Show only the difference column when using add_difference function in gt_summary

Question

When creating a table with the add_difference function like this

  trial %>%
  select(trt, age, marker, response, death) %>%
  tbl_summary(
    by = trt,
    missing = "no"
  ) %>%
  add_difference()

You get a table with the difference column and p_value and CI columns.

I want to remove those columns as I want to add p_values for a nonparametric test for continuous variables, which to my understanding from this documentation, can't be done using add_difference.

If I use modify_column_hide(columns = c(p.value, ci)) and than add_p i'll get an error:

trial %>%
  select(trt, age, marker, response, death) %>% tbl_summary(
    by = trt,
    missing = "no" ) %>%
  add_difference() %>% 
  modify_column_hide(columns = c(p.value, ci)) %>% 
  add_p()

#Error: `add_p()` cannot be run after `add_difference()` or `add_p()` when a #'p.value' column is already present.

What is the correct approach to do so?

Answer 1

Here is how i would do it...

library(gtsummary)
packageVersion("gtsummary")
#> [1] '1.7.0'

trial %>%
  select(trt, age, marker, response, death) %>% tbl_summary(
    by = trt,
    missing = "no" ) %>%
  add_difference() %>% 
  modify_column_hide(columns = c(p.value, ci)) %>% # hide the CI and p-value
  modify_table_body(~.x %>% select(-p.value)) %>% # remove the p-value from the table 
  modify_footnote(estimate = NA) %>% # remove difference footnote
  add_p() %>% # add new p-values
  as_kable() # convert to kable so it will display on stackoverflow

Characteristic	Drug A, N = 98	Drug B, N = 102	Difference	p-value
Age	46 (37, 59)	48 (39, 56)	-0.44	0.7
Marker Level (ng/mL)	0.84 (0.24, 1.57)	0.52 (0.19, 1.20)	0.20	0.085
Tumor Response	28 (29%)	33 (34%)	-4.2%	0.5
Patient Died	52 (53%)	60 (59%)	-5.8%	0.4

^{Created on 2023-01-26 with reprex v2.0.2}