Subject Level Summary Table • carver

Introduction

This article describes creating reports which can be classified as ADSL Summary Tables using adsl_summary function. Example here is created using ADSL domain.

Note: All examples assume CDISC ADaM format as input unless otherwise specified.

Programming Flow

Read in Data
Preprocess Data
ADSL Summary Table
ADSL Table Display

Read in Data

library(carver)
library(dplyr)
library(purrr) # nolint
data(adsl)

Preprocess Data

Before creating adsl_summary, it is required to use the pre-processing function: mentry which serves as an essential step in preparing the data for summarisation.

`mentry()`

The next step in pre-processing involves the usage of mentry function to read and process data with subsets and variables. It takes in the adsl as input . The input data gets processed based on TRT(mandatory variable for generating columns in the final table), BY(optional variable used to group rows in the final table) and/or SUBGROUP(optional variable used to add subgroups under the columns-TRT in the final table) variables and filtered by subset conditions.

Note: For additional information such as parameters info, refer to the mentry() documentation.

mentry_df <- adsl |>
  mentry(
    subset = NA_character_,
    byvar = NA_character_,
    trtvar = "TRT01A",
    trtsort = "TRT01AN",
    subgrpvar = NA_character_,
    trttotalyn = "Y",
    add_grpmiss = "N",
    pop_fil = "SAFFL"
  )

ADSL Summary Table

This step is used in creating the ADSL summary table using the ‘adsl_summary’ function.

The following are the arguments for the table:

datain: The input data for table generation. The output data from mentry_df function.
vars: Names of adsl variables to display in final output which are tilde-separated, Add "-S" for numeric variables.
stat_vars: Statistics to display in table for numeric vars, tilde-separated.
pctdisp: Denominator to calculate percentages by. Values: "TRT", "VAR", "COL", "SUBGRP", "SGRPN", "CAT", "NONE", "NO", "DPTVAR".
total_catyn To return a ‘Total’ row for categorical analysis in vars. Values: "Y"/"N"
total_catlabel Label for total category row. eg- “All”/“Total”
miss_catyn To include empty/blank values as miss_catlabel in categories of dptvar variable or not. Values: "Y"/"N"
miss_catlabel Label for missing values
a_subset Analysis Subset condition; tilde-separated for each variable in vars.
denom_subset Subset condition to be applied to dataset for calculating denominator, tilde-separated for categorical variables within vars.

adsl_sum <- mentry_df |>
  adsl_summary(
    vars = "AGEGR1/AGEGR1N~AGE-S~SEX/SEXN~RACE/RACEN",
    stat_vars = "N~Range~Meansd~Median~IQR",
    pctdisp = "TRT",
    total_catyn = "N",
    total_catlabel = "Total",
    miss_catyn = "N",
    miss_catlabel = "Missing",
    a_subset = "AGE<65~AGE>80~SEX=='F'~NA",
    denom_subset = NA_character_
  )

ADSL Table Display

Prepare data for general table display

This step is used creating the ADSL summary table using ‘tbl_processor’ function to prepare data for general table display.

The following are the arguments:

datain Input dataframe. The output data from adsl_summary function.
dptlabel Tilde-separated labels to set to category variables in data. If analysis vars (DPTVAR) contains AGEGR1, RACE; dptlabel may be "Age Group~Race".
statlabel Tilde-separated labels corresponding to each Statistic in data.
extra_df Additional dataframe, to merge with datain.
extra_mergeby Variables to merge extra_df by, if present.
dropvars Variables additional to standard present in input data to be removed from output
disp_value_col Hide/Show value column from the final display.
addrowvars Group Variable(s) to be removed as a column and instead used as row headers in the value column.

Create flextable output from display templates

This step is used creating the ADSL summary table using ‘tbl_display’ to create flextable output from display templates.

The following are the arguments:

datain Input dataframe. The output data from tbl_processor function.
bylabel Change BYVAR names to label, if any.
dpthead String to become name of the column containing categories (DPTVAL) in output.
font Font face for text inside table
fontsize Font size for text inside table

ADSL Table Display Output

The ADSL Summary Table is generated successfully. Below is the Table generated:

adsl_sum |>
  display_bign_head(mentry_data = mentry_df) |>
  tbl_processor(
    statlabel = "N~Range~Meansd~Median~IQR",
    dptlabel = "Age Group~NONE~Sex~Race",
    addrowvar = "DPTVAR"
  ) |>
  tbl_display() |>
  flextable::autofit()

	Placebo (N=86)	Xanomeline Low Dose (N=84)	Xanomeline High Dose (N=84)	Total (N=254)
Age Group
<65	14 (16.28%)	8 ( 9.52%)	11 (13.10%)	33 (12.99%)
NONE
N	30	29	18	77
Range	_NO_STAT	_NO_STAT	_NO_STAT	_NO_STAT
Meansd	_NO_STAT	_NO_STAT	_NO_STAT	_NO_STAT
Median	84.00	83.00	83.50	84.00
IQR	4.75	2.00	2.75	3.00
Sex
F	53 (61.63%)	50 (59.52%)	40 (47.62%)	143 (56.30%)
Race
WHITE	78 (90.70%)	78 (92.86%)	74 (88.10%)	230 (90.55%)
BLACK OR AFRICAN AMERICAN	8 ( 9.30%)	6 ( 7.14%)	9 (10.71%)	23 ( 9.06%)
AMERICAN INDIAN OR ALASKA NATIVE	0	0	1 ( 1.19%)	1 ( 0.39%)