Build {gtsummary} with ARDs

We will use the tbl_summary() function to illustrate how to build a new gtsummary table.

Imagine we wish to construct the table below: it can be done both by passing a data frame to tbl_summary() or an ARD to tbl_ard_summary(). These functions share most of their internals.

library(gtsummary)
library(cards)

tbl_summary(
  trial, 
  by = trt, 
  include = c(marker, age, grade),
  type = age ~ "continuous2",
  statistic = all_continuous2() ~ c("{mean} ({sd})", "{median} ({p25}, {p75})"),
  missing = "no"
)

Characteristic	Drug A N = 98 1	Drug B N = 102 1
Marker Level (ng/mL)	0.84 (0.23, 1.60)	0.52 (0.18, 1.21)
Age
Mean (SD)	47 (15)	47 (14)
Median (Q1, Q3)	46 (37, 60)	48 (39, 56)
Grade
I	35 (36%)	33 (32%)
II	32 (33%)	36 (35%)
III	31 (32%)	33 (32%)
1 Median (Q1, Q3); n (%)

# build the ARD with the {cards} package
ard_for_table <- 
  ard_stack(
    trial,
    ard_continuous(variables = c(marker, age)),
    ard_categorical(variables = grade),
    .by = trt,
    .missing = TRUE,                     # we use missing information in gtsummary
    .attributes = TRUE,                  # we also use attributes like column labels
    .total_n = TRUE                      # add total N to the ARD
  )

# then convert it to a table with `tbl_ard_summary()`
ard_for_table |> 
  tbl_ard_summary(
    by = trt,
    type = age ~ "continuous2",
    statistic = all_continuous2() ~ c("{mean} ({sd})", "{median} ({p25}, {p75})")
  )

Characteristic	Drug A 1	Drug B 1
Marker Level (ng/mL)	0.8 (0.2, 1.6)	0.5 (0.2, 1.2)
Age
Mean (SD)	47.0 (14.7)	47.4 (14.0)
Median (Q1, Q3)	46.0 (37.0, 60.0)	48.0 (39.0, 56.0)
Grade
I	35 (35.7%)	33 (32.4%)
II	32 (32.7%)	36 (35.3%)
III	31 (31.6%)	33 (32.4%)
1 Median (Q1, Q3); n (%)