Substituting treatment arm with drug exposure (continuous variable) in RCT

I think there is some small risk of a collider bias, which is why dose-response analysis (dose vs. clinical outcomes) is emphasized. But the analyses incorporating achieved blood concentration will be extremely helpful.

1 Like