admiralophtha (version 1.2.0)

derive_var_bcvacritxfl: Adds CRITx/CRITxFL pairs to BCVA dataset

Description

The derive_var_bcvacritxfl() function has been superseded in favor of admiral::derive_vars_crit_flag().

Adds a criterion variables CRITx and their corresponding flags CRITxFL to a dataset containing BCVA records

Usage

derive_var_bcvacritxfl(
  dataset,
  crit_var,
  bcva_ranges = NULL,
  bcva_uplims = NULL,
  bcva_lowlims = NULL,
  additional_text = "",
  critxfl_index = NULL
)

Value

The input BCVA dataset with additional column pairsCRITx, CRITxFL.

Arguments

dataset

Input dataset containing BCVA data (usually ADBCVA).

crit_var

Variable with respect to which CRITx/CRITxFL are derived (usually CHG or AVAL).

bcva_ranges

List containing one or more numeric vectors of length 2. For each vector c(a,b) in bcva_ranges, a pair of variables CRITx, CRITxFL is created with the condition: a <= crit_var <= b. If criterion flags of that type are not required, then leave as NULL.

bcva_uplims

List containing one or more numeric elements. For each element a in bcva_uplims, a pair of variables CRITx, CRITxFL is created with the condition: crit_var <= a. If criterion flags of that type are not required, then leave as NULL.

bcva_lowlims

List containing one or more numeric elements. For each element b in bcva_lowlims, a pair of variables CRITx, CRITxFL is created with the condition: crit_var >= b. If criterion flags of that type are not required, then leave as NULL.

additional_text

string containing additional text to append to CRITx

critxfl_index

positive integer detailing the first value of x to use in CRITxFL. If not supplied, the function takes the first available value of x, counting up from x = 1.

Author

Edoardo Mancini

Details

This function works by calling derive_var_bcvacritxfl() once for each of the elements in bcva_ranges, bcva_uplims and bcva_lowlims. NOTE: if crit_var is equal to NA, then the resulting criterion flag is also marked as NA.

Examples

library(tibble)
library(admiral)
library(admiraldev)

adbcva1 <- tribble(
  ~STUDYID, ~USUBJID, ~AVISIT, ~BASETYPE, ~PARAMCD, ~CHG,
  "XXX001", "P01", "BASELINE", "LAST", "SBCVA", 0,
  "XXX001", "P01", "WEEK 2", "LAST", "FBCVA", 2,
  "XXX001", "P02", "BASELINE", "LAST", "SBCVA", -13,
  "XXX001", "P02", "WEEK 2", "LAST", "FBCVA", 5,
  "XXX001", "P03", "BASELINE", "LAST", "SBCVA", NA,
  "XXX001", "P03", "WEEK 2", "LAST", "FBCVA", 17
)

derive_var_bcvacritxfl(
  dataset = adbcva1,
  crit_var = exprs(CHG),
  bcva_ranges = list(c(0, 5), c(-5, -1), c(10, 15)),
  bcva_uplims = list(5, 10),
  bcva_lowlims = list(8),
  additional_text = ""
)

adbcva2 <- tribble(
  ~STUDYID, ~USUBJID, ~AVISIT, ~BASETYPE, ~PARAMCD, ~AVAL, ~CHG,
  "XXX001", "P01", "BASELINE", "LAST", "SBCVA", 4, NA,
  "XXX001", "P01", "BASELINE", "LAST", "SBCVA", 6, NA,
  "XXX001", "P01", "AVERAGE BASELINE", "AVERAGE", "SBCVA", 5, NA,
  "XXX001", "P01", "WEEK 2", "LAST", "SBCVA", -3, NA,
  "XXX001", "P01", "WEEK 4", "LAST", "SBCVA", -10, NA,
  "XXX001", "P01", "WEEK 6", "LAST", "SBCVA", 12, NA,
  "XXX001", "P01", "WEEK 2", "AVERAGE", "SBCVA", -2, -7,
  "XXX001", "P01", "WEEK 4", "AVERAGE", "SBCVA", 6, 1,
  "XXX001", "P01", "WEEK 6", "AVERAGE", "SBCVA", 3, -2
)

restrict_derivation(
  adbcva2,
  derivation = derive_var_bcvacritxfl,
  args = params(
    crit_var = exprs(CHG),
    bcva_ranges = list(c(0, 5), c(-10, 0)),
    bcva_lowlims = list(5),
    additional_text = " (AVERAGE)"
  ),
  filter = PARAMCD %in% c("SBCVA", "FBCVA") & BASETYPE == "AVERAGE"
)