Five examples of survival analysis problems
are briefly mentioned here. The first is a study
that follows leukemia patients in remission
over several weeks to see how long they stay
in remission. The second example follows a
disease-free cohort of individuals over several
years to see who develops heart disease. A third
example considers a 13-year follow-up of an
elderly population (60þ years) to see how long
subjects remain alive. A fourth example follows
newly released parolees for several weeks to
see whether they get rearrested. This type of
problem is called a recidivism study. The fifth
example traces how long patients survive after
receiving a he art transplant.
All of the above examples are survival analysis
problems because the outcome variable is time
until an event occurs. In the first example,
involving leukemia patients, the event of inter-
est (i.e., failure) is “going out of remission,”
and the outcome is “time in weeks until a
person goes out of remission.” In the second
example, the event is “developing heart dis-
ease,” and the outcome is “time in years until
a person develops heart disease.” In the third
example, the event is “death” and the outcome
is “time in years until death.” Example four,
a sociological rather than a medical study, con-
siders the event of recidi vism (i.e., getting rear-
rested), and the outcome is “time in weeks until
rearrest.” Finally, the fifth example considers
the event “death,” with the outcome being
“time until death (in months from receiving
a transplant).”
We will return to some of these examples later
in this presentation and in later presentations .
II. Censored Data
Most surviva l analyses must consider a key
analytical problem called censoring. In essence,
censoring occurs when we have some informa-
tion about individual survival time, but we don’t
know the survival time exactly.
EXAMPLE
1. Leukemia patients/time in remission
(weeks)
2. Disease-free cohort/time until heart
disease (years)
3. Elderly (60þ) population/time until
death (years)
4. Parolees (recidivism study)/time
until rearrest (weeks)
5. Heart transplants/time until death
(months)
Censoring: don’t know survival
time exactly
Presentation: II. Censored Data 5
我的内容管理
展开
