R: Derive Treatment-emergent Flag

derive_var_trtemfl {admiral}

R Documentation

Derive Treatment-emergent Flag

Description

Derive treatment emergent analysis flag (e.g., TRTEMFL).

Usage

derive_var_trtemfl(
  dataset,
  new_var = TRTEMFL,
  start_date = ASTDTM,
  end_date = AENDTM,
  trt_start_date = TRTSDTM,
  trt_end_date = NULL,
  end_window = NULL,
  ignore_time_for_trt_end = TRUE,
  initial_intensity = NULL,
  intensity = NULL
)

Arguments

`dataset`	Input dataset The variables specified by `start_date`, `end_date`, `trt_start_date`, `trt_end_date`, `initial_intensity`, and `intensity` are expected.
`new_var`	New variable
`start_date`	Event start date Permitted Values: A symbol referring to a date or datetime variable of the input dataset
`end_date`	Event end date Permitted Values: A symbol referring to a date or datetime variable of the input dataset
`trt_start_date`	Treatment start date Permitted Values: A symbol referring to a date or datetime variable of the input dataset
`trt_end_date`	Treatment end date Permitted Values: A symbol referring to a date or datetime variable of the input dataset or `NULL`
`end_window`	If the argument is specified (in 'days'), events starting more than the specified number of days after end of treatment, are not flagged. Permitted Values: A non-negative integer or `NULL`
`ignore_time_for_trt_end`	If the argument is set to `TRUE`, the time part is ignored for checking if the event occurred more than `end_window` days after end of treatment. Permitted Values: `TRUE`, `FALSE`
`initial_intensity`	Initial severity/intensity or toxicity This derivation assumes AE data collection method as single record per AE with "initial" and "most extreme" severity/intensity recorded separately. If the argument is specified, events which start before treatment start and end after treatment start (or are ongoing) and worsened (i.e., the intensity is greater than the initial intensity), are flagged. The values of the specified variable must be comparable with the usual comparison operators. I.e., if the intensity is greater than the initial intensity `initial_intensity < intensity` must evaluate to `TRUE`. Permitted Values: A symbol referring to a variable of the input dataset or `NULL`
`intensity`	Severity/intensity or toxicity If the argument is specified, events which start before treatment start and end after treatment start (or are ongoing) and worsened (i.e., the intensity is greater than the initial intensity), are flagged. The values of the specified variable must be comparable with the usual comparison operators. I.e., if the intensity is greater than the initial intensity `initial_intensity < intensity` must evaluate to `TRUE`. Permitted Values: A symbol referring to a variable of the input dataset or `NULL`

Details

For the derivation of the new variable the following cases are considered in this order. The first case which applies, defines the value of the variable.

not treated: If trt_start_date is NA, it is set to NA_character_.
event before treatment: If end_date is before trt_start_date (and end_date is not NA), it is set to NA_character_.
no event date: If start_date is NA, it is set to "Y" as in such cases it is usually considered more conservative to assume the event was treatment-emergent.
event started during treatment:
- if end_window is not specified: if start_date is on or after trt_start_date, it is set to "Y",
- if end_window is specified: if start_date is on or after trt_start_date and start_date is on or before trt_end_date + end_window days, it is set to "Y",
event started before treatment and (possibly) worsened on treatment:
- if initial_intensity and intensity is specified: if initial_intensity < intensity and start_date is before trt_start_date and end_date is on or after trt_start_date or end_date is NA, it is set to "Y".
Otherwise it is set to NA_character_.

Value

The input dataset with the variable specified by new_var added

Examples


library(tibble)
library(dplyr, warn.conflicts = FALSE)
library(lubridate)

adae <- expected <- tribble(
  ~USUBJID, ~ASTDTM,            ~AENDTM,            ~AEITOXGR, ~AETOXGR,
  # before treatment
  "1",      "2021-12-13T20:15", "2021-12-15T12:45", "1",       "1",
  "1",      "2021-12-14T20:15", "2021-12-14T22:00", "1",       "3",
  # starting before treatment and ending during treatment
  "1",      "2021-12-30T20:00", "2022-01-14T11:00", "1",       "3",
  "1",      "2021-12-31T20:15", "2022-01-01T01:23", "1",       "1",
  # starting during treatment
  "1",      "2022-01-01T12:00", "2022-01-02T23:25", "3",       "4",
  # after treatment
  "1",      "2022-05-10T11:00", "2022-05-10T13:05", "2",       "2",
  "1",      "2022-05-11T11:00", "2022-05-11T13:05", "2",       "2",
  # missing dates
  "1",      "",                 "",                 "3",       "4",
  "1",      "2021-12-30T09:00", "",                 "3",       "4",
  "1",      "2021-12-30T11:00", "",                 "3",       "3",
  "1",      "",                 "2022-01-04T09:00", "3",       "4",
  "1",      "",                 "2021-12-24T19:00", "3",       "4",
  "1",      "",                 "2022-06-04T09:00", "3",       "4",
  # without treatment
  "2",      "",                 "2021-12-03T12:00", "1",       "2",
  "2",      "2021-12-01T12:00", "2021-12-03T12:00", "1",       "2",
  "2",      "2021-12-06T18:00", "",                 "1",       "2"
) %>%
  mutate(
    ASTDTM = ymd_hm(ASTDTM),
    AENDTM = ymd_hm(AENDTM),
    TRTSDTM = if_else(USUBJID == "1", ymd_hm("2022-01-01T01:01"), ymd_hms("")),
    TRTEDTM = if_else(USUBJID == "1", ymd_hm("2022-04-30T23:59"), ymd_hms(""))
  )

# derive TRTEMFL without considering treatment end and worsening
derive_var_trtemfl(adae) %>% select(ASTDTM, AENDTM, TRTSDTM, TRTEMFL)

# derive TRTEM2FL taking treatment end and worsening into account
derive_var_trtemfl(
  adae,
  new_var = TRTEM2FL,
  trt_end_date = TRTEDTM,
  end_window = 10,
  initial_intensity = AEITOXGR,
  intensity = AETOXGR
) %>% select(ASTDTM, AENDTM, AEITOXGR, AETOXGR, TRTEM2FL)

[Package admiral version 1.0.2 Index]