all cause mortality is what is important. what if i told you that a drug puts you at a 2 fold risk of death from cancer compared to comparator, but a 4 fold risk of NOT dying from heart attack? you can read these things many different ways but all cause mort is usually much more significant. you'd have to look closer than an abstract to find out the details of the results (ie who smoked etc) as well, but who wants to do that....