Composite outcomes

You add multiple outcomes together… this is great because you can increase the event rate.. if you are just looking at death lots of people might now die but if you look at death and hospitalizations well then it is easier to get to your expected event rate because its much easier to be hospitalized. This make it so you can have smaller sample sizes and it wont be nearly as expensive.

The problem is all the events are given the same importance and we know they are not the same importance to patients

Death and hospitalizations NOT THE SAME

PE and death—not the same

3 questions to asked. 

Part of the endpoints of similar importance to patient's?

Due to the more or less significant endpoints occur with the same frequency?

Due to the endpoints share similar relative risk reductions?