| > This is not a strong signal. The adjusted odds ratio is 0.03 (95% CI: 0.003 - 0.25) Are you for real? Smallest effect in the 95% CI is a 4x reduction in ICU admission. Yes, it's probably closer to the bottom end of this range, but this is a fantastically different. > For randomised trials, it makes no sense to say that the difference in baseline is not significant For randomized trials, the baselines are not likely to be massively different. And, of course, the authors compared risk factors and ages to rule out some of the ways that the trial could be ridiculously tilted from the outset. > In this case it's pretty obvious the effect of vitamin D would easily be non significant if you shift around some variables. ...??? The endpoint was pre-declared, and we're just comparing two pre-randomized groups. Exactly what variables would you shift around? > Reeks of p hacking imo If you're going to cast aspersions, be a little more concrete. |