[ptest1]

The problem I see with this trial is that it’s open-label, small n, and has subjective endpoints (ICU admission as the primary outcome). The study is underpowered to detect mortality. Given the open-label nature of the trial, the subjective outcome with the small n makes this result less strong.

Why is this the case? Well, for this trial the physicians treating patients knew who got the Vitamin D and who didn’t, and thus they may have been more likely to admit those who didn’t to the ICU (subjective). Something like mortality is not as subjective, but there are too few study participants to detect a mortality signal in this trial.

If I had been conducting this small open label trial I would have picked some less subjective outcomes, like maybe P/F Ratio.