Perform Meta-Analysis

Conducts fixed-effect or random-effects meta-analysis on a set of studies. Supports binary, continuous, and time-to-event outcomes.

Usage

meta_analysis(
  data = NULL,
  yi = NULL,
  sei = NULL,
  ni = NULL,
  study_labels = NULL,
  effect_measure = c("hr", "or", "rr", "rd", "md", "smd"),
  model = c("random", "fixed"),
  method = c("REML", "DL", "PM"),
  knapp_hartung = TRUE,
  conf_level = 0.95,
  prediction = TRUE
)

Arguments

data

Data frame with study-level summary statistics. StudySet input is planned but not yet implemented: the current S7::S7_inherits(data, StudySet) code path aborts with ph_abort("StudySet input not yet implemented"). For now, provide yi and sei directly.

yi

Numeric vector of effect estimates (log scale for ratios)

sei

Numeric vector of standard errors

ni

Numeric vector of sample sizes (optional)

study_labels

Character vector of study names

effect_measure

Character. Type of effect: "hr", "or", "rr", "rd", "md", "smd"

model

Character. "fixed" or "random". Default: "random"

method

Character. Estimation method for random effects: "DL" (DerSimonian-Laird), "REML", "PM" (Paule-Mandel). Default: "REML"

knapp_hartung

Logical. Apply Knapp-Hartung adjustment. Default: TRUE

conf_level

Numeric. Confidence level. Default: 0.95

prediction

Logical. Calculate prediction interval. Default: TRUE

Value

A MetaResult S7 object with pooled estimate and heterogeneity stats

Examples

# Random-effects meta-analysis of 5 studies with hazard ratios
result <- meta_analysis(
  yi = log(c(0.75, 0.82, 0.68, 0.91, 0.77)),
  sei = c(0.12, 0.15, 0.18, 0.14, 0.11),
  study_labels = paste("Study", 1:5),
  effect_measure = "hr",
  model = "random",
  method = "REML",
  knapp_hartung = TRUE
)
result@estimate
#> [1] 0.7823178
result@ci
#> [1] 0.6840484 0.8947045
result@heterogeneity$I2
#> [1] 0