Skip to contents

COXT01 (Default) Cox Regression Model Table.

coxt01.Rd

Cox models are the most commonly used methods to estimate the magnitude of the effect in survival analyses. It assumes proportional hazards; that is, it assumes that the ratio of the hazards of the two groups (e.g. two arms) is constant over time. This ratio is referred to as the "hazard ratio" and is one of the most commonly reported metrics to describe the effect size in survival analysis.

Usage

coxt01_main(
  adam_db,
  arm_var = "ARM",
  time_var = "AVAL",
  event_var = "EVENT",
  covariates = c("SEX", "RACE", "AAGE"),
  strata = NULL,
  lbl_vars = "Effect/Covariate Included in the Model",
  multivar = FALSE,
  ...
)

coxt01_pre(adam_db, ...)

coxt01_post(tlg, prune_0 = FALSE, ...)

coxt01

Format

An object of class chevron_t of length 1.

Arguments

adam_db

(list of data.frames) object containing the ADaM datasets

arm_var

(string) the arm variable used for arm splitting.

time_var

(string) the time variable in a Cox proportional hazards regression model.

event_var

(string) the event variable in a Cox proportional hazards regression model.

covariates

(character) will be fitted and the corresponding effect will be estimated.

strata

(character) will be fitted for the stratified analysis.

lbl_vars

(string) text label for the a Cox regression model variables.

multivar

(flag) indicator of whether multivariate cox regression is conducted.

...

Further arguments passed to tern::control_coxreg().

tlg

(TableTree, Listing or ggplot) object typically produced by a main function.

prune_0

(flag) remove 0 count rows

Details

  • The reference arm will always the first level of arm_var. Please change the level if you want to change the reference arms.

  • The table allows confidence level to be adjusted, default is two-sided 95%.

  • The stratified analysis is with DISCRETE tie handling (equivalent to tern::control_coxreg(ties = "exact") in R).

  • Model includes treatment plus specified covariate(s) as factor(s) or numeric(s), with "SEX", "RACE" and "AAGE" as default candidates.

  • The selection of the covariates and whether or not there is a selection process (vs. a fixed, pre-specified list) needs to be pre-specified.

  • For pairwise comparisons using the hazard ratio, the value for the control group is the denominator.

  • Keep zero-count rows unless overridden with prune_0 = TRUE.

Functions

  • coxt01_main(): Main TLG function

  • coxt01_pre(): Preprocessing

  • coxt01_post(): Postprocessing

Note

  • adam_db object must contain an adtte table with "PARAMCD", "ARM", "AVAL", "CNSR, and the columns specified by "covariates" which is denoted as c("SEX", "RACE", "AAGE") by default.

Examples

library(dunlin)

proc_data <- log_filter(syn_data, PARAMCD == "CRSD", "adtte")
proc_data <- log_filter(proc_data, ARMCD != "ARM C", "adsl")
proc_data$adtte$ARM <- droplevels(proc_data$adtte$ARM)
run(coxt01, proc_data)
#>   Effect/Covariate Included in the Model       Treatment Effect Adjusted for Covariate     
#>                                              n     Hazard Ratio       95% CI       p-value 
#>   —————————————————————————————————————————————————————————————————————————————————————————
#>   Treatment:                                                                               
#>     B: Placebo vs control (A: Drug X)       268        1.43        (1.06, 1.94)     0.0204 
#>   Covariate:                                                                               
#>     Sex                                     268        1.43        (1.06, 1.94)     0.0208 
#>     RACE                                    268        1.44        (1.06, 1.96)     0.0208 
#>     Age (yr)                                268        1.46        (1.07, 1.98)     0.0154 

run(coxt01, proc_data, covariates = c("SEX", "AAGE"), strata = c("RACE"), conf_level = 0.90)
#>   Effect/Covariate Included in the Model       Treatment Effect Adjusted for Covariate     
#>                                              n     Hazard Ratio       90% CI       p-value 
#>   —————————————————————————————————————————————————————————————————————————————————————————
#>   Treatment:                                                                               
#>     B: Placebo vs control (A: Drug X)       268        1.42        (1.09, 1.84)     0.0274 
#>   Covariate:                                                                               
#>     Sex                                     268        1.42        (1.09, 1.84)     0.0273 
#>     Age (yr)                                268        1.44        (1.11, 1.87)     0.0211