NMAforest: Generate Forest Plot with Proportion Contributions for...

In NMAforest: Forest Plots for Network Meta-Analysis with Proportion for Paths and Studies

Generate Forest Plot with Proportion Contributions for Network Meta-analysis

Description

This function extends the capabilities of network meta-analysis visualization by generating a forest plot that incorporates direct, indirect, and network meta-analysis (NMA) treatment effects, along with contribution proportions from individual studies and comparison paths. It builds on the statistical foundation provided by the 'netmeta' package and is particularly useful for evaluating the influence of study evidence in complex treatment networks.

Usage

NMAforest(
  data,
  sm,
  reference,
  model = "random",
  comparison,
  study = "study",
  treat = "t",
  event = "r",
  N = "n",
  mean = "y",
  sd = "sd",
  study_id = "study_id",
  study_path = TRUE
)

Arguments

data	A data frame in long format with required variables study (study identifier) and t (treatment label), plus outcome columns: for binary outcomes, r (events) and N (sample size); for continuous outcomes, y (mean), sd (standard deviation), and N (sample size). Column names are configurable via study, treat, event, N, mean, and sd.
sm	A Character string specifying the summary measure to use (e.g., "OR" for odds ratio, "RR" for risk ratio, "RD" for risk difference, "MD" for mean difference, "SMD" for standardized mean difference, etc.).
reference	Specify the reference treatment for comparisons.
model	Choose "random" or "fixed" effect model.
comparison	A vector of two treatments to compare.
study	Column name identifying the study label.
treat	Column name for treatment assignment.
event	Column name for event counts (for binary outcomes).
N	Column name for total sample size (for binary and continuous outcomes).
mean	Column name for mean values (for continuous outcomes).
sd	Column name for standard deviations (for continuous outcomes).
study_id	Column name used to uniquely identify each study arm (default = "study_id"). If the column is not present in the original dataset, the updated data frame will be returned with this column added.
study_path	Logical. TRUE to include study combination contributions, FALSE to exclude.

Details

This package includes an implementation of 'comparisonStreams()' originally developed by Papakonstantinou et al. (2018) and released with the paper's supplementary material.

Value

A list containing:

plot

A ggplot object of the forest plot.

output

A data frame summarizing the results for the specified treatment comparison, including study-level and overall estimates. Columns contain the effect sizes with their 95% confidence intervals, standard errors, study weights, and contribution proportions, along with labels indicating whether the row corresponds to a study, direct effect, indirect effect, or the overall NMA estimate.

updated_df

The input data frame after preprocessing, with a unique study identifier column (from the argument study_id) added when not already present. This guarantees consistent referencing of studies in both the analysis and the output.

References

Balduzzi, S., Rücker, G., Nikolakopoulou, A., Papakonstantinou, T., Salanti, G., Efthimiou, O., & Schwarzer, G. (2023). netmeta: An R Package for Network Meta-Analysis Using Frequentist Methods. *Journal of Statistical Software*, 106(2), 1–40. https://doi.org/10.18637/jss.v106.i02

Csardi, G., & Nepusz, T. (2006). The igraph software package for complex network research. *InterJournal*, Complex Systems, 1695. https://igraph.org

Csárdi, G., Nepusz, T., Traag, V., Horvát, Sz., Zanini, F., Noom, D., & Müller, K. (2025). *igraph: Network Analysis and Visualization in R*. https://doi.org/10.5281/zenodo.7682609

Wickham, H. (2016). *ggplot2: Elegant Graphics for Data Analysis*. Springer-Verlag New York. https://ggplot2.tidyverse.org

Papakonstantinou, T., Nikolakopoulou, A., Rücker, G., Chaimani, A., Schwarzer, G., Egger, M., & Salanti, G. (2018). Estimating the contribution of studies in network meta-analysis: paths, flows and streams. *F1000Research*, 7:610. https://doi.org/10.12688/f1000research.14527.3

flow_contribution GitHub repository: https://github.com/esm-ispm-unibe-ch/flow_contribution

Examples

# Example 1: Binary outcome data
data(example_data)

NMAforest(
  data = example_data,
  sm = "OR",
  reference = "x",
  model = "random",
  comparison = c("x", "y"),
  study = "study",
  treat = "t",
  event = "r",
  N = "n",
  study_id = "id",
  study_path = TRUE
)

# Example 2: Continuous outcome data
data("parkinson", package = "pcnetmeta")

NMAforest(
  data = parkinson,
  sm = "MD",
  reference = "1",
  model = "random",
  comparison = c("1", "3"),
  study = "s.id",
  treat = "t.id",
  mean = "mean",
  sd = "sd",
  N = "n",
  study_path = TRUE
)

NMAforest documentation built on Feb. 10, 2026, 9:06 a.m.