oceCens (version 0.1.2)

coxph2WR: Take coxph object and translate results to win ratios.

Description

Let cout a coxph object, then Using normal approximations and the output from the cout$coefficients and cout$var. If the cluster argument is used in the coxph call, then cout$var is the robust variance (see coxph.

Usage

coxph2WR(coutput, conf.level = 0.95)

Value

A vector or matrix with 4 elements (or columns) giving the win ratio, the lower and upper confidence limits, and the two-sided p-value.

Arguments

coutput

a coxph object created by coxph.

conf.level

confidence level.

Details

The function takes a beta coefficient and returns the win ratio version: exp(-beta). Confidence intervals are calculated by exp(-beta -/+ qnorm(1-(1-conf.level)/2)*sqrt(coutput$var)). P-values are two-sided.

References

Follmann, D., Fay, M. P., Hamasaki, T., and Evans, S. (2020). Analysis of ordered composite endpoints. Statistics in Medicine, 39(5), 602-616.

Examples

data(simScenario5)
xform<-oceFormat(data=simScenario5,oceTime=c("T1","T2","T3"),
   oceStatus=c("I1","I2","I3"),
   group="Z",outputDataFrame=TRUE)
# perform cox regression using time varying treatment efects, IZ1,IZ2, IZ3
# associated with the 3 prioritized endpoints
cout<- coxph(Surv(START, STOP, status) ~ IZ1+IZ2+IZ3, data=xform$data)
coxph2WR(cout)