This is the same study that has been posted here repeatedly, probably without people realizing it is the same one. Again, the entire conclusion rests on how ~260 cases bin between cohorts that number ~65000 each. In both cohorts, doubled vaxxed and previously infected, the central observation is that second infections are rare. Given that the events of interest (ie secondary infection) are very small, there are statistical issues and cohort matching issues associated with the study. These problems are touched upon at some length by various people posting within the comments section of the preprint server on which the study is posted. As of last week, the paper still remains under review. Given the length of time that has passed, it is possible that that the authors have been asked to provide extensive additional documentation, or to demonstrate that their findings actually have predictive value in a larger, random sampling.
That Science chose to highlight this study-as a preprint-is I think unfortunate, and an example of how the scientific community and journal editors need to adopt more rigorous best practices to ensure studies like this are not misappropriated. The immunologist quoted in the Science highlight (the one who says "don't try this home") is a leading figure in the field and, at the time this preprint was posted, was about to have a big paper coming out showing how the the clean up on aisle 5 that results from the massive cell lysis associated with viral infection stimulates formation of cellular structures that can then become super-primed by subsequent vaccination. So they were sort of interested in this type of synergy between infection and vaccination and I think that is why Science decided to highlight it. I was bothered by the cavalier attitude at Science and wrote the editor with like "do you realize what you are messing with". The response was largely, well, if the numbers don't hold up then it doesn't matter. But that's wrong. For this study it is now too late for whether the numbers are right.
