Funnel plot from a meta-analysis

ggfunnel() draws a funnel plot: each study's effect estimate against its standard error (a measure of precision), with pseudo confidence-interval contours around the pooled effect. Asymmetry can signal small-study effects or publication bias. Like ggforest(), it accepts a meta object or a tidy data frame and returns an ordinary ggplot, so a forest and a funnel plot compose on one canvas with, for example, patchwork.

Usage

ggfunnel(x, ...)

# Default S3 method
ggfunnel(x, ...)

# S3 method for class 'meta'
ggfunnel(x, ..., ref = c("common", "random"), level = 0.95)

# S3 method for class 'data.frame'
ggfunnel(
  x,
  centre = NULL,
  sm = NULL,
  level = 0.95,
  xlab = NULL,
  ylab = "Standard error",
  title = NULL,
  point_args = list(),
  contour_args = list(),
  ref_args = list(),
  ...
)

Arguments

x

A meta object, or a data frame with estimate and se columns (study effects and standard errors on the analysis scale).

...

Additional arguments passed to methods.

ref

Which pooled estimate to centre the funnel on: "common" (fixed effect, default) or "random".

level

Confidence level(s) for the funnel contours, e.g. 0.95 or c(0.95, 0.99). Default 0.95.

centre

Effect the funnel is centred on. Defaults to the inverse-variance (common-effect) estimate of the supplied studies.

sm

Summary measure (e.g. "RR", "PLOGIT"), used to label the x-axis and, for transformed measures (ratios, proportions, rates, correlations), to show back-transformed axis labels. Optional.

xlab, ylab

Axis labels. ylab defaults to "Standard error".

title

Plot title. Default NULL.

point_args, contour_args, ref_args

Lists of arguments used to restyle the study points (ggplot2::geom_point()), the funnel contours (geom_funnel_contour()), and the vertical reference line (ggplot2::geom_vline()). For example point_args = list(size = 3, fill = "black") or contour_args = list(colour = "grey70", linetype = "dotted").

Value

A ggplot object.

Examples

# \donttest{
library(meta)
m <- metabin(event.e, n.e, event.c, n.c,
  data = data.frame(
    event.e = c(12, 8, 25, 18, 30, 15), n.e = c(120, 90, 200, 150, 250, 130),
    event.c = c(20, 14, 30, 28, 35, 25), n.c = c(118, 92, 205, 148, 245, 128)
  ),
  studlab = paste("Study", 1:6), sm = "RR"
)
ggfunnel(m)

# }