You can modify the format functions and the expected denominator as follows. Default values are listed in the functions' documentation (e.g. for `?count_occurrences` it is `"count_fraction_fixed_dp"`). Notice that format default for that is `tern::format_count_fraction_fixed_dp` (that I copied and modified according to your needs). The parameter `denom `is the usual handler for changing the denominator. Can be seen in the same documentation page (`?count_occurrences`).
```r library(pharmaverseadam) library(tern) library(dplyr)
adsl <- adsl %>% df_explicit_na()
adae <- adae %>% df_explicit_na()
adae <- adae %>% var_relabel( AEBODSYS = "MedDRA System Organ Class", AEDECOD = "MedDRA Preferred Term" ) %>% dplyr::filter(SAFFL == "Y")
# Define the split function split_fun <- drop_split_levels lyt <- basic_table(show_colcounts = TRUE) %>% split_cols_by(var = "ACTARM") %>% add_overall_col(label = "All Patients") %>% analyze_num_patients( vars = "USUBJID", .stats = c("unique", "nonunique"), .labels = c( unique = "Total number of patients with at least one adverse event", nonunique = "Overall total number of events" ) ) %>% split_rows_by( "AEBODSYS", child_labels = "visible", nested = FALSE, split_fun = split_fun, label_pos = "topleft", split_label = obj_label(adae$AEBODSYS) ) %>% summarize_num_patients( var = "USUBJID", .stats = c("unique", "nonunique"), .labels = c( unique = "Total number of patients with at least one adverse event", nonunique = "Total number of events" ) ) %>% # Here denom is changed and .formats is specified. Notice that this is a copy of # tern::format_count_fraction_fixed_dp count_occurrences( vars = "AEDECOD", .indent_mods = -1L, .stats = "count_fraction_fixed_dp", denom = "n", .formats = list("count_fraction_fixed_dp" = function(x, ...) { attr(x, "label") <- NULL if (any(is.na(x))) { return("NA") } checkmate::assert_vector(x) checkmate::assert_integerish(x[1]) assert_proportion_value(x[2], include_boundaries = TRUE) result <- if (x[1] == 0) { "0" } else if (x[2] == 1) { sprintf("%d (100%%)", x[1]) } else { # browser() sprintf("%d/%d (%.1f%%)", x[1], round(x[1] / x[2], 0), x[2] * 100) } return(result) }) ) %>% append_varlabels(adae, "AEDECOD", indent = 1L)
result <- build_table(lyt, df = adae, alt_counts_df = adsl) |> head(n = 10)
result #> MedDRA System Organ Class Placebo Xanomeline High Dose Xanomeline Low Dose All Patients #> MedDRA Preferred Term (N=86) (N=72) (N=96) (N=306) #> ————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————— #> Total number of patients with at least one adverse event 69 (80.2%) 70 (97.2%) 86 (89.6%) 225 (73.5%) #> Overall total number of events 301 436 454 1191 #> CARDIAC DISORDERS #> Total number of patients with at least one adverse event 13 (15.1%) 15 (20.8%) 16 (16.7%) 44 (14.4%) #> Total number of events 27 30 34 91 #> ATRIAL FIBRILLATION 1/13 (7.7%) 2/15 (13.3%) 2/16 (12.5%) 5/44 (11.4%) #> ATRIAL FLUTTER 0 1/15 (6.7%) 1/16 (6.2%) 2/44 (4.5%) #> ATRIAL HYPERTROPHY 1/13 (7.7%) 0 0 1/44 (2.3%) #> ATRIOVENTRICULAR BLOCK FIRST DEGREE 1/13 (7.7%) 0 1/16 (6.2%) 2/44 (4.5%) #> ATRIOVENTRICULAR BLOCK SECOND DEGREE 2/13 (15.4%) 1/15 (6.7%) 2/16 (12.5%) 5/44 (11.4%) ```
<sup>Created on 2025-07-10 with [reprex v2.1.1](https://reprex.tidyverse.org)</sup>

I would go for something like the following. Note that you can change and make the analysis function fancier if you need.
``` ``` r library(dplyr) library(tern)
adsl <- random.cdisc.data::cadsl adae <- random.cdisc.data::cadae
# Ensure character variables are converted to factors and empty strings and NAs are explicit missing levels. adsl <- df_explicit_na(adsl) adae <- df_explicit_na(adae) %>% var_relabel( AEBODSYS = "MedDRA System Organ Class", AEDECOD = "MedDRA Preferred Term" ) %>% filter(ANL01FL == "Y")
# Define the split function split_fun <- drop_split_levels
# Define the column functions for analysis colfuns <- list( function(x, .spl_context) { # browser() # Debugging point to inspect the context rcell(table(as.character(x)) |> unlist(), format = "xx.x", label = .spl_context$value[nrow(.spl_context)]) }, function(x, .N_row, .spl_context) { # see ?.additional_fun_params rcell(table(as.character(x)) |> unlist() / .N_row , format = "xx.x%", label = .spl_context$value[nrow(.spl_context)]) } )
# build layout lyt <- basic_table() %>% split_rows_by( "AEBODSYS", child_labels = "visible", nested = FALSE, split_fun = split_fun, label_pos = "topleft", split_label = obj_label(adae$AEBODSYS) ) %>% split_rows_by( "AEDECOD", child_labels = "hidden", nested = TRUE, split_fun = split_fun, label_pos = "topleft", split_label = obj_label(adae$AEDECOD) ) %>% split_cols_by("ARM") |> # Add the analysis variables split_cols_by_multivar(vars = c("ARM", "ARM"), varlabels = c("n", "%")) |> # Add the analysis functions analyze_colvars(afun = colfuns, format = "xx.xx")

result <- build_table(lyt, df = adae, alt_counts_df = adsl) result #> MedDRA System Organ Class A: Drug X B: Placebo C: Combination #> MedDRA Preferred Term n % n % n % #> ———————————————————————————————————————————————————————————————————————————— #> cl A.1 #> dcd A.1.1.1.1 57.0 32.8% 43.0 24.7% 74.0 42.5% #> dcd A.1.1.1.2 58.0 32.8% 56.0 31.6% 63.0 35.6% #> cl B.1 #> dcd B.1.1.1.1 44.0 32.1% 43.0 31.4% 50.0 36.5% #> cl B.2 #> dcd B.2.1.2.1 52.0 32.1% 51.0 31.5% 59.0 36.4% #> dcd B.2.2.3.1 50.0 28.9% 55.0 31.8% 68.0 39.3% #> cl C.1 #> dcd C.1.1.1.3 47.0 30.9% 51.0 33.6% 54.0 35.5% #> cl C.2 #> dcd C.2.1.2.1 39.0 28.7% 40.0 29.4% 57.0 41.9% #> cl D.1 #> dcd D.1.1.1.1 52.0 33.3% 40.0 25.6% 64.0 41.0% #> dcd D.1.1.4.2 54.0 36.5% 44.0 29.7% 50.0 33.8% #> cl D.2 #> dcd D.2.1.5.3 49.0 28.7% 57.0 33.3% 65.0 38.0% ```
<sup>Created on 2025-07-01 with </sup>[<sup>reprex v2.1.1</sup>](https://reprex.tidyverse.org)