Skip to contents

The EGT03 Table entries provide the number of patients by baseline assessment and minimum or maximum post-baseline assessment. Percentages are based on the total number of patients in a treatment group. Baseline is the patient's last observation prior to initiation of study drug.

Usage

egt03_main(
  adam_db,
  arm_var = "ACTARMCD",
  summaryvar = "BNRIND",
  splitvar = "ANRIND",
  visitvar = "AVISIT",
  page_var = "PARAMCD",
  ...
)

egt03_pre(adam_db, ...)

egt03_post(tlg, prune_0 = FALSE, ...)

egt03

Format

An object of class chevron_t of length 1.

Arguments

adam_db

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

arm_var

(character) the arm variables used for row split, typically "ACTARMCD".

summaryvar

(character) variables to be analyzed, typically "BNRIND". Labels of the corresponding columns are used as subtitles.

splitvar

(character) variables to be analyzed, typically "ANRIND". Labels of the corresponding columns are used as subtitles.

visitvar

(string) typically "AVISIT" or user-defined visit incorporating "ATPT".

page_var

(string) variable name prior to which the row split is by page.

...

not used.

tlg

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

prune_0

(flag) remove 0 count rows

Value

the main function returns an rtables object.

the preprocessing function returns a list of data.frame.

the postprocessing function returns an rtables object or an ElementaryTable (null report).

Details

  • ADEG data are subsetted to contain only "POST-BASELINE MINIMUM"/"POST-BASELINE MAXIMUM" visit according to the preprocessing.

  • Percentages are based on the total number of patients in a treatment group.

  • Split columns by Analysis Reference Range Indicator, typically ANRIND.

  • Does not include a total column by default.

  • Sorted based on factor level.

Functions

  • egt03_main(): Main TLG function

  • egt03_pre(): Preprocessing

  • egt03_post(): Postprocessing

Note

  • adam_db object must contain an adeg table with a "ACTARMCD" column as well as columns specified in summaryvar and splitvar.

Examples

library(dunlin)

proc_data <- log_filter(syn_data, PARAMCD == "HR", "adeg")
run(egt03, proc_data)
#>   Actual Arm Code                            Minimum Post-Baseline Assessment     
#>     Baseline Reference Range Indicator      LOW         NORMAL      HIGH   Missing
#>   ————————————————————————————————————————————————————————————————————————————————
#>   Heart Rate                                                                      
#>     ARM A (N=15)                                                                  
#>       LOW                                    0             0         0        0   
#>       NORMAL                             4 (26.7%)     9 (60.0%)     0        0   
#>       HIGH                                   0         2 (13.3%)     0        0   
#>       Missing                                0             0         0        0   
#>     ARM B (N=15)                                                                  
#>       LOW                                    0         1 (6.7%)      0        0   
#>       NORMAL                             4 (26.7%)    10 (66.7%)     0        0   
#>       HIGH                                   0             0         0        0   
#>       Missing                                0             0         0        0   
#>     ARM C (N=15)                                                                  
#>       LOW                                    0             0         0        0   
#>       NORMAL                             4 (26.7%)    10 (66.7%)     0        0   
#>       HIGH                                   0         1 (6.7%)      0        0   
#>       Missing                                0             0         0        0