SmoothHazard (version 1.0.9)

lifexpect.idmPl: Predictions of life expectancies from an illness-death model using a penalized likelihood approach.

Description

Predict life expectancies from an object of class idmPl. Life expectancies are calculated at time s for a subject who has the covariates values Z01, Z02, Z12. Confidence intervals are calculated.

Usage

lifexpect.idmPl(object, s, Z01, Z02, Z12, nsim=1000, CI=TRUE, 
...)

Arguments

object

an idmPl class objects returned by a call to the idm function with hazard="Splines".

s

time at prediction.

Z01

vector for the values of the covariates on the transition 0 --> 1 (in the same order as the covariates within the call. The default values are all 0.

Z02

vector for the values of the covariates on the transition 0 --> 2 (in the same order as the covariates within the call. The default values are all 0.

Z12

vector for the values of the covariates on the transition 1 --> 2 (in the same order as the covariates within the call. The default values are all 0.

nsim

number of simulations for the confidence intervals calculations. The default is 1000.

CI

boolean: with (TRUE) or without (WRONG) confidence intervals for the life expectancies. The default is TRUE.

...

others arguments.

Value

• a list containing:
• life.in.0.expectancy
  life expectancy in state 0 and confidence interval.
• life.expectancy.nondis
  life expectancy of a non-diseased subject and confidence interval.
• life.expectancy.dis
  life expectancy of a diseased subject and confidence interval.

See Also

idm

Examples

data(Paq1000)
d <- Paq1000
names(d) <- c("dementia","mort","entry","L","R","time","certif","death")

fit <- idm(formula02=Hist(time,event=death,entry=entry)~certif,
formula01=Hist(time=list(L,R),event=dementia)~certif,data=d,hazard="Splines") 

pred <- lifexpect.idmPl(fit,s=70,t=80,Z01=c(1),Z02=c(1),Z12=c(1))