Skip to contents

The CDISC Analysis Results Standard aims to facilitate automation, reproducibility, reusability, and traceability of analysis results data (ARD). The {cards} package creates these CDISC Analysis Result Data Sets.

Use cases:

  1. Quality Control (QC) of existing tables and figures.

  2. Pre-calculate statistics to be summarized in tables and figures.

  3. Medical writers may easily access statistics and place in reports without copying and pasting from reports.

  4. Provides a consistent format for results and lends results to be combined across studies for re-use and re-analysis.

Installation

Install cards from CRAN with:

You can install the development version of cards from GitHub with:

# install.packages("devtools")
devtools::install_github("insightsengineering/cards")

Extensions

cardx website

The {cards} package exports three types of functions:

  1. Functions to create basic ARD objects.

  2. Utilities to create new ARD objects.

  3. Functions to work with existing ARD objects.

The {cardx} R package is an extension to {cards} that uses the utilities from {cards} and exports functions for creating additional ARD objects––including functions to summarize t-tests, Wilcoxon Rank-Sum tests, regression models, and more.

Getting Started

Review the Getting Started page for examples using ARDs to calculate statistics to later include in tables.

library(cards)

ard_continuous(ADSL, by = "ARM", variables = "AGE")
#> {cards} data frame: 24 x 10
#>    group1 group1_level variable stat_name stat_label   stat
#> 1     ARM      Placebo      AGE         N          N     86
#> 2     ARM      Placebo      AGE      mean       Mean 75.209
#> 3     ARM      Placebo      AGE        sd         SD   8.59
#> 4     ARM      Placebo      AGE    median     Median     76
#> 5     ARM      Placebo      AGE       p25         Q1     69
#> 6     ARM      Placebo      AGE       p75         Q3     82
#> 7     ARM      Placebo      AGE       min        Min     52
#> 8     ARM      Placebo      AGE       max        Max     89
#> 9     ARM    Xanomeli…      AGE         N          N     84
#> 10    ARM    Xanomeli…      AGE      mean       Mean 74.381
#> ℹ 14 more rows
#> ℹ Use `print(n = ...)` to see more rows
#> ℹ 4 more variables: context, fmt_fn, warning, error

Other Resources