Does anyone know of an R function for computing the Greenland-Robins variance for Mantel-Haenszel relative risks? Thanks Frank -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt University