Skip to main content

Table 5 Estimates of log hazard ratio in scenario 2 (with an effect modifier)

From: Performance of unanchored matching-adjusted indirect comparison (MAIC) for the evidence synthesis of single-arm trials with time-to-event outcomes

Statistical properties Unweighted estimate Weighted estimate
   All variables included in weight Omitted the effect modifier (menopausal status) from weight Omitted the effect modifier (menopausal status) and tumor grade from weight
Bias 0.173 0.035 0.079 0.122
MCV 0.002 0.003 0.003 0.002
MSE 0.032 0.004 0.009 0.017
CI coverage (%) 7.7 89.9 68.7 34.1
mean ESS NA 791 827 888
  1. Abbreviation: MCV Monte-Carlo variance; MSE Mean squared error; CI 95% Confidence interval; ESS Effective sample size