All Individual Subject Deaths, Safety Population, Pooled Analysis (or Trial X)

FDA Table 09

FDA IG Shell

Programming Notes

Background and Instructions

Table 9: All Individual Subject Deaths provides a list of all subject deaths from the AE and disposition datasets. AEs with an outcome of death from AE datasets and subjects who died due to natural causes are only listed in the disposition datasets. The trial day of death below shows the date of death, not the date of onset of SAE leading to death.

Customization

After analyzing the initial presentation of Table 9, the clinical reviewer may discuss with the CDS whether an additional column for investigator’s assessment of relatedness or clinical reviewer’s assessment of relatedness is desired. The clinical reviewer may also request from the CDS a graphical subject profile for subjects who died.

Output

Full Code

# FDA TABLE 09 ---------------------------------------------------------------
# Self-contained template. Run top-to-bottom to build the objects below.
# Used by tests via run_template() and published in the catalog.

library(dplyr)
library(gtsummary)
library(crane)

adae <- pharmaverseadam::adae
adex <- pharmaverseadam::adex

# Pre-processing --------------------------------------------
# deaths
adae <- adae |>
  filter(
    # safety population
    SAFFL == "Y",
    # deaths
    DTHFL == "Y"
  ) |>
  # select variables from `adae` to include in final result
  select(USUBJID, TRT01A, AGE, SEX, DTHADY, DTHCAUS) |>
  # keep one row per unique ID
  distinct(USUBJID, DTHCAUS, DTHADY, .keep_all = TRUE)

# dosing
adex <- adex |>
  filter(
    # safety population
    SAFFL == "Y",
    # total dosages
    PARAMCD == "TDOSE"
  ) |>
  # select variables from `adex` to include in final result
  select(USUBJID, AVAL, TRTSDT, TRTEDT) |>
  mutate(
    # derive dosage duration
    DOSDUR = (TRTEDT - TRTSDT + 1) |> as.character()
  )

# combine all data
data <- left_join(adae, adex, by = "USUBJID") |>
  select(TRT01A, USUBJID, AGE, SEX, DOSDUR, DTHADY, DTHCAUS) |>
  arrange()

# Table 9 is one of the FDA IG's subject-level listings (titled "All
# Individual Subject Deaths"), so it uses tbl_listing() rather than a summary
# constructor. The other IG listings (Tables 39, 41, 42, 54, 55) are not in
# the catalog yet; when added they should use tbl_listing() too.
tbl <- crane::tbl_listing(data) |>
  # set table header labels
  modify_header(
    TRT01A = "**Treatment Arm**",
    USUBJID = "**Unique  \n Subject ID**",
    AGE = "**Age**",
    SEX = "**Sex**",
    DOSDUR = "**Dosing**  \n **Duration**  \n **(Days)**",
    DTHADY = "**Study**  \n **Day of**  \n **Death**",
    DTHCAUS = "**Cause of Death**"
  ) |>
  # align all columns left
  modify_column_alignment(everything(), align = "left")

# A listing has no ARD; the full unformatted dataset (`data`) is the result.

Table

tbl

Treatment Arm	Unique Subject ID	Age	Sex	Dosing Duration (Days)	Study Day of Death	Cause of Death
Xanomeline Low Dose	01-701-1211	76	F	59	61	SUDDEN DEATH
Placebo	01-704-1445	75	M	175	175	COMPLETED SUICIDE
Placebo	01-710-1083	89	F	11	12	MYOCARDIAL INFARCTION

Data

This table is a listing, so there is no ARD. The underlying dataset is the result.

withr::local_options(width = 9999, tibble.print_max = Inf)
data

TRT01A	USUBJID	AGE	SEX	DOSDUR	DTHADY	DTHCAUS
Xanomeline Low Dose	01-701-1211	76	F	59	61	SUDDEN DEATH
Placebo	01-704-1445	75	M	175	175	COMPLETED SUICIDE
Placebo	01-710-1083	89	F	11	12	MYOCARDIAL INFARCTION