derive_param_computed: Adds a Parameter Computed from the Analysis Value of Other...
In admiral: ADaM in R Asset Library

derive_param_computed

R Documentation

Adds a Parameter Computed from the Analysis Value of Other Parameters

Description

Adds a parameter computed from the analysis value of other parameters. It is expected that the analysis value of the new parameter is defined by an expression using the analysis values of other parameters. For example mean arterial pressure (MAP) can be derived from systolic (SYSBP) and diastolic blood pressure (DIABP) with the formula

MAP = \frac{SYSBP + 2 DIABP}{3}

Usage

derive_param_computed(
  dataset = NULL,
  dataset_add = NULL,
  by_vars,
  parameters,
  set_values_to,
  filter = NULL,
  constant_by_vars = NULL,
  constant_parameters = NULL,
  keep_nas = FALSE
)

Arguments

`dataset`	Input dataset The variables specified by the `by_vars` argument are expected to be in the dataset. `PARAMCD` is expected as well. The variable specified by `by_vars` and `PARAMCD` must be a unique key of the input dataset after restricting it by the filter condition (`filter` parameter) and to the parameters specified by `parameters`.
`dataset_add`	Additional dataset The variables specified by the `by_vars` parameter are expected. The variable specified by `by_vars` and `PARAMCD` must be a unique key of the additional dataset after restricting it to the parameters specified by `parameters`. If the argument is specified, the observations of the additional dataset are considered in addition to the observations from the input dataset (`dataset` restricted by `filter`).
`by_vars`	Grouping variables For each group defined by `by_vars` an observation is added to the output dataset. Only variables specified in `by_vars` will be populated in the newly created records. Permitted Values: list of variables created by `exprs()` e.g. `exprs(USUBJID, VISIT)`
`parameters`	Required parameter codes It is expected that all parameter codes (`PARAMCD`) which are required to derive the new parameter are specified for this parameter or the `constant_parameters` parameter. If observations should be considered which do not have a parameter code, e.g., if an SDTM dataset is used, temporary parameter codes can be derived by specifying a list of expressions. The name of the element defines the temporary parameter code and the expression the condition for selecting the records. For example `parameters = exprs(HGHT = VSTESTCD == "HEIGHT")` selects the observations with `VSTESTCD == "HEIGHT"` from the input data (`dataset` and `dataset_add`), sets `PARAMCD = "HGHT"` for these observations, and adds them to the observations to consider. Unnamed elements in the list of expressions are considered as parameter codes. For example, `parameters = exprs(WEIGHT, HGHT = VSTESTCD == "HEIGHT")` uses the parameter code `"WEIGHT"` and creates a temporary parameter code `"HGHT"`. Permitted Values: A character vector of `PARAMCD` values or a list of expressions
`set_values_to`	Variables to be set The specified variables are set to the specified values for the new observations. The values of variables of the parameters specified by `parameters` can be accessed using `⁠<variable name>.<parameter code>⁠`. For example exprs( AVAL = (AVAL.SYSBP + 2 * AVAL.DIABP) / 3, PARAMCD = "MAP" ) defines the analysis value and parameter code for the new parameter. Variable names in the expression must not contain more than one dot. Permitted Values: List of variable-value pairs
`filter`	Filter condition The specified condition is applied to the input dataset before deriving the new parameter, i.e., only observations fulfilling the condition are taken into account. Permitted Values: a condition
`constant_by_vars`	By variables for constant parameters The constant parameters (parameters that are measured only once) are merged to the other parameters using the specified variables. (Refer to Example 2) Permitted Values: list of variables created by `exprs()` e.g. `exprs(USUBJID, VISIT)`
`constant_parameters`	Required constant parameter codes It is expected that all the parameter codes (`PARAMCD`) which are required to derive the new parameter and are measured only once are specified here. For example if BMI should be derived and height is measured only once while weight is measured at each visit. Height could be specified in the `constant_parameters` parameter. (Refer to Example 2) If observations should be considered which do not have a parameter code, e.g., if an SDTM dataset is used, temporary parameter codes can be derived by specifying a list of expressions. The name of the element defines the temporary parameter code and the expression the condition for selecting the records. For example `constant_parameters = exprs(HGHT = VSTESTCD == "HEIGHT")` selects the observations with `VSTESTCD == "HEIGHT"` from the input data (`dataset` and `dataset_add`), sets `PARAMCD = "HGHT"` for these observations, and adds them to the observations to consider. Unnamed elements in the list of expressions are considered as parameter codes. For example, `constant_parameters = exprs(WEIGHT, HGHT = VSTESTCD == "HEIGHT")` uses the parameter code `"WEIGHT"` and creates a temporary parameter code `"HGHT"`. Permitted Values: A character vector of `PARAMCD` values or a list of expressions
`keep_nas`	Keep observations with `NA`s If the argument is set to `TRUE`, observations are added even if some of the values contributing to the computed value are `NA`.

