Funnel Plot

Authors

[Editor] Hu Zheng;

[Contributors]

Can be used to show potential bias factors in Meta-analysis.

Setup

System Requirements: Cross-platform (Linux/MacOS/Windows)
Programming language: R
Dependent packages: data.table; jsonlite; FunnelPlotR; gridExtra

# Install packages
if (!requireNamespace("data.table", quietly = TRUE)) {
  install.packages("data.table")
}
if (!requireNamespace("jsonlite", quietly = TRUE)) {
  install.packages("jsonlite")
}
if (!requireNamespace("FunnelPlotR", quietly = TRUE)) {
  install.packages("FunnelPlotR")
}
if (!requireNamespace("gridExtra", quietly = TRUE)) {
  install.packages("gridExtra")
}

# Load packages
library(data.table)
library(jsonlite)
library(FunnelPlotR)
library(gridExtra)

Data Preparation

# Load data
data <- data.table::fread(jsonlite::read_json("https://hiplot.cn/ui/basic/funnel-plot/data.json")$exampleData$textarea[[1]])
data <- as.data.frame(data)

# View data
head(data)

  los hmo white died age80 type type1 type2 type3 provnum     prds
1   4   0     1    0     0    1     1     0     0   30001 9.667315
2   9   1     1    0     0    1     1     0     0   30001 8.956472
3   3   1     1    1     1    1     1     0     0   30001 6.856678
4   9   0     1    0     0    1     1     0     0   30001 9.667315
5   1   0     1    1     1    1     1     0     0   30001 7.400868
6   4   0     1    1     0    1     1     0     0   30001 7.561051

Visualization

# Funnel Plot
p <- funnel_plot(
  data, numerator = los, denominator = prds,  group = provnum, data_type = "SR",
  limit = 99, label = "outlier", sr_method = "SHMI", trim_by=0.1, 
  title = "Funnel Plot", x_range = "auto", y_range = "auto"
  )

p

A funnel plot object with 54 points of which 9 are outliers. 
Plot is adjusted for overdispersion.