I am in the process of creating gtsummary table based on the mortality status (the variable "fate") of the Bernard data included in the pubh package.

The issue I am facing is that I want to add the percentage of "Dead" and "Alive" next to their count. But since this is the grouping variable, I haven't been able to configure it

This is my sample code for the table:

library(pubh)
library(dplyr)
library(gtsummary)

data("Bernard")


na.omit(Bernard)  %>% select(fate, race, apache) %>%
  tbl_summary(by = fate,
              
     type =  list(race ~ "categorical", apache ~ "continuous"),
     statistic = list(all_continuous() ~ "{min}, {max}", all_categorical() ~ "{p}%"),
     digits = list(all_continuous() ~ 2, all_categorical() ~ 2),
     missing_text = "(Missing)" ) %>% 
                
     add_stat_label() %>%
     modify_header(label ~ "**Variable**") %>%
     modify_caption("**Table 1. Summary statistics by  Mortality Status**") %>%
     modify_spanning_header(c("stat_1", "stat_2") ~ "**Fate**") %>%
     bold_labels() %>%
     italicize_labels() %>%
     italicize_levels() 

And this is the output

Ideally, I would like to have the table show : Alive, N = 96 (67..%) Dead, N = 47 (32..%)

I have tried listing the "fate" variable as categorical and then providing the statistic for percentage:

               type =  list(c(race, **fate**) ~ "categorical", apache ~ "continuous"),
               
               statistic = list(all_continuous() ~ "{min}, {max}", all_categorical() ~ "{p}%", **fate ~ "{p}%"**),

This did not work

And I was also thinking that using mutate to create a new variable before using tbl_summary() would probably work, but I am curious if this can be configured explicitly within tbl_summary()

One could generalize the question to: whether summary statistics other than the number of observations can be shown for the grouping variable in gtsummary

CodePudding user response：

You can add the percentage to the header using the modify_header() function. Example below!

library(gtsummary)
packageVersion("gtsummary")
#> [1] '1.6.3'

trial %>%
  tbl_summary(
    by = trt, 
    include = age
  ) %>%
  modify_header(all_stat_cols() ~ "**{level}**, N={n} ({style_percent(p)}%)") %>%
  as_kable() # convert to kable to display on stackoverflow

^{Created on 2022-12-24 with reprex v2.0.2}