Skip to main content

Table 2 Median (IQR) of heterogeneity metrics for the treatment effect when no effect modification was presenta

From: Quantifying heterogeneity in individual participant data meta-analysis with binary outcomes

τ 2 Prevalence (%) Number of studies \(I^{2}_{\mathrm {two-stage}}\) \(R^{2}_{\mathrm {two-stage}}\) \(I^{2}_{\mathrm {one-stage}}\) \(R^{2}_{\mathrm {two-stage}}\)
0.5 30 15 0.10 (0.33) 1.22 (0.80) 0.58 (0.41) 1.75 (1.13)
1.0 30 15 0.44 (0.36) 2.30 (1.76) 0.83 (0.15) 3.29 (2.41)
1.5 30 15 0.58 (0.30) 3.22 (2.53) 0.89 (0.09) 5.84 (4.24)
0.5 70 15 0.01 (0.26) 1.00 (0.65) 0.52 (0.47) 1.76 (1.11)
1.0 70 15 0.39 (0.42) 2.12 (1.69) 0.80 (0.18) 3.33 (2.26)
1.5 70 15 0.55 (0.32) 3.07 (2.67) 0.87 (0.11) 5.66 (4.36)
0.5 30 30 0.12 (0.28) 1.29 (0.72) 0.60 (0.32) 1.79 (0.84)
1.0 30 30 0.47 (0.25) 2.44 (1.29) 0.84 (0.09) 3.49 (1.61)
1.5 30 30 0.62 (0.19) 3.54 (1.98) 0.90 (0.05) 6.32 (3.26)
0.5 70 30 0.06 (0.22) 1.14 (0.57) 0.56 (0.34) 1.81 (0.82)
1.0 70 30 0.42 (0.28) 2.30 (1.30) 0.82 (0.11) 3.52 (1.80)
1.5 70 30 0.59 (0.20) 3.37 (1.93) 0.88 (0.07) 6.36 (3.14)
  1. aPlease note that I 2 is presented here as a proportion varying from 0 to 1, rather than as a percentage