Introduction to {crane}

What is {crane}?

The {crane} package supplements the {gtsummary} package to ease the process of creating Roche clinical trial tables. {gtsummary} is a general framework for building tables, and {crane} utilizes the frame work and exports functions meant to simplify the creation of summary tables in pharmaceutical industry.

The {crane} package includes wrapper functions for {gtsummary} functions to make it easier to apply common pharma-specific modifications to table formats. For example, demographics tables which use certain variables and formatting conventions as a default.

When the crane package is loaded into your R session, a couple of things occur.

• As expected, the {crane} package is loaded, but the {gtsummary}, {cards}, and {cardx} packages are also loaded.
• The Roche gtsummary theme is set by running crane::theme_gtsummary_roche(). This theme changes a few default behaviors for {gtsummary} tables; for example, all p-values are rounded with label_roche_pvalue(), and tables are printed with {flextable} instead of {gt}. See the function’s help file for details.

library(crane)
#> Loading required package: gtsummary
theme_gtsummary_roche()
#> Setting theme "Roche"

When should I add a new function to {crane}?

While all Roche tables can be created using {gtsummary}, for some tables the code can become long. In these cases, adding a new function to {crane} is warranted.

For example, our standard demographics table can be created with gtsummary::tbl_summary(). But for such a common table, the code is somewhat long, and we therefore created crane::tbl_roche_summary().

To create the Roche standard demographics table, we simple call:

cards::ADSL |>
  tbl_roche_summary(
    by = ARM,
    include = c(AGE, RACE),
    nonmissing = "always"
  )

	Placebo (N = 86)	Xanomeline High Dose (N = 84)	Xanomeline Low Dose (N = 84)
Age
n	86	84	84
Mean (SD)	75 (9)	74 (8)	76 (8)
Median	76	76	78
Min - Max	52 - 89	56 - 88	51 - 88
Race
n	86	84	84
AMERICAN INDIAN OR ALASKA NATIVE	0	1 (1.2%)	0
BLACK OR AFRICAN AMERICAN	8 (9.3%)	9 (10.7%)	6 (7.1%)
WHITE	78 (90.7%)	74 (88.1%)	78 (92.9%)

To create this table with gtsummary::tbl_summary(), we’d need the following code. As some of these steps are not the most straightforward, we created the tbl_roche_summary() wrapper.

cards::ADSL |> 
  tbl_summary(
    by = ARM, 
    include = c(AGE, RACE),
    type = list(AGE = "continuous2"),
    statistic = list(AGE = c("{mean} ({sd})", "{median}", "{min} - {max}")),
    digits = all_categorical() ~ list(p = 1),
    missing = "always",
    missing_stat = "{N_nonmiss}",
    missing_text = "n"
  ) |>
    # remove default footnote
    remove_footnote_header(columns = everything()) |>
    # remove the default "Characteristic" header
    modify_header(label = "") |>
    # convert "0 (0.0%)" to "0"
    gtsummary::modify_post_fmt_fun(
      fmt_fun = ~ifelse(. == "0 (0.0%)", "0", .),
      columns = all_stat_cols()
    ) |>
    # sort the missing row to just below the header row
    modify_table_body(
      function(x) {
        x |>
          dplyr::mutate(
            .by = "variable",
            dplyr::across(dplyr::everything(), ~ .[order(row_type != "label", row_type != "missing")])
          )
      }
    )

Another case where we may create helper functions in {crane}, is when the table can only be created in {gtsummary} using an ARD-first approach. Generally, creating an ARD of all results in a summary table, then creating then table is more cumbersome. In these cumbersome cases, a new function in {crane} is warranted.