Skip to contents

Forest Plot of Adverse Events

Source: R/ae_forestplot.R
ae_forest_plot.Rd

ae_forest_plot() takes arguments for displaying Risk and counts data (from risk_stat()) in the form of a forest plot with elements - list of AE terms corresponding to scatter plot, forest of risk statistics and p-values according to treatment pair.

Usage

ae_forest_plot(
  datain,
  series_opts,
  trtpair_color = rep("black", 10),
  axis_opts,
  term_label = "",
  risk_ref = 1,
  ht_dispyn = "N",
  pvalue_dispyn = "Y",
  text_size = 2.4,
  highlight_sig = "N",
  pvalue_sig = 0.05,
  pairwise = "N",
  terms_perpg = NULL,
  rel_widths = c(0.55, 0.35, 0.1),
  interactive = "N"
)

Arguments

datain

Input data from risk_stat() output, to be used for plotting

series_opts

Series Variable styling options, a list containing shape, color and size; applicable to percent scatter plot with TRTVAR as series variable

trtpair_color

Series colors for treatment pair statistic (i.e. error bars in risk plot and p-value table)

axis_opts

A list of axis specific options retrieved from plot_axis_opts(); applied primarily to plot of risk statistics

term_label

X axis label to display above list of AE terms.

risk_ref

X intercept Value to draw vertical reference line in risk plot

ht_dispyn

Display column for Event High Term corresponding to events. Values: "Y"/"N"

pvalue_dispyn

Display column(s) for p-value corresponding to term. Values: "Y"/"N"

text_size

Size of text for the AE terms as well as p-values Used only if ht_dispynand/or pvalue_dispyn = Y.

highlight_sig

To highlight points in scatter plot having significant differences reflected as percentage per treatment/event? Values: "Y"/"N"

pvalue_sig

p value cutoff to determine and display significant differences from control treatment, if highlight_sig = "Y"

pairwise

Split plot as 1 treatment pair per page? Values: "Y"/"N". If N, all pairs are displayed together.

terms_perpg

Number of AE terms to display per page. If NULL, all terms are shown.

rel_widths

Relative Widths of each sub-plot in the order: High Term (optional)~Percent Scatter Plot~Risk Plot~p value table (optional). Number of values should change per number of subplots, but total should be equal to 1. eg; c(0.25, 0.38, 0.27, 0.10)/"0.25~0.38~0.27~0.1", if ht_dispyn and pvalues_dispyn = Y. eg. If only ht_dispyn = Y; "0.2~0.45~0.35" or if only pvalue_dispyn = Y, c(0.45, 0.35, 0.2)

interactive

Return interactive plot (ggplotly)? Values: "Y"/"N"

Value

: a list of plots/plotly objects.

Details

  • Treatments with Risk statistic of a statistically significant p-value (below pvalue_sig) are highlighted as 'Significantly Higher' or 'Significantly Lower' in the legend, relative to the incidence rate of each term on the control treatment.

  • Color, shape and size for each treatment in scatter plot passed in series_opts and "TRTTXT" variable for legend labels are recommended to be created from plot_aes_opts()

  • Recommended values for risk_ref are 0 for Risk Difference and 1 for Risk Ratio plots.

  • axis_opts applies primarily to risk plot - with only X axis related values being used. Elements xsize and xtsize alone are applicable to both percent plot and risk plot.

Examples

data("adae")
ae_pre <- ae_pre_processor(
  datain = adae,
  obs_residual = 0,
  fmq_data = NA
)
ae_entry <- mentry(
  datain = ae_pre$data,
  subset = NA,
  byvar = "AEBODSYS",
  trtvar = "TRTA",
  trtsort = "TRTAN",
  subgrpvar = NA,
  trttotalyn = "N",
  add_grpmiss = "N",
  sgtotalyn = "N",
  pop_fil = "SAFFL"
)
ae_risk_forest <- risk_stat(
  datain = ae_entry,
  a_subset = ae_pre$a_subset,
  summary_by = "Patients",
  eventvar = "AEDECOD",
  ctrlgrp = "Placebo",
  trtgrp = "Xanomeline High Dose",
  statistics = "Risk Ratio",
  alpha = 0.05,
  cutoff_where = "FREQ >5",
  sort_opt = "Ascending",
  sort_var = "Count"
) |>
  plot_display_bign(
    ae_entry,
    bignyn = "Y"
  )
#> mcatstat success

ae_forest_plot(
  datain = ae_risk_forest,
  series_opts = plot_aes_opts(
    datain = ae_risk_forest,
    series_color = c("black", "royalblue2"),
    series_size = rep(1, 5)
  ),
  axis_opts = plot_axis_opts(
    xaxis_label = "Risk Ratio",
    xopts = list(labelsize = 8)
  ),
  text_size = 2.4,
  term_label = "Reported Term for the Adverse Event",
  risk_ref = 1,
  highlight_sig = "N"
)[[1]]
#> Warning: Ignoring unknown parameters: `linewidth`