Skip to contents

Creating ADFACE

Source: vignettes/adface.Rmd
adface.Rmd

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

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
  )