Creating ADFACE

Introduction

This article describes about creating ADFACE ADaM dataset which is part of Vaccine - Reactogenicity based on the Center for Biologics Evaluation and Research (CBER) guidelines.

See the below links for more information:

Center for Biologics Evaluation and Research (CBER) Guidelines

Therapeutic Area Data Standards User Guide for Vaccines (TAUG-Vax)

Examples are currently tested using ADSL (ADaM) and face, vs, ex (SDTM) inputs.

Note: All examples assume CDISC SDTM and/or ADaM format as input unless otherwise specified.

Programming Workflow

• Read in Data
• Pre-processing of Input Dataset
• Merge FACE with EX
• Merge Required ADSL Variables Needed for Analysis
• Derive Fever Records from VS Domain
• Derive/Impute Numeric Date/Time and Analysis Day (ADT, ADTM, ADY, ADTF, ATMF)
• Derive Period Variables (e.g. APxxSDT, APxxEDT, …)
• Derive Direct Mapping Variables
• Derive Severity Records for Administration Site Events
• Derive Maximum Records
• Assign PARAMCD, PARAM, PARAMN, PARCAT1
• Derive Maximum Severity Flag
• Derive Event Occurrence Flag
• Post-processing of the Dataset
• Add ADSL Variables

Read in Data

To start, all data frames needed for the creation of ADFACE should be read into the environment. Some of the data frames needed are VS,EX and FACE.

library(admiral)
library(admiralvaccine)
library(admiraldev)
library(pharmaversesdtm)
library(dplyr, warn.conflicts = FALSE)
library(lubridate)
library(stringr)
library(tidyr)
library(tibble)

data("face_vaccine")
data("suppface_vaccine")
data("ex_vaccine")
data("suppex_vaccine")
data("vs_vaccine")
data("admiralvaccine_adsl")

face <- convert_blanks_to_na(face_vaccine)
ex <- convert_blanks_to_na(ex_vaccine)
vs <- convert_blanks_to_na(vs_vaccine)
suppface <- convert_blanks_to_na(suppface_vaccine)
suppex <- convert_blanks_to_na(suppex_vaccine)
adsl <- convert_blanks_to_na(admiralvaccine_adsl)

Pre-processing of Input Dataset

This step involves company-specific pre-processing of required input dataset for further analysis. In this step, we will filter records that has only reactogenicity events and combine the face and ex with their supplementary datasets suppface and suppex respectively.

face <- face %>%
  filter(FACAT == "REACTOGENICITY" & grepl("ADMIN|SYS", FASCAT)) %>%
  mutate(FAOBJ = str_to_upper(FAOBJ)) %>%
  metatools::combine_supp(suppface)
ex <- metatools::combine_supp(ex, suppex)

Merge FACE with EX

In this step, we will merge face with ex domain and add required variables from ex domain to the input dataset. If subjects have multiple vaccination at same visit then this function will not merge input dataset with ex dataset and throws a warning.

The function derive_vars_merged_vaccine() is used to merge face with ex domain.

adface <- derive_vars_merged_vaccine(
  dataset = face,
  dataset_ex = ex,
  by_vars_sys = exprs(USUBJID, FATPTREF = EXLNKGRP),
  by_vars_adms = exprs(USUBJID, FATPTREF = EXLNKGRP, FALOC = EXLOC, FALAT = EXLAT),
  ex_vars = exprs(EXTRT, EXDOSE, EXSEQ, EXSTDTC, EXENDTC, VISIT, VISITNUM)
)

This call would return the input dataset with columns from ex_vars added if the subjects does not have multiple vaccination at same visit.

Though the function will throw warning if subjects have multiple vaccination at same visit, this call would return the input dataset merging it with supplementary dataset.

Merge Required ADSL Variables Needed for Analysis

At this step, it may be useful to join ADSL to your face domain. Only the ADSL variables used for derivations are selected at this step. The rest of the relevant ADSL variables would be added later.

adsl_vars <- exprs(RFSTDTC, RFENDTC)

adface <- derive_vars_merged(
  face,
  dataset_add = adsl,
  new_vars = adsl_vars,
  by_vars = get_admiral_option("subject_keys")
)

This call would return the input dataset with columns RFSTDTC, RFENDTC added.

