Figure 3
From: Assessing discriminative ability of risk models in clustered data
Meta-analysis pooling weights for European centers. For methods 1 to 8 the weights are represented by the height of the bars on the right hand side of the Figure. C-indexes are printed in the bars if the cluster weight was at least equal to the average weight. 1 Equal = Fixed effect meta-analysis with equal weights; 2 Patients = Fixed effect meta-analysis with number of patients as weights; 3 Events = Fixed effect meta-analysis with number of events as weights; 4 Pairs = Fixed effect meta-analysis with number of usable patient pairs as weights; 5 Inv var FE = Fixed effect meta-analysis with inverse variance weights; 6 Inv var RE = Random effects meta-analysis with inverse variance weights; 7 Inv var FE logit = Fixed effect meta-analysis with inverse variance weights on log-odds scale; 8 Inv var RE logit = Random effects meta-analysis with inverse variance weights on log-odds scale.