Reading the abstract to the Lancet article [0] would probably convince any decently trained science reporter that all is good. They tick all the boxes for what you are supposed to do when you do a case-control study. It reads as very professional and well done. I certainly thought that it passed the bullshit test on first read. The problem is with the data itself, and it isn't obvious it is fabricated because it is such a ridiculous over the top lie that you wouldn't think someone would try to make such a lie.

0. https://www.thelancet.com/journals/lancet/article/PIIS0140-6...