Derive Fever Records from VS Domain

In this step, we will merge fever records from the VS domain with the input dataset if the fever records does not present in the input dataset.

The function derive_fever_records() is used to merge fever records. These records will also be used in maximum temperature calculation.

adface <- derive_fever_records(
  dataset = adface,
  dataset_source = ungroup(vs),
  filter_source = VSCAT == "REACTOGENICITY" & VSTESTCD == "TEMP",
  faobj = "FEVER"
)

This call returns the input dataset with FEVER records added if the input dataset does not have FEVER records. If the input dataset has FEVER records, the output dataset will be same as the input dataset.

Derive/Impute Numeric Date/Time and Analysis Day (ADT, ADTM, ADTF, ATMF, ADY)

The function derive_vars_dt() can be used to derive ADT. This function allows the user to impute the date as well.

Similarly, ADTM can be created using the function derive_vars_dtm(). Imputation can be done on both the date and time components of ADTM.

Example calls:

adface <- adface %>%
  derive_vars_dt(
    new_vars_prefix = "A",
    dtc = FADTC
  ) %>%
  derive_vars_dtm(
    new_vars_prefix = "A",
    dtc = FADTC,
    highest_imputation = "n"
  )

Once ADT is derived, the function derive_vars_dy() can be used to derive ADY. This example assumes both ADT and RFSTDTC exist on the data frame.

adface <- adface %>%
  mutate(RFSTDTC = as.Date(RFSTDTC)) %>%
  derive_vars_dy(reference_date = RFSTDTC, source_vars = exprs(ADT))

Derive Period Variables (e.g. APxxSDT, APxxEDT, …)

The admiral core package has separate functions to handle period variables since these variables are study specific.

See the “Visit and Period Variables” vignette for more information.

If the variables are not derived based on a period reference dataset, they may be derived at a later point of the flow. For example, phases like “Treatment Phase” and “Follow up” could be derived based on treatment start and end date.

period_ref <- create_period_dataset(
  dataset = adsl,
  new_vars = exprs(
    APERSDT = APxxSDT, APEREDT = APxxEDT, TRTA = TRTxxA,
    TRTP = TRTxxP
  )
)

adface <- derive_vars_joined(
  adface,
  dataset_add = period_ref,
  by_vars = get_admiral_option("subject_keys"),
  filter_join = ADT >= APERSDT & ADT <= APEREDT,
  join_type = "all"
)

Derive Direct Mapping Variables

In this step,we will create the user defined function to assign AVAL values from AVALC which will be used in further steps.

The user defined functions would look like the following:

sev_to_numeric <- function(x, y) {
  case_when(
    x == "NONE" ~ 0,
    x == "MILD" ~ 1,
    x == "MODERATE" ~ 2,
    x == "SEVERE" ~ 3,
    TRUE ~ y
  )
}

The mapping of these variables is left to the User. An example mapping may be:

adface <- adface %>%
  mutate(
    AVALC = as.character(FASTRESC),
    AVAL = suppressWarnings(as.numeric(FASTRESN)),
    AVAL = sev_to_numeric(AVALC, AVAL),
    ATPTREF = FATPTREF,
    ATPT = FATPT,
    ATPTN = FATPTNUM
  )

Derive AN01FL

Creating ANL01FL which would flag the records that will be considered for analysis and if there is any Investigator and Subject record for the same day, it would flag the Investigator record over the subject record.

