I am trying to determine whether a procedure (o2_mon) is associated with an outcome (mortality). I have 2:1 matched data (2 controls for every 1 patient that got the procedu