mets (version 1.3.3)

phreg_lt: Lu-Tsiatis More Efficient Log-Rank for Randomized studies with baseline covariates

Description

Efficient implementation of the Lu-Tsiatis improvement using baseline covariates. Results almost equivalent with the speffSurv function of the speff2trial function. A dynamic censoring augmentation regression is also computed to gain even more from the censoring augmentation.

Usage

phreg_lt(
  formula,
  data,
  augmentR = NULL,
  treat.model = ~+1,
  augmentC = NULL,
  km = TRUE,
  cens.code = 0,
  level = 0.95,
  cens.model = NULL,
  typeII = NULL,
  ...
)

Arguments

formula

formula with 'Surv' outcome (see coxph) and treatment (randomization 0/1)

data

data frame

augmentR

formula for the randomization augmentation (~age+sex)

treat.model

propensity score model, default is ~+1, assuming RCT study

augmentC

formula for the censoring augmentation (~age+sex)

km

use Kaplan-Meier for the censoring weights (stratified on treatment)

cens.code

censoring code

level

of confidence intervals

cens.model,

default is censoring model ~strata(treatment) but any model can be used to make censoring martingales

typeII

if 1 then computes also alternative formulae that are based on the censoring martingale rather than the robust processes of Lu-Tsiatis computations.

...

Additional arguments to phreg function

Author

Thomas Scheike

References

Lu, Tsiatis (2008), Improving the efficiency of the log-rank test using auxiliary covariates, Biometrika, 679--694

Examples

## Lu, Tsiatis simulation
data <- mets:::simLT(0.7,100)

out <- phreg_lt(Surv(time,status)~Z,data=data,augmentR=~X,augmentC=~factor(Z):X)
out$coefs