pharmaverse champion R/pharma pharmaverse Friday Workshop

Ari Siggaard Knoph
Ross Farrugia

Intro

  • 2 pharmaverse council reps from Novo Nordisk & Roche
  • Apologies we are not Kangjie Zhang from Bayer (as advertised)

Agenda

  • pharmaverse overview
  • Select packages walkthrough and examples
  • admiral exercises
  • Closing thoughts

pharmaverse overview


pharmaverse overview

Help at your fingertips!

install.packages("pharmaverse")

pharmaverse overview

Explore!

pharmaverse::explore()

pharmaverse overview

Join communities!

pharmaverse::join("slack")
pharmaverse::join("linkedin")
pharmaverse::join("github")

pharmaverse overview

See pharmaverse examples!

pharmaverse::examples()

Packages Walkthrough - intro

  • metacore
  • metatools
  • xportr
  • admiral

Packages Walkthrough - example usage

admiral exercises

library(admiral)
library(dplyr, warn.conflicts = FALSE)
library(pharmaversesdtm)
library(lubridate)
library(stringr)

data("dm")
data("ds")
data("ex")
data("ae")

dm <- convert_blanks_to_na(dm)
ds <- convert_blanks_to_na(ds)
ex <- convert_blanks_to_na(ex)
ae <- convert_blanks_to_na(ae)

admiral exercises - 1

  • Create a new dataset ds_ext with all of the ds, plus derive a numeric date version of DS.DSSTDTC without any imputation as DSSTDT.
    Hint: derive_vars_dt()
10:00

admiral exercises - 2

  • Create a new dataset adsl_1 with all of dm, plus derive an end of study date (EOSDT) using the DSSTDT from ds_ext where DSCAT == "DISPOSITION EVENT" & DSDECOD != "SCREEN FAILURE".
    Hint: derive_vars_merged()
10:00

admiral exercises - 3

  • Create a new dataset adsl_2 with all of adsl_1, plus derive a safety population flag (SAFFL) as Y for any patient from ex with a record where EXDOSE > 0.
    Hint: derive_var_merged_exist_flag()
10:00

admiral exercises - 4

  • Create a new dataset adsl_3 with all of adsl_2, plus derive a last known alive date (LSTALVDT) as the latest date of the following variables from ae: AESTDTC and AEENDTC, imputing missing day and month to the first.
    Hint: derive_var_extreme_dt()
10:00

admiral exercises - 5

10:00

Closing Thoughts

  • Anyone can be become a pharmaverse champion as one of the criteria we look for in pharmaverse recommended packages is great user guides, as hopefully demonstrated wit the sites shown today
  • We were only able to show a small number of pharmaverse packages today, but please delve deeper to explore the full breadth
  • If you’d like to contribute to pharmaverse, check out options at pharmaverse.org
  • Finally, be sure to join our community on Slack!