WRrec: Generalized win ratio tests

Description

Perform stratified two-sample test of possibly recurrent nonfatal event and death using the recommended last-event assisted win ratio (LWR), and/or naive win ratio (NWR) and first-event assisted win ratio (FWR) (Mao et al., 2022). The LWR and FWR reduce to the standard win ratio of Pocock et al. (2012).

Usage

WRrec(ID, time, status, trt, strata = NULL, naive = FALSE)

Arguments

A vector of unique patient identifiers.

time

A numeric vector of event times.

status

A vector of event type variable; 2 = recurrent event, 1 = death, and 0 = censoring.

trt

A vector of binary treatment indicators.

strata

A vector of categorical variable for strata; Default is NULL, which leads to unstratified analysis.

naive

If TRUE, results for NWR and FWR will be provided in addition to LWR; Default is FALSE, which gives LWR only.

Value

An object of class WRrec, which contains the following elements.

theta

A bivariate vector of win/loss fractions by LWR.

log.WR, se

Log-win ratio estimate and its standard error by LWR.

pval

\(p\)-value by the LWR test.

theta.naive

A bivariate vector of win/loss fractions by NWR.

log.WR.naive, se.naive

Log-win ratio estimate and its standard error by NWR.

theta.FI

A bivariate vector of win/loss fractions by FWR.

log.WR.FI, se.FI

Log-win ratio estimate and its standard error by FWR.

...

References

Mao, L., Kim, K. and Li, Y. (2022). On recurrent-event win ratio. Statistical Methods in Medical Research, under review.

Pocock, S., Ariti, C., Collier, T., and Wang, D. (2012). The win ratio: a new approach to the analysis of composite endpoints in clinical trials based on clinical priorities. European Heart Journal, 33, 176--182.

Examples

Run this code

# NOT RUN {
## load the HF-ACTION trial data
library(WR)
head(hfaction_cpx9)
dat<-hfaction_cpx9
## Comparing exercise training to usual care by LWR, FWR, and NWR
obj<-WRrec(ID=dat$patid,time=dat$time,status=dat$status,
          trt=dat$trt_ab,strata=dat$age60,naive=TRUE)
## print the results
obj
# }