Fits a random effects model describing the dependence in the cumulative incidence curves for subjects within a cluster. Given the gamma distributed random effects it is assumed that the cumulative incidence curves are indpendent, and that the marginal cumulative incidence curves are on the form $$ P(T \leq t, cause=1 | x,z) = P_1(t,x,z) = 1- exp( -x^T A(t) exp(z^T \beta)) $$ We allow a regression structure for the random effects variances that may depend on cluster covariates.
random.cif(
cif,
data,
cause = NULL,
cif2 = NULL,
cause1 = 1,
cause2 = 1,
cens.code = NULL,
cens.model = "KM",
Nit = 40,
detail = 0,
clusters = NULL,
theta = NULL,
theta.des = NULL,
sym = 1,
step = 1,
same.cens = FALSE,
var.link = 0,
score.method = "nr",
entry = NULL,
trunkp = 1,
...
)a model object from the comp.risk function with the marginal cumulative incidence of cause2, i.e., the event that is conditioned on, and whose odds the comparision is made with respect to
a data.frame with the variables.
specifies the causes related to the death times, the value cens.code is the censoring value.
specificies model for cause2 if different from cause1.
cause of first coordinate.
cause of second coordinate.
specificies the code for the censoring if NULL then uses the one from the marginal cif model.
specified which model to use for the ICPW, KM is Kaplan-Meier alternatively it may be "cox"
number of iterations for Newton-Raphson algorithm.
if 0 no details are printed during iterations, if 1 details are given.
specifies the cluster structure.
specifies starting values for the cross-odds-ratio parameters of the model.
specifies a regression design for the cross-odds-ratio parameters.
1 for symmetry 0 otherwise
specifies the step size for the Newton-Raphson algorith.m
if true then censoring within clusters are assumed to be the same variable, default is independent censoring.
if var.link=1 then var is on log-scale.
default uses "nlminb" optimzer, alternatively, use the "nr" algorithm.
entry-age in case of delayed entry. Then two causes must be given.
gives probability of survival for delayed entry, and related to entry-ages given above.
extra arguments.
returns an object of type 'cor'. With the following arguments:
estimate of proportional odds parameters of model.
variance for gamma.
the derivative of the used score.
scores at final stage.
scores at final stage.
matrix of iid decomposition of parametric effects.
A Semiparametric Random Effects Model for Multivariate Competing Risks Data, Scheike, Zhang, Sun, Jensen (2010), Biometrika.
Cross odds ratio Modelling of dependence for Multivariate Competing Risks Data, Scheike and Sun (2012), work in progress.
## Reduce Ex.Timings
d <- simnordic.random(5000,delayed=TRUE,cordz=0.5,cormz=2,lam0=0.3,country=TRUE)
times <- seq(50,90,by=10)
add1 <- timereg::comp.risk(Event(time,cause)~-1+factor(country)+cluster(id),data=d,
times=times,cause=1,max.clust=NULL)
### making group indidcator
mm <- model.matrix(~-1+factor(zyg),d)
out1<-random.cif(add1,data=d,cause1=1,cause2=1,theta=1,same.cens=TRUE)
summary(out1)
#> Random effect variance for variation due to clusters
#>
#> Cause 1 and cause 1
#>
#>
#> Coef. SE z P-val Cross odds ratio SE
#> intercept 1.226821 0.1520889 8.066471 6.661338e-16 2.226821 0.1520889
out2<-random.cif(add1,data=d,cause1=1,cause2=1,theta=1,
theta.des=mm,same.cens=TRUE)
summary(out2)
#> Random effect variance for variation due to clusters
#>
#> Cause 1 and cause 1
#>
#>
#> Coef. SE z P-val Cross odds ratio
#> factor(zyg)MZ 2.2934222 0.3267745 7.018363 2.244871e-12 3.293422
#> factor(zyg)DZ 0.4308398 0.1401646 3.073812 2.113423e-03 1.430840
#> SE
#> factor(zyg)MZ 0.3267745
#> factor(zyg)DZ 0.1401646
#########################################
##### 2 different causes
#########################################
add2 <- timereg::comp.risk(Event(time,cause)~-1+factor(country)+cluster(id),data=d,
times=times,cause=2,max.clust=NULL)
out3 <- random.cif(add1,data=d,cause1=1,cause2=2,cif2=add2,sym=1,same.cens=TRUE)
summary(out3) ## negative dependence
#> Random effect variance for variation due to clusters
#>
#> Cause 1 and cause 2
#>
#>
#> Coef. SE z P-val Cross odds ratio
#> intercept -0.2235706 0.03021707 -7.398816 1.374456e-13 0.7764294
#> SE
#> intercept 0.03021707
out4 <- random.cif(add1,data=d,cause1=1,cause2=2,cif2=add2,theta.des=mm,sym=1,same.cens=TRUE)
summary(out4) ## negative dependence
#> Random effect variance for variation due to clusters
#>
#> Cause 1 and cause 2
#>
#>
#> Coef. SE z P-val Cross odds ratio
#> factor(zyg)MZ -0.46488910 0.02998657 -15.503242 0.0000000 0.5351109
#> factor(zyg)DZ 0.08958262 0.06140977 1.458768 0.1446289 1.0895826
#> SE
#> factor(zyg)MZ 0.02998657
#> factor(zyg)DZ 0.06140977