| Type: | Package |
| Title: | Inference for Model-Free Between-Group Parameters for Censored Survival Data |
| Version: | 1.0-5 |
| Date: | 2017-06-23 |
| Author: | Lu Tian, Hajime Uno, Miki Horiguchi |
| Maintainer: | Miki Horiguchi <horiguchimiki@gmail.com> |
| Description: | Performs inference of several model-free group contrast measures, which include difference/ratio of cumulative incidence rates at given time points, quantiles, and restricted mean survival times (RMST). Two kinds of covariate adjustment procedures (i.e., regression and augmentation) for inference of the metrics based on RMST are also included. |
| Imports: | flexsurv, plotrix, survival |
| License: | GPL-2 |
| Encoding: | UTF-8 |
| LazyData: | true |
| RoxygenNote: | 6.0.1 |
| Suggests: | knitr, rmarkdown |
| NeedsCompilation: | no |
| Packaged: | 2017-06-22 21:09:19 UTC; Miki |
| Repository: | CRAN |
| Date/Publication: | 2017-06-23 08:51:58 UTC |
Inference for Model-Free Between-Group Parameters For Censored Survival Data
Description
Performs inference of several model-free group contrast measures, which include difference/ratio of cumulative incidence rates at given time points, quantiles, and restricted mean survival times (RMST). Two kinds of covariate adjustment procedures (i.e., regression and augmentation) for inference of the metrics based on RMST are also included.
Author(s)
Lu Tian, Hajime Uno, Miki Horiguchi
Maintainer: Miki Horiguchi <horiguchimiki@gmail.com>
References
Tian L, Zhao L, Wei LJ. Predicting the restricted mean event time with the subject's baseline covariates in survival analysis. Biostatistics 2014, 15, 222-233.
Zhao L, Tian L, Uno H, Solomon S, Pfeffer M, Schindler J, Wei LJ. Utilizing the integrated difference of two survival functions to quantify the treatment contrast for designing, monitoring, and analyzing a comparative clinical study. Clinical Trials 2012, 9, 570-577.
See Also
flexsurv plotrix survival
Edit pbc data to run sample code
Description
Edit pbc data in survival package and make it ready to run the sample code in this manual.
Usage
pbc.sample()
See Also
pbc in survival package
Plot method for surv2sample objects
Description
Creates plots from a surv2sample object.
Usage
## S3 method for class 'surv2sample'
plot(x, measure = NULL, baseline = 0, ...)
Arguments
x |
surv2sample object |
measure |
The type of measure used for the plot. When default(=NULL), plot.survfit() is called and KM plots are given. When "relative time" is specified, a plot of relative percentiles with corresponidng 0.95 confidence intervals is generatead. |
baseline |
Indicates the baseline group, 0/1. Default is 0. |
... |
For further method |
See Also
plotCI in plotrix package
Adjusted difference/ratio of restricted mean survival times
Description
Compares restricted mean survival time between two groups, adjusting for imbalance of baseline factors.
Usage
rmstaug(y, delta, x, arm, tau, type="difference", conf.int=0.95)
Arguments
y |
The follow-up time. |
delta |
The censoring indicator, 1=event, and 0=censoring. |
x |
The covariate matrix. The group indicator, arm (below) should not be included in this matrix. |
arm |
The group indicator, 1/0. |
tau |
The value indicates the restricted time point on the follow-up time to calculate the restricted mean survival time. |
type |
The type of the between-group contrast measure: "difference"(default), "ratio" or "lossratio". |
conf.int |
The level for computation of the confidence intervals. The default is 0.95. |
Author(s)
Lu Tian
References
Tian L, Zhao L, Wei LJ. Predicting the restricted mean event time with the subject's baseline covariates in survival analysis. Biostatistics 2014, 15, 222-233.
Examples
D=pbc.sample()
rmstaug(D$time, D$status, D$covariates, D$group, tau=8, type="difference")
Adjusted difference/ratio of restricted mean survival times
Description
Compares restricted mean survival time between two groups, adjusting for imbalance of baseline factors via a regression model.
Usage
rmstreg(y, delta, x, arm, tau, type="difference", conf.int=0.95)
Arguments
y |
The follow-up time. |
delta |
The censoring indicator, 1=event, and 0=censoring. |
x |
The covariate matrix. The first colomn of this matrix should be the group indicator, arm (below). |
arm |
The group indicator, 1/0. |
tau |
The value indicates the restricted time point on the follow-up time to calculate the restricted mean survival time. |
type |
The type of the between-group contrast measure: "difference"(default), "ratio" or "lossratio". |
conf.int |
The level for computation of the confidence intervals. The default is 0.95. |
Author(s)
Lu Tian
References
Tian L, Zhao L, Wei LJ. Predicting the restricted mean event time with the subject's baseline covariates in survival analysis. Biostatistics 2014, 15, 222-233.
Examples
D=pbc.sample()
x=cbind(D$group, D$covariates)
rmstreg(D$time, D$status, x, D$group, tau=8, type="difference")
Inference of model-free between-group contrasts with censored survival data
Description
Performs inference of several model-free group contrast measures, which include difference/ratio of cumulative incidence rates, quantiles, restricted mean survival times (RMST), and integrated survival rates.
Usage
surv2sample(time, status, arm, npert=1000,
timepoints=c(12, 24, 36, 40), quanprobs=c(0.1, 0.15, 0.2),
tau_start=0, tau, SEED=NULL, procedure="KM", conf.int=0.95)
Arguments
time |
The follow-up time. |
status |
The censoring indicator, 1=event, and 0=censoring. |
arm |
The indicator for groups to compare 1/0. |
npert |
The number of resampling. The default is 1000. |
timepoints |
specifies the time points at which difference and ratio of the survival rates are computed. |
quanprobs |
specifies the probabilities at which difference and ratio of the corresponding quantiles are computed. |
tau_start |
The value indicates time point on the follow-up time to calculate the restricted mean survival time beyond the time point. The default is 0. |
tau |
The value indicates the restricted time point on the follow-up time to calculate the restricted mean survival time. (i.e., the minimum of the largest observed time in each of the two groups) |
SEED |
A random seed used for the resampling. Default is NULL. |
procedure |
Specifies the inference procedure. A non-parametric procedure by the method of Kaplan-Meier ("KM") is the default. Another option is a parametric inference procedure by fitting a generalized gamma distribution to each group ("GG"). |
conf.int |
The level for computation of the confidence intervals. The default is 0.95. |
Author(s)
Hajime Uno, Miki Horiguchi
References
Tian L, Zhao L, Wei LJ. Predicting the restricted mean event time with the subject's baseline covariates in survival analysis. Biostatistics 2014, 15, 222-233.
Zhao L, Tian L, Uno H, Solomon S, Pfeffer M, Schindler J, Wei LJ. Utilizing the integrated difference of two survival functions to quantify the treatment contrast for designing, monitoring, and analyzing a comparative clinical study. Clinical Trials 2012, 9, 570-577.
Examples
D=pbc.sample()
surv2sample(D$time, D$status, D$group, npert=500, timepoints=c(2,4,6,8),
quanprobs =c(0.2, 0.3), tau=8, procedure="KM")