Regression Analysis for Outcome-Dependent Sampling Design under the Covariate-Adjusted