Note: Please, consider which assessment is needed for your analysis. If you want to prioritize Investigator assessment, please proceed as follows. Otherwise, change FAEVAL order.

adface <- adface %>% derive_var_extreme_flag(
  by = exprs(STUDYID, USUBJID, FATPTREF, FAOBJ, FATESTCD, FATPTNUM),
  order = exprs(STUDYID, USUBJID, FATPTREF, FAOBJ, FATESTCD, FATPTNUM, FAEVAL),
  new_var = ANL01FL,
  mode = "first",
  true_value = "Y",
  false_value = NA_character_
)

Derive Severity Records for Administration Site Events

The function derive_diam_to_sev_records() is used to derive the severity records from the diameter records for an event.

The severity records created will be useful for calculating the maximum severity.

adface <- derive_diam_to_sev_records(
  dataset = adface,
  filter_add = ANL01FL == "Y",
  diam_code = "DIAMETER",
  faobj_values = c("REDNESS", "SWELLING"),
  testcd_sev = "SEV",
  test_sev = "Severity/Intensity",
  none = 0,
  mild = 2,
  mod = 5,
  sev = 10
)

This call returns the input dataset with severity records derived from the diameter records for an event.

By default, we will populate SEV and Severity/Intensity as the FATESTCD and FATEST for the newly added records. The function allows the user to change the FATESTCD and FATEST as well.

Derive Maximum Records

In this step, we will derive maximum records for severity, diameter, temperature using the function derive_extreme_records().

adface <- derive_extreme_records(
  dataset = adface,
  dataset_add = adface,
  filter_add = FATESTCD == "SEV" & ANL01FL == "Y",
  by_vars = exprs(USUBJID, FAOBJ, ATPTREF),
  order = exprs(AVAL),
  check_type = "none",
  mode = "last",
  set_values_to = exprs(
    FATEST = "Maximum Severity",
    FATESTCD = "MAXSEV"
  )
)

adface <- derive_extreme_records(
  dataset = adface,
  dataset_add = adface,
  filter_add = FAOBJ %in% c("REDNESS", "SWELLING") & FATESTCD == "DIAMETER" & ANL01FL == "Y",
  by_vars = exprs(USUBJID, FAOBJ, FALNKGRP),
  order = exprs(AVAL),
  check_type = "none",
  mode = "last",
  set_values_to = exprs(
    FATEST = "Maximum Diameter",
    FATESTCD = "MAXDIAM"
  )
)

adface <- derive_extreme_records(
  dataset = adface,
  dataset_add = adface,
  filter_add = FAOBJ == "FEVER" & ANL01FL == "Y",
  by_vars = exprs(USUBJID, FAOBJ, ATPTREF),
  order = exprs(VSSTRESN),
  check_type = "none",
  mode = "last",
  set_values_to = exprs(
    FATEST = "Maximum Temperature",
    FATESTCD = "MAXTEMP"
  )
)

This call returns the input dataset with maximum records added for the severity, diameter, temperature.

Assign PARAMCD, PARAM, PARAMN, PARCAT1

To assign parameter level values such as PARAMCD, PARAM, PARAMN, etc., a lookup needs to be created to join to the source data. PARCAT1, PARCAT2 variables are assigned from FACAT, FASCAT variables.

For example, when creating ADFACE dataset, a lookup based on the SDTM --TESTCD value can be created:

FATESTCD	PARAMCD	PARAMN	FATEST
SEV	SEVREDN	1	Severity
DIAMETER	DIARE	2	Diameter
MAXDIAM	MDIRE	3	Maximum Diameter cm
MAXTEMP	MAXTEMP	4	Maximum Temperature
MAXSEV	MAXSWEL	5	Maximum Severity
OCCUR	OCFEVER	6	Occurrence Indicator
OCCUR	OCERYTH	7	Occurrence Indicator
SEV	SEVPAIN	8	Severity
OCCUR	OCPAIN	9	Occurrence Indicator
OCCUR	OCSWEL	10	Occurrence Indicator

This lookup can now be joined to the source data:

adface <- derive_vars_params(
  dataset = adface,
  lookup_dataset = lookup_dataset,
  merge_vars = exprs(PARAMCD, PARAMN)
)

PARAMCD will be always derived from lookup dataset whereas PARAMN, PARAM, PARCAT1, PARCAT2 can be either derived from lookup dataset if mentioned in merge_vars argument or derived in the function.

