Right? If the likelihood of someone with mild COVID needing hospitalisation or dying is less than 1% anyway, if n is around 100 then of course you’re not going to get any statistically significant results! This feels like it was designed to deliver this result.