CRITy, CRITyFL, and CRITyFNThe function derives ADaM compliant criterion flags, e.g., to facilitate subgroup analyses.
If a criterion flag can't be derived with this function, the derivation is not ADaM compliant. It helps to ensure that:
the condition of the criterion depends only on variables of the same row,
the CRITyFL is populated with valid values, i.e, either "Y" and NA or
"Y", "N", and NA,
the CRITy variable is populated correctly, i.e.,
set to a constant value within a parameter if CRITyFL is populated with
"Y", "N", and NA and
set to a constant value within a parameter if the criterion condition is
fulfilled and to NA otherwise if CRITyFL is populated with "Y", and
NA
derive_vars_crit_flag(
dataset,
crit_nr = 1,
condition,
description,
values_yn = FALSE,
create_numeric_flag = FALSE
)The input dataset with the variables CRITy, CRITyFL, and
optionally CRITyFN added.
Input dataset
a dataset, i.e., a data.frame or tibble
none
The criterion number, i.e., the y in CRITy
a positive integer, e.g. 2 or 5
1
Condition for flagging records
See description of the values_yn argument for details on how the
CRITyFL variable is populated.
an unquoted condition, e.g., AVISIT == "BASELINE"
none
The description of the criterion
The CRITy variable is set to the specified value.
An expression can be specified to set the value depending on the parameter. Please note that the value must be constant within a parameter.
an unquoted expression which evaluates to a character
(in dataset)
none
Should "Y" and "N" be used for CRITyFL?
If set to TRUE, the CRITyFL variable is set to "Y" if the condition
(condition) evaluates to TRUE, it is set to "N" if the condition
evaluate to FALSE, and to NA if it evaluates to NA.
Otherwise, the CRITyFL variable is set to "Y" if the condition
(condition) evaluates to TRUE, and to NA otherwise.
"TRUE", "FALSE"
FALSE
Create a numeric flag?
If set to TRUE, the CRITyFN variable is created. It is set to 1 if
CRITyFL == "Y", it set to 0 if CRITyFL == "N", and to NA otherwise.
"TRUE", "FALSE"
FALSE
The following examples use the BDS dataset below as a basis.
library(tibble, warn.conflicts = FALSE)adbds <- tribble(
~PARAMCD, ~AVAL,
"AST", 42,
"AST", 52,
"AST", NA_real_,
"ALT", 33,
"ALT", 51
)
"Y" and NA
(condition, description)
The following call is a simple application of derive_vars_crit_flag()
to derive a criterion flag/variable pair in a BDS dataset.
The new variables are named CRIT1/CRIT1FL because the argument
crit_nr has not been passed.
Since the argument values_yn has also not been passed and thus is
set to its default of FALSE, CRIT1FL is set to Y only if
condition evaluates to TRUE. For example, in both the
first and third records, where condition is respectively FALSE
and NA, we set CRIT1FL = NA_character_. The fourth record also
exhibits this behavior. Also, as per CDISC standards, in this case
CRIT1 is populated only for records where condition evaluates
to TRUE.
derive_vars_crit_flag(
adbds,
condition = AVAL > 50,
description = "Absolute value > 50"
)
#> # A tibble: 5 × 4
#> PARAMCD AVAL CRIT1FL CRIT1
#> <chr> <dbl> <chr> <chr>
#> 1 AST 42 <NA> <NA>
#> 2 AST 52 Y Absolute value > 50
#> 3 AST NA <NA> <NA>
#> 4 ALT 33 <NA> <NA>
#> 5 ALT 51 Y Absolute value > 50The description argument also accepts expressions which depend
on other variables in the input dataset. This can be useful to
dynamically populate CRITx, for instance in the case below where
we improve the CRIT1 text because the same flag/variable pair is
actually being used for multiple parameters.
derive_vars_crit_flag(
adbds,
condition = AVAL > 50,
description = paste(PARAMCD, "> 50"),
)
#> # A tibble: 5 × 4
#> PARAMCD AVAL CRIT1FL CRIT1
#> <chr> <dbl> <chr> <chr>
#> 1 AST 42 <NA> <NA>
#> 2 AST 52 Y AST > 50
#> 3 AST NA <NA> <NA>
#> 4 ALT 33 <NA> <NA>
#> 5 ALT 51 Y ALT > 50"Y", "N" and NA
(values_yn)
The next call builds on the previous example by using
value_yn = TRUE to distinguish between the cases
where condition is FALSE and those where it is
not evaluable at all.
As compared to the previous example, for the first record condition
evaluates to FALSE and so we set CRIT1FL = "N", whereas for the
third record, condition evaluates to NA because AVAL is
missing and so we set CRIT1FL to NA.
Note also that because we are using the values "Y", "N" and NA
for the flag, as per CDISC standards CRIT1 is now
populated for all records rather than just for the "Y" records.
derive_vars_crit_flag(
adbds,
condition = AVAL > 50,
description = paste(PARAMCD, "> 50"),
values_yn = TRUE
)
#> # A tibble: 5 × 4
#> PARAMCD AVAL CRIT1FL CRIT1
#> <chr> <dbl> <chr> <chr>
#> 1 AST 42 N AST > 50
#> 2 AST 52 Y AST > 50
#> 3 AST NA <NA> AST > 50
#> 4 ALT 33 N ALT > 50
#> 5 ALT 51 Y ALT > 50If the user wishes to set the criterion flag to "N" whenever
the condition is not fulfilled, condition can be updated using
an if_else call, where the third argument determines the behavior
when the condition is not evaluable.
derive_vars_crit_flag(
adbds,
condition = if_else(AVAL > 50, TRUE, FALSE, FALSE),
description = paste(PARAMCD, "> 50"),
values_yn = TRUE
)
#> # A tibble: 5 × 4
#> PARAMCD AVAL CRIT1FL CRIT1
#> <chr> <dbl> <chr> <chr>
#> 1 AST 42 N AST > 50
#> 2 AST 52 Y AST > 50
#> 3 AST NA N AST > 50
#> 4 ALT 33 N ALT > 50
#> 5 ALT 51 Y ALT > 50crit_nr, create_numeric_flag).
The user can manually specify the criterion variable/flag number
to use to name CRITy/CRITyFL by passing the crit_nr argument. This
may be necessary if, for instance, other criterion flags already exist
in the input dataset.
The user can also choose to create an additional, equivalent numeric
flag CRITyFN by setting create_numeric_flag to TRUE.
derive_vars_crit_flag(
adbds,
condition = AVAL > 50,
description = paste(PARAMCD, "> 50"),
values_yn = TRUE,
crit_nr = 2,
create_numeric_flag = TRUE
)
#> # A tibble: 5 × 5
#> PARAMCD AVAL CRIT2FL CRIT2 CRIT2FN
#> <chr> <dbl> <chr> <chr> <int>
#> 1 AST 42 N AST > 50 0
#> 2 AST 52 Y AST > 50 1
#> 3 AST NA <NA> AST > 50 NA
#> 4 ALT 33 N ALT > 50 0
#> 5 ALT 51 Y ALT > 50 1BDS-Findings Functions that returns variable appended to dataset:
derive_basetype_records(),
derive_var_analysis_ratio(),
derive_var_anrind(),
derive_var_atoxgr(),
derive_var_atoxgr_dir(),
derive_var_base(),
derive_var_chg(),
derive_var_ontrtfl(),
derive_var_pchg(),
derive_var_shift()