Skip to main content

Table 2 Simulation results for comparisons of methods in estimating the difference in restricted mean survival time. Scenario with 5 trials and 200 patients per trial, and with non-proportional hazards

From: Bias and precision of methods for estimating the difference in restricted mean survival time from an individual patient data meta-analysis

Heterogeneity scenario Methods β = 0 β = −0.7
t * = 5 True rmstD = 0 t * = 10 True rmstD = 0 t * = 5 True rmstD = −0.3 t * = 10 True rmstD = 0.3
Bias ESE ASE Bias ESE ASE Bias ESE ASE Bias ESE ASE
22) = (0.01;0.01) Naïve Kaplan-Meier 0.00 0.12 0.11 0.00 0.28 0.24 0.00 0.12 0.11 0.00 0.28 0.24
Pooled Kaplan-Meier 0.00 0.12 0.13 0.01 0.29 0.30 0.01 0.12 0.13 −0.05 0.29 0.29
Pooled Exponential 0.00 0.10 0.11 0.00 0.27 0.28 0.40 0.12 0.12 0.07 0.30 0.30
Peto-quintile 0.00 0.11 0.10 0.00 0.24 0.21 0.02 0.11 0.10 −0.19 0.24 0.21
22) = (0.01;0.10) Naïve Kaplan-Meier 0.00 0.19 0.11 0.02 0.50 0.24 0.00 0.19 0.11 0.00 0.48 0.24
Pooled Kaplan-Meier 0.01 0.19 0.18 0.02 0.50 0.46 0.00 0.19 0.18 −0.05 0.48 0.46
Pooled Exponential 0.00 0.19 0.17 0.01 0.49 0.45 0.39 0.20 0.18 0.05 0.50 0.47
Peto-quintile 0.00 0.18 0.10 0.01 0.43 0.22 0.02 0.18 0.10 −0.18 0.42 0.22
22) = (0.10;0.01) Naïve Kaplan-Meier 0.00 0.12 0.11 0.00 0.28 0.24 0.00 0.12 0.11 −0.01 0.27 0.24
Pooled Kaplan-Meier 0.00 0.12 0.13 0.00 0.29 0.29 0.00 0.12 0.13 −0.07 0.28 0.29
Pooled Exponential 0.00 0.10 0.11 0.00 0.27 0.28 0.38 0.12 0.12 0.04 0.30 0.31
Peto-quintile 0.00 0.11 0.10 0.00 0.24 0.21 0.00 0.11 0.10 −0.21 0.25 0.22
22) = (0.10;0.10) Naïve Kaplan-Meier −0.01 0.19 0.11 −0.01 0.46 0.24 0.00 0.20 0.11 0.01 0.49 0.24
Pooled Kaplan-Meier −0.01 0.19 0.18 −0.02 0.47 0.45 0.00 0.20 0.18 −0.05 0.48 0.45
Pooled Exponential −0.01 0.18 0.17 −0.01 0.45 0.44 0.38 0.20 0.18 0.05 0.50 0.46
Peto-quintile −0.01 0.18 0.10 −0.01 0.41 0.22 0.01 0.19 0.10 −0.18 0.44 0.22
  1. A DerSimonian-Laird random effects meta-analysis model was used for Pooled Kaplan-Meier, Pooled Exponential and Peto-quintile
  2. β Size of treatment effect (=log(HR)), σ 2 baseline hazard heterogeneity, τ 2 treatment effect heterogeneity, ASE average standard error, CI Confidence interval, ESE empirical standard error, rmstD difference in restricted mean survival time, t * time horizon