They changed the study target effect (which hour range), design (interventional vs observational) and inclusion/ exclusion criteria multiple times.

I don't really care at that point what their conclusion says, because I have no idea how to interpret the statistics in a theoretically sound way now.