an advantage they would then have tried to convince the regulatory bodies that the effect was the same RR in vaccinated patients. That is one of the advantages of RR is that you can do things like that, although it is has assumptions that haven't been evaluated. and then you can turn them into absolute risks for the population.

I think this drug has been tried for other diseases so that would give some information on adverse events.

The interim is very dodgy. You would need to look at the protocol to find out what the chosen significance level was for the interim. It obviously didn't meet it, so the trial continued, but to then quote a p value for the interim is wrong without comparison to the chosen significance level. In other words the drug didn't work at the interim, however they had written the paper. There are reasons to use a one sided test so that the trial wont be stopped if the placebo is better than the drug. There is also sometimes a futility analysis, where if the drug is performing too poorly at the interim they give up.

I gather that Pfizer has a drug that actually works, so there isn't much point in continuing.


Ken: One-sided is not the proper way to run trials. I'd be shocked if that is pre-specified. We don't have the protocol nor do we have any scientific paper associated with this trial. Neither do we have anything on Pfizer other than press releases, if I'm not mistaken

The comments to this entry are closed.

Get new posts by email:
Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR, Wired.

See my Youtube and Flickr.


  • only in Big Data
Numbers Rule Your World:
Amazon - Barnes&Noble

Amazon - Barnes&Noble

Junk Charts Blog

Link to junkcharts

Graphics design by Amanda Lee

Next Events

Jan: 10 NYPL Data Science Careers Talk, New York, NY

Past Events

Aug: 15 NYPL Analytics Resume Review Workshop, New York, NY

Apr: 2 Data Visualization Seminar, Pasadena, CA

Mar: 30 ASA DataFest, New York, NY

See more here

Principal Analytics Prep

Link to Principal Analytics Prep