Change statistics argument so, that dichotomous variables show number of both manifestations separated by slash ( / )

I need to prepare a clinical characteristics table showing not just the frequency of one manifestation of a dichotomous variable, but both on the same line seperated by a / (slash).

For example:

Sex (female, male): 10/8

However, I'm not able to do that. The only thing I can do is number in relation to total number of observations:

data("mtcars")
require(gt_summary)

tbl_summary(mtcars,
            statistic = list(
            all_continuous() ~ "{mean}±{sd}",
            all_dichotomous() ~ "{n}/{N}"
        )
            )

I also tried:

tbl_summary(mtcars,
            statistic = list(
            all_continuous() ~ "{mean}±{sd}",
            all_dichotomous() ~ "{n}/{N-n}"
        )
            )

but that won't work.

The journal unfortunately forces this style.

Any ideas?

This required some workaround to achieve using the "stargazer" package. See code below.

install.packages(c("tidyverse","foreach","stargazer"))
library(tidyverse)

data(mtcars)
mtcars[c(2,8:11)] <- lapply(mtcars[c(2,8:11)], function(x) factor(x, ordered = FALSE)

a <- mtcars %>% 
select_if(~is.numeric(.)) %>% 
summarise(across(everything(), list(N = ~str_c(round(mean(.),2), " ± ", round(sd(.),2))))) %>% 
as.data.frame() %>% 
pivot_longer(
  cols = everything(),
  names_to = c("Characteristic", ".value"),
  names_sep = "_"
)

categorical_var <- mtcars %>% select_if(~!is.numeric(.)) %>% colnames(.) %>% as.vector()

library(foreach)

b <- foreach(i = categorical_var, .combine = 'rbind') %do% {
  temp <- mtcars %>% 
select(i) %>% 
group_by(get(i)) %>% 
count() %>% 
rename(Characteristic = `get(i)`, N = n)
  temp[[1]] <- str_c(i, " ( ", paste0(temp[[1]], collapse=", "), " )")
  temp[[2]] <- str_c(temp[[2]], collapse=" / ")
  temp
}

b <- b %>% distinct(.)

c <- rbind(a,b)

library(stargazer)
stargazer(c, type="html", summary=F)

Running the function "stargazer" with type set to "html" will print out the HTML code for the table in the R console. You can copy-paste the code generated into a text editor (I prefer Notepad++) to save it as a HTML file type, and open it in a web browser of your choice (Edge, Chrome, Firefox, etc.). The final table will look like this:

Note - All the individual code chunks can be packaged nicely into a custom function as well. Feel free to modify as needed.