c*********4 发帖数: 3 | 1 My research is based on how the treatments influence the patient's hazard
rate. It is kind of similar to the transplant treatment fixed effect and
time-dependent effect in biostatistics. Does anyone who knows SAS well give
me some advice on this code?
my email: c*********[email protected]
Data description:
depend=0, as dependent variable, represent the patient is dead, =1
represents the patient is censored.
trt=1 represent one treatment, =2, two treatment.
treatment1 means the time (year)from the st | l****r 发帖数: 57 | 2 If you want to predict the patient's death influenced by the treatments.
Consider the following the idea:
If all your independent variables has two values and time-dependent, and
future state only depends on current state, you could construct a discrete-
time Markov chain. Don't forget to add one state as death state. For
estimating the transition matrix composed by transition probabilities
between two states especially from one risk state to death, you could use
logistic regression model.
give
【在 c*********4 的大作中提到】 : My research is based on how the treatments influence the patient's hazard : rate. It is kind of similar to the transplant treatment fixed effect and : time-dependent effect in biostatistics. Does anyone who knows SAS well give : me some advice on this code? : my email: c*********[email protected] : Data description: : depend=0, as dependent variable, represent the patient is dead, =1 : represents the patient is censored. : trt=1 represent one treatment, =2, two treatment. : treatment1 means the time (year)from the st
|
|