Fully sequential Bayesian clinical trial design for in-hospital treatment of COVID-19

[f2harrell]

Great questions Eric.

@ENB

First, is the outcome for each patient the worst outcome experienced by the patient or the outcome on day 15 after randomization?

I know of one study being planned that will measure the ordinal scale daily. For others I think it’s the status at day 14 regardless of the path to get there. I may be wrong.

@ENB

Second, the document states that the SOFA score is going to have a linear effect in the model. I may be mistaken but isn’t this a point-based score that could be broken up into a number of sub-scores or even the continuous underlying measurements? As a practical matter, is it required that all measurements be performed on the same day and how do multiple measurements come into play?

Sample sizes are not large enough to allow for adjustment of individual SOFA components, I have not discussed with the critical care docs how the component measurements are made but I assume they are measured on the day of randomization.

Regarding data entry the best solution is REDCap and I am talking to our PIs about sharing the REDCap metadata/forms online.