"For the researcher, they show that studies using in-hospital mortality will be heavily biased by factors influencing whether death occurs in-hospital or shortly after discharge." . . . .