Skip to main content

Table 1 Log odds ratio estimates from simulation study

From: A comparison of methods to estimate the survivor average causal effect in the presence of missing data: a simulation study

ORUYORUZEstimation methodMonotonicity
ValidViolated
EstimateaSESBMSECoveragebEstimateaSESBMSECoverageb
0.50.5Average τ0.920.09   0.920.09   
Marginal structural model−0.540.12−290.0191.6− 0.530.11−140.0194.6
Sensitivity analysis          
SP = 0.5−0.340.111640.04 −0.420.10920.02 
SP = 1−0.540.11−260.01 −0.520.10−40.01 
SP = 2−0.840.11− 3050.12 −0.670.10− 1590.04 
0.52.0Average τ1.100.11   1.100.11   
Marginal structural model−0.480.11270.0193.6−0.500.10110.0192.4
Sensitivity analysis          
SP = 0.5−0.330.101760.04 −0.420.10950.02 
SP = 1−0.480.10300.01 −0.490.10200.01 
SP = 2−0.710.10− 1970.05 −0.610.10− 1060.02 
2.00.5Average τ1.120.09   1.120.09   
Marginal structural model−0.460.10540.0190.4−0.490.09230.0192.6
Sensitivity analysis          
SP = 0.5−0.260.092660.07 −0.380.091430.02 
SP = 1−0.470.09460.01 −0.480.09330.01 
SP = 2−0.760.09− 2710.07 −0.640.09− 1400.02 
2.02.0Average τ0.900.08   0.890.07   
Marginal structural model−0.530.09− 250.0192.9−0.520.08−90.0194.3
Sensitivity analysis          
SP = 0.5−0.360.081830.03 −0.430.081040.01 
SP = 1−0.520.08−70.01 −0.500.0890.01 
SP = 2−0.750.08− 2900.06 −0.620.08− 1440.02 
  1. aEstimates of the log odds ratio have been averaged over 1200 simulated datasets from each scenario
  2. bCoverage indicates the percentage of datasets in each scenario where the true value of the SACE was within the bias-corrected bootstrap confidence interval of the marginal structural model
  3. MSE Mean square error, SACE Survivor average causal effect, SB Standardized bias as a percentage, SE Empirical standard error, SP Sensitivity parameter.
  4. ORUY is the odds ratio effect of U on the outcome. ORUZ is the odds ratio effect of U on survival. τ is the ratio of the odds of the outcome following high iron intake between compliant-survivors and always-survivors. True SACE log odds ratio = ln(0.6) = − 0.511.