Skip to contents

A closer look at meta_forestly()

Source: vignettes/meta-forestly.Rmd
meta-forestly.Rmd
library(forestly)
#> Error in get(paste0(generic, ".", class), envir = get_method_env()) : 
#>   object 'type_sum.accel' not found
library(metalite)

Overview

meta_forestly() is the first step in the interactive forest plot generation workflow and is a wrapper of several functions from metalite. Specifically, it defines a metadata, including ADaM datesets, population, observations and parameters to be used in the forest plot. In this vignette we will take a closer look at the parameter_term argument in meta_forestly() to better understand how to generate the forest plot for different parameters of interest.

parameter_term examples

Prior to turning to the examples, just a reminder that the treatment group variable in the ADaM datasets should be set as a factor, otherwise a warning will be generated. For example, if this is what you have in the TRT01A:

unique(forestly_adsl$TRT01A)
#> [1] "Placebo"             "Xanomeline Low Dose"

you can easily turn it into a factor variable with the desired labels using the following code:

forestly_adsl$TRTA <- factor(
  forestly_adsl$TRT01A, 
  levels = c("Xanomeline Low Dose", "Placebo"), 
  labels = c("Low Dose", "Placebo")
  )

A similar approach needs to be applied to the TRTA:

forestly_adae$TRTA <- factor(
  forestly_adae$TRTA, 
  levels = c("Xanomeline Low Dose", "Placebo"), 
  labels = c("Low Dose", "Placebo")
  )

If the deliverable requires a forest plot of all adverse events and serious adverse events, then the following code could be used with any and ser parameters in parameter_term argument respectively:

meta_forestly(
  dataset_adsl = forestly_adsl,
  dataset_adae = forestly_adae,
  parameter_term = "any;ser",
  population_subset = SAFFL == "Y",
  observation_subset = SAFFL == "Y"
) |>
  prepare_ae_forestly() |>
  format_ae_forestly() |>
  ae_forestly()

However, if grade 3-5 adverse events need to be plotted as well, we can add g35 parameter in the parameter_term. Click on the AE criteria below to see that a forest plot for grade 3-5 adverse events was added.

meta_forestly(
  dataset_adsl = forestly_adsl,
  dataset_adae = forestly_adae,
  parameter_term = "any;ser;g35",
  population_subset = SAFFL == "Y",
  observation_subset = SAFFL == "Y"
) |>
  prepare_ae_forestly() |>
  format_ae_forestly() |>
  ae_forestly()

Adding a new parameter

Below we demonstrate how to add a new parameter to the meta_forestly(). Consider, for instance, that now we need to plot adverse events which resulted in dose reductions, a parameter that is currently not available in the predefined options of the parameter_term. Instead of using meta_forestly() directly, we will define meta_new object using metalite core functions as demonstrated below.

Specifically, define_parameter() is where the new parameter reduc is defined by filtering forestly_adae with AEACN == "DOSE REDUCED" condition in order to identify the adverse events of interest:

meta_new <- meta_adam(
  population = forestly_adsl,
  observation = forestly_adae
  ) |>
  define_plan(plan = metalite::plan(
    analysis = "ae_forestly",
    population = "apat",
    observation = "all",
    parameter = "any;ser;g35;reduc"
    )
  ) |>
  define_population(
    name = "apat",
    group = "TRTA",
  ) |>
  define_observation(
    name = "all",
    group = "TRTA",
    label = "All AEs"
  ) |>
  define_parameter(
    name = "reduc",
    subset = AEACN == "DOSE REDUCED",
    label = "adverse events resulting in dose reduction"
  ) |>
  define_analysis(
    name = "ae_forestly",
    label = "Interactive forest plot"
  ) |>
  meta_build()

Now, similar to the originally generated forest plot above, we need to apply prepare_ae_forestly(), format_ae_forestly(), and ae_forestly() to generate the plot that will include adverse events which resulted in dose reductions: