Thread by @EricTopol

For evaluating the vaccine trials (and any trials), it's important to look at the number of endpoints and their severity.
The J&J trial has the same number of endpoints as @Pfizer/ @BioNTech_Group, @moderna_tx and @Novavax (1st interim) combined /1

The total number of participants was nearly the same for J&J and Pfizer trials, but the number of events was 2.75X higher for the former. While the J&J trial had 25% of participants in South Africa (w/ 95% B.1.351 variant) that doesn't account for this difference /2

The vaccine elimination of deaths and hospitalizations in the J&J trial (including South Africa) is noteworthy since there were so few of these events in the other 3 trials. Most (~90%) of the events in all the trials are mild to moderate infections /3

The gestalt from these 4 trials suggests that the mRNA and protein vaccines do well at blocking infections whereas, even though J&J's Ad vaccine is not as effective overall, it is highly efficient for preventing the ones that lead to hospitalizations or death /4

There are implications in these differences for asymptomatic transmission (carriers), #LongCovid (more frequent with mild/moderate infections), what it will take to get to population-level immunity. Nonetheless, we've got 4 very effective vaccines, which is terrific /5