Is anyone familiar with competing risk model in survival analysis. I have
gone through couple of publications. The concept was really blur and
conflicting.
The question: overdose mortality among injection drug users change with HIV.
The problem: competing risk between overdose mortality and AIDS death, as
well as other deaths in
IDUs.
I read the Paul Allison's survival book that has one chapter of competing
risk. BUt he was really talking about the cause-specific hazard, not
competing risk hazard. Also, recent paper from CASCADE study in Europe, the
outcome is AIDS defined illness and they calculated competing risk hazard by
censoring subjects who had uninterested illness at the end of follow-up,
instead of the date of that illness happened. But this strategy obviously
doesn't work for mortality.
Any help/thought. Appreciate!
