perform metaregression with loop on results of meta-analysis loop

Question

I have used code below to perform meta-analysis through loop in r

    meta_results <- list()
meta_summaries_c <- list()
    for (outcom in unique(df$outcom)) {
      outcom_data <- df[df$outcom == outcom, ]
      meta_result <- metagen(
                            TE = SMD,   
                            seTE = seTE,  
                            studlab = paste(study),
                            data = outcom_data,           
                            sm = "SMD",              
                            method.tau = "REML")
      meta_results[[stage]] <- meta_result
    }
meta_summaries_c[[outcom]] <- summary(meta_result)
    print(meta_results)

I want to use my variables and some of outcomes that have more than five study for metaregression

here are my vars and outcoms

variables <- c("var1", "var2", "var3", "var4")
outcomes <- c("outcom1", "outcom2", "outcom3", "outcom4", "outcom5", "outcom6", "outcom7", "outcom8", "outcom9", "outcom10", "outcom11", "outcom12")

Now How can i run a loop through each outcom for each variable to perform metaregression using results of meta_summaries_c?

Answer 1

Perform meta-regression on meta-analysis results :

meta_regression_results <- list()

for (outcom in outcomes) {
  if (outcom %in% names(meta_results)) {
    meta_result <- meta_results[[outcom]]
    outcom_data <- df[df$outcom == outcom, ]
    
    # Loop through each variable for meta-regression
    for (var in variables) {
      
      # Perform meta-regression
      meta_regression <- metareg(
        meta_result,
        formula = as.formula(paste(var, "~ 1")),
        data = outcom_data
      )
      
      # Store results
      meta_regression_results[[paste(outcom, var, sep = "_")]] <- meta_regression
    }
  }
}

# Print meta-regression results
print(meta_regression_results)

This approach ensures you explicitly control the outcomes used in the meta-analysis and meta-regression by specifying them in the outcomes vector.