Derive Maximum Severity Flag

The function derive_vars_max_flag() is used to derive flag variable for the maximum values of an event.

flag1 - Flags the maximum value per subject per event per Vaccination. flag2 - Flags the maximum value per subject per event for Overall.

adface <- derive_vars_max_flag(
  dataset = adface,
  flag1 = "ANL02FL",
  flag2 = "ANL03FL"
)
#> Joining with `by = join_by(STUDYID, DOMAIN, USUBJID, FASEQ, FALNKGRP, PARAM,
#> FALAT, FALNKID, FALOC, FATESTCD, FATEST, FAOBJ, FACAT, FASCAT, FAEVAL, FAORRES,
#> FAORRESU, EPOCH, FADTC, FADY, FATPT, FATPTNUM, FATPTREF, FARFTDTC, FAEVLINT,
#> FAEVINTX, FASTAT, FAREASND, FASTRESC, FASTRESN, FASTRESU, IDVARVAL, CLTYP,
#> RFSTDTC, RFENDTC, VSSTRESN, ADT, ADTM, ADY, APERIOD, APERSDT, APEREDT, TRTA,
#> TRTP, AVALC, AVAL, ATPTREF, ATPT, ATPTN, ANL01FL, PARAMCD, PARAMN, PARCAT1,
#> PARCAT2)`
#> Joining with `by = join_by(STUDYID, DOMAIN, USUBJID, FASEQ, FALNKGRP, PARAM,
#> FALAT, FALNKID, FALOC, FATESTCD, FATEST, FAOBJ, FACAT, FASCAT, FAEVAL, FAORRES,
#> FAORRESU, EPOCH, FADTC, FADY, FATPT, FATPTNUM, FATPTREF, FARFTDTC, FAEVLINT,
#> FAEVINTX, FASTAT, FAREASND, FASTRESC, FASTRESN, FASTRESU, IDVARVAL, CLTYP,
#> RFSTDTC, RFENDTC, VSSTRESN, ADT, ADTM, ADY, APERIOD, APERSDT, APEREDT, TRTA,
#> TRTP, AVALC, AVAL, ATPTREF, ATPT, ATPTN, ANL01FL, PARAMCD, PARAMN, PARCAT1,
#> PARCAT2, ANL02FL)`

This call would return the input dataset with columns ANL02FL, ANL03FL added by default. This function allows the user to change the name of the new variables created.

Derive Event Occurrence Flag

The function derive_vars_event_flag() is used to derive flag variable for the events that occurred.

new_var1 - Flags the record if at least one of the event occurred within the observation period. new_var2 - Flags the record if the event is occurred.

adface <- derive_vars_event_flag(
  dataset = adface,
  by_vars = exprs(USUBJID, FAOBJ, ATPTREF),
  aval_cutoff = 2.5,
  new_var1 = EVENTFL,
  new_var2 = EVENTDFL
)

This call would return the input dataset with columns EVENTFL, EVENTDFL added by default. This function allows the user to change the name of the new variables created as well.

Post-processing of the Dataset

In this step, we will remove values for all the derived records in SDTM variables.

adface <- post_process_reacto(
  dataset = adface,
  filter_dataset = FATESTCD %in% c("MAXDIAM", "MAXSEV", "MAXTEMP") |
    (FATESTCD %in% c("OCCUR", "SEV") & FAOBJ %in% c("FEVER", "REDNESS", "SWELLING"))
)

Add ADSL variables

If needed, the other ADSL variables can now be added. List of ADSL variables already merged held in vector adsl_vars

adsl <- adsl %>%
  convert_blanks_to_na() %>%
  filter(!is.na(USUBJID))

adface <- derive_vars_merged(
  dataset = adface,
  dataset_add = select(adsl, !!!negate_vars(adsl_vars)),
  by_vars = get_admiral_option("subject_keys")
)

Example Script

ADaM	Sample Code
ADFACE	ad_adface.R