risksetROC (version 1.0.4.1)

SchoenSmooth: Incident/Dynamic (I/D) ROC curve, AUC and integrated AUC (iAUC) estimation of censored survival data

Description

This function smooths the Schoenfeld residuals using Epanechnikov's optimal kernel.

Usage

SchoenSmooth(fit, Stime, status, span=0.40, order=0, entry=NULL)

Arguments

fit

the result of fitting a Cox regression model, using the coxph function

Stime

Survival times in case of right censored data and exit time for left truncated data

status

Survival status

span

bandwidth parameter that controls the size of a local neighborhood

order

0 or 1, locally mean if 0 and local linear if 1

entry

entry time when left censored data is considered, default is NULL for only right censored data

Value

Returns a list of following items:

time

failure times

beta

estimate of time-varying parameter \(\beta(t)\) at each unique failure time.

Details

This function smooths the Schoenfeld residuals to get an estimate of time-varying effect of the marker using Epanechnikov's optimal kernel using either local mean or local linear smoother.

References

Heagerty, P.J., Zheng Y. (2005) Survival Model Predictive Accuracy and ROC curves Biometrics, 61, 92 -- 105

Examples

# NOT RUN {
data(pbc)
## considering only randomized patients
pbc1 <- pbc[1:312,]
## create new censoring variable combine 0,1 as 0, 2 as 1
survival.status <- ifelse( pbc1$status==2, 1, 0)
survival.time <- pbc1$fudays
pbc1$status1 <- survival.status
fit <- coxph( Surv(fudays,status1) ~ log(bili) +
                                     log(protime) +
                                     edema +
                                     albumin +
                                     age,
              data=pbc1 )
eta5 <- fit$linear.predictors
x <- eta5
nobs <- length(survival.time[survival.status==1])
span <- 1.5*(nobs^(-0.2))
fitCox5 <- coxph( Surv(survival.time,survival.status) ~ x )
bfnx1.1 <- SchoenSmooth( fit=fitCox5, Stime=survival.time, status=survival.status,
                       span=span, order=1)
bfnx1.0 <- SchoenSmooth( fit=fitCox5, Stime=survival.time, status=survival.status,
                       span=span, order=0)
plot(bfnx1.1$time, bfnx1.1$beta, type="l", xlab="Time", ylab="beta(t)")
lines(bfnx1.0$time, bfnx1.0$beta, lty=3)
# }