Details

For each group (with respect to the variables specified for the by_vars parameter) an observation is added to the output dataset if the filtered input dataset (dataset) or the additional dataset (dataset_add) contains exactly one observation for each parameter code specified for parameters.

For the new observations the variables specified for set_values_to are set to the provided values. The values of the other variables of the input dataset are set to NA.

Value

The input dataset with the new parameter added. Note, a variable will only be populated in the new parameter rows if it is specified in by_vars.

Examples

library(tibble)
library(dplyr)
library(lubridate)

# Example 1a: Derive MAP
advs <- tribble(
  ~USUBJID, ~PARAMCD, ~PARAM, ~AVAL, ~AVALU, ~VISIT,
  "01-701-1015", "DIABP", "Diastolic Blood Pressure (mmHg)", 51, "mmHg", "BASELINE",
  "01-701-1015", "DIABP", "Diastolic Blood Pressure (mmHg)", 50, "mmHg", "WEEK 2",
  "01-701-1015", "SYSBP", "Systolic Blood Pressure (mmHg)", 121, "mmHg", "BASELINE",
  "01-701-1015", "SYSBP", "Systolic Blood Pressure (mmHg)", 121, "mmHg", "WEEK 2",
  "01-701-1028", "DIABP", "Diastolic Blood Pressure (mmHg)", 79, "mmHg", "BASELINE",
  "01-701-1028", "DIABP", "Diastolic Blood Pressure (mmHg)", 80, "mmHg", "WEEK 2",
  "01-701-1028", "SYSBP", "Systolic Blood Pressure (mmHg)", 130, "mmHg", "BASELINE",
  "01-701-1028", "SYSBP", "Systolic Blood Pressure (mmHg)", 132, "mmHg", "WEEK 2"
) %>%
  mutate(
    ADT = case_when(
      VISIT == "BASELINE" ~ as.Date("2024-01-10"),
      VISIT == "WEEK 2" ~ as.Date("2024-01-24")
    ),
    ADTF = NA_character_
  )

derive_param_computed(
  advs,
  by_vars = exprs(USUBJID, VISIT),
  parameters = c("SYSBP", "DIABP"),
  set_values_to = exprs(
    AVAL = (AVAL.SYSBP + 2 * AVAL.DIABP) / 3,
    PARAMCD = "MAP",
    PARAM = "Mean Arterial Pressure (mmHg)",
    AVALU = "mmHg",
    ADT = ADT.SYSBP
  )
)

# Example 1b: Using option `keep_nas = TRUE` to derive MAP in the case where some/all values
# of a variable used in the computation are missing

derive_param_computed(
  advs,
  by_vars = exprs(USUBJID, VISIT),
  parameters = c("SYSBP", "DIABP"),
  set_values_to = exprs(
    AVAL = (AVAL.SYSBP + 2 * AVAL.DIABP) / 3,
    PARAMCD = "MAP",
    PARAM = "Mean Arterial Pressure (mmHg)",
    AVALU = "mmHg",
    ADT = ADT.SYSBP,
    ADTF = ADTF.SYSBP
  ),
  keep_nas = TRUE
)

# Example 2: Derive BMI where height is measured only once
advs <- tribble(
  ~USUBJID,      ~PARAMCD, ~PARAM,        ~AVAL, ~AVALU, ~VISIT,
  "01-701-1015", "HEIGHT", "Height (cm)", 147.0, "cm",   "SCREENING",
  "01-701-1015", "WEIGHT", "Weight (kg)",  54.0, "kg",   "SCREENING",
  "01-701-1015", "WEIGHT", "Weight (kg)",  54.4, "kg",   "BASELINE",
  "01-701-1015", "WEIGHT", "Weight (kg)",  53.1, "kg",   "WEEK 2",
  "01-701-1028", "HEIGHT", "Height (cm)", 163.0, "cm",   "SCREENING",
  "01-701-1028", "WEIGHT", "Weight (kg)",  78.5, "kg",   "SCREENING",
  "01-701-1028", "WEIGHT", "Weight (kg)",  80.3, "kg",   "BASELINE",
  "01-701-1028", "WEIGHT", "Weight (kg)",  80.7, "kg",   "WEEK 2"
)

derive_param_computed(
  advs,
  by_vars = exprs(USUBJID, VISIT),
  parameters = "WEIGHT",
  set_values_to = exprs(
    AVAL = AVAL.WEIGHT / (AVAL.HEIGHT / 100)^2,
    PARAMCD = "BMI",
    PARAM = "Body Mass Index (kg/m^2)",
    AVALU = "kg/m^2"
  ),
  constant_parameters = c("HEIGHT"),
  constant_by_vars = exprs(USUBJID)
)

# Example 3: Using data from an additional dataset and other variables than AVAL
qs <- tribble(
  ~USUBJID, ~AVISIT,   ~QSTESTCD, ~QSORRES, ~QSSTRESN,
  "1",      "WEEK 2",  "CHSF112", NA,               1,
  "1",      "WEEK 2",  "CHSF113", "Yes",           NA,
  "1",      "WEEK 2",  "CHSF114", NA,               1,
  "1",      "WEEK 4",  "CHSF112", NA,               2,
  "1",      "WEEK 4",  "CHSF113", "No",            NA,
  "1",      "WEEK 4",  "CHSF114", NA,               1
)

adchsf <- tribble(
  ~USUBJID, ~AVISIT,  ~PARAMCD, ~QSSTRESN, ~AVAL,
  "1",      "WEEK 2", "CHSF12", 1,             6,
  "1",      "WEEK 2", "CHSF14", 1,             6,
  "1",      "WEEK 4", "CHSF12", 2,            12,
  "1",      "WEEK 4", "CHSF14", 1,             6
) %>%
  mutate(QSORRES = NA_character_)

derive_param_computed(
  adchsf,
  dataset_add = qs,
  by_vars = exprs(USUBJID, AVISIT),
  parameters = exprs(CHSF12, CHSF13 = QSTESTCD %in% c("CHSF113", "CHSF213"), CHSF14),
  set_values_to = exprs(
    AVAL = case_when(
      QSORRES.CHSF13 == "Not applicable" ~ 0,
      QSORRES.CHSF13 == "Yes" ~ 38,
      QSORRES.CHSF13 == "No" ~ if_else(
        QSSTRESN.CHSF12 > QSSTRESN.CHSF14,
        25,
        0
      )
    ),
    PARAMCD = "CHSF13"
  )
)

# Example 4: Computing more than one variable
adlb_tbilialk <- tribble(
  ~USUBJID, ~PARAMCD, ~AVALC, ~ADTM,        ~ADTF,
  "1",      "ALK2",   "Y",    "2021-05-13", NA_character_,
  "1",      "TBILI2", "Y",    "2021-06-30", "D",
  "2",      "ALK2",   "Y",    "2021-12-31", "M",
  "2",      "TBILI2", "N",    "2021-11-11", NA_character_,
  "3",      "ALK2",   "N",    "2021-04-03", NA_character_,
  "3",      "TBILI2", "N",    "2021-04-04", NA_character_
) %>%
  mutate(ADTM = ymd(ADTM))

derive_param_computed(
  dataset_add = adlb_tbilialk,
  by_vars = exprs(USUBJID),
  parameters = c("ALK2", "TBILI2"),
  set_values_to = exprs(
    AVALC = if_else(AVALC.TBILI2 == "Y" & AVALC.ALK2 == "Y", "Y", "N"),
    ADTM = pmax(ADTM.TBILI2, ADTM.ALK2),
    ADTF = if_else(ADTM == ADTM.TBILI2, ADTF.TBILI2, ADTF.ALK2),
    PARAMCD = "TB2AK2",
    PARAM = "TBILI > 2 times ULN and ALKPH <= 2 times ULN"
  ),
  keep_nas = TRUE
)

admiral documentation built on June 22, 2024, 9:06 a.m.