ABC.perm: The permutation test of area between the CIF curves for...

In ComparisonCR: Comparison of Cumulative Incidence Between Two Groups Under Competing Risks

Description

Statistical inference of area between the CIF curves (ABC) test with a permutation procedure for competing risk data.

Usage

ABC.perm(time, status, group, t0=0, tau=NULL, nperm=1000, seed=12345, bias=FALSE)

Arguments

time	The followed up time for testing data.
status	The status indicator, should be coded as 0= censored, 1= event of interest, 2= all other competing events.
group	The group indicator for comparison, and the elements of this vector should take either 0 or 1. Normally, 0= control group, 1= treatment group.
t0	The start time point for testing, the default is set as t0=0 for overall test.
tau	The truncation time point for testing, which needs to be smaller than or equal to the minimum of the largest observed time in each of the two groups. When tau=NULL, the default value is used. See details in reference.
nperm	The times of permutation, with default nperm=1000.
seed	The seed number, with default seed=12345.
bias	If bias=TRUE, the results will print the mean difference for delta of simulated datasets and original data. The default is bias=FALSE

Value

t0	The start time point for testing.
tau	The truncation time point for testing.
delta	The alsolute difference of two cumulative incidence functions.
var(delta)	The variance for delta of simulated datasets.
bias	The mean difference for delta of simulated datasets and original data.
Pvalue	The P value of this test.

References

Lyu J, Chen J, Hou Y, Chen Z. Comparison of two treatments in the presence of competing risks. Pharmaceutical Statistics, 2020. DOI: 10.1002/pst.2028.

Examples

#get dataset from package
data(crossdata)
#just for an example, we set resampling times at 10
#overall test
ABC.perm(crossdata$time, crossdata$status, crossdata$group, nperm=10)
#arbitary test for detecting difference after 2 years
ABC.perm(crossdata$time, crossdata$status, crossdata$group, t0=2, tau=NULL, nperm=10)
#arbitary test for detecting difference between 2 years and 4 years
ABC.perm(crossdata$time, crossdata$status, crossdata$group, t0=2, tau=4, nperm=10)

Example output

The permutation resampling times =10


  t0 tau delta var.delta. Pvalue
1  0 4.6   0.6 0.01052004      0
The permutation resampling times =10


  t0 tau delta  var.delta. Pvalue
1  2 4.6 0.256 0.007491378    0.1
The permutation resampling times =10


  t0 tau delta  var.delta. Pvalue
1  2   4 0.223 0.005915733    0.1

ComparisonCR documentation built on July 1, 2020, 8:12 p.m.