One thing before we dive in, there are some formatting changes that could be better like font or presentation (PICO). Overall, I think this hits the major points. There are lot of concerns with time to randomization, mITT, severity of illness, and the amount of people excluded compared to included (can we apply this to our pts?). I do find it interesting the trends towards vasopressin especially in lower severity of illness. I would say subgroup result should always be used for hypothesis generation weather the trial is positive or not, so I agree with the authors recommendations but not because of their rational. If you were to do a trial similar to this, how would you design it in…
Thank you for taking the time to comment and for the thoughful feedback. In my opinion, the time to randomization did not matter, as I don't think starting the drug earlier would have made a big difference. From an external validity point of view, any pressor is the same as another to me, based on multiple systematic reviews / meta-analysis; network meta-analysis and RTCs that have failed to show a difference in mortality. I think it is ok to ask the question, but I think it is unlikely that any pressor will show superiority to another in mortality. They are all different mechanisms of achieving a same MAP goal. Based on the above, I would not design a trial to…
One thing before we dive in, there are some formatting changes that could be better like font or presentation (PICO). Overall, I think this hits the major points. There are lot of concerns with time to randomization, mITT, severity of illness, and the amount of people excluded compared to included (can we apply this to our pts?). I do find it interesting the trends towards vasopressin especially in lower severity of illness. I would say subgroup result should always be used for hypothesis generation weather the trial is positive or not, so I agree with the authors recommendations but not because of their rational. If you were to do a trial similar to this, how would you design it in…