Create IQWiG-Compliant Forest Plot for Bayesian Meta-Analysis — create_bayesian_forest_plot

Generates a forest plot formatted according to IQWiG guidelines with:

Study weights displayed
Pooled effect with credible interval
Heterogeneity statistics
Prediction interval (if applicable)
Proper scaling and formatting

Usage

create_bayesian_forest_plot_iqwig(
  bayesian_result,
  study_data = NULL,
  show_weights = TRUE,
  show_prediction_interval = TRUE,
  digits_estimate = 3L,
  locale = get_locale(),
  title = NULL,
  subtitle = NULL,
  null_value = NULL,
  base_size = 11,
  point_size = 2,
  ci_linewidth = 0.6
)

Arguments

bayesian_result: A bayesian_meta_result object from bayesian_meta_analysis()
study_data: Data frame with study data containing yi, sei, and study_labels columns. If NULL, uses metadata from bayesian_result if available.
show_weights: Logical. Display study weights (default: TRUE)
show_prediction_interval: Logical. Show prediction interval (default: TRUE)
digits_estimate: Integer. Decimal places for estimates (default: 3)
locale: Character. Locale for formatting: "en" or "de"
title: Character. Plot title (default: NULL)
subtitle: Character. Plot subtitle with heterogeneity info (default: auto-generated)
null_value: Numeric. Reference line value (default: 1 for ratios, 0 for differences)
base_size: Numeric. Base font size (default: 11)
point_size: Numeric. Size of study point estimates (default: 2)
ci_linewidth: Numeric. Line width for CI lines (default: 0.6)

Value

A ClinicalPlot object containing the forest plot

Note

The prediction interval shown (when enabled) is an approximation for visualization, based on the pooled effect and an average within-study variance term; it is not a full posterior predictive interval.

Examples

if (FALSE) { # \dontrun{
result <- bayesian_meta_analysis(yi = yi, sei = sei, study_labels = labels)
study_df <- data.frame(yi = yi, sei = sei, study_labels = labels)
plot <- create_bayesian_forest_plot_iqwig(result, study_data = study_df)
print(plot)
} # }