Hi,

I have a question about the behavior of subgroup analysis when tau.common = TRUE. When this option is used, the confidence interval for this common τ estimate does not appear to be calculated or stored in the resulting object.

I've reviewed the package documentation but couldn't find a specific explanation for why the CI is not calculated in this case. Am I missing something in my understanding, or is this the intended behavior?

library(meta)
#> Loading required package: metadat
#> Loading 'meta' package (version 8.2-1).
#> Type 'help(meta)' for a brief overview.

set.seed(123)

data <- data.frame(
  studlab = paste("Study", 1:10),
  event.e = rpois(10, 25),
  n.e = rpois(10, 100),
  event.c = rpois(10, 30),
  n.c = rpois(10, 100),
  subgroup = rep(c("Drug A", "Drug B"), each = 5)
)

# Using tau.common = FALSE
meta <- metabin(
  event.e = event.e,
  n.e = n.e,
  event.c = event.c,
  n.c = n.c,
  studlab = studlab,
  data = data,
  subgroup = subgroup,
  tau.common = FALSE
)

meta[["upper.tau.w"]]
#>    Drug A    Drug B 
#> 0.5903846 0.9797538

# Using tau.common = TRUE
meta = update(meta, tau.common = TRUE)

meta[["upper.tau.w"]]
#> Drug A Drug B 
#>     NA     NA
meta[["upper.tau.resid"]]
#> [1] NA

^{Created on 2025-09-17 with reprex v2.1.1}

Edit:
Some other situations, confidence intervals are not calculated:
1- Confidence interval for I2 and tau2 measures when k = 2
2- Confidence interval for I2 when using method.I2 = "tau2"
3- Prediction interval when using method.predict = "HK" and k = 2

for those three additional situations, I think they are normally calculated in metafor package
I think providing documentation for this behavior would be beneficial

and thanks again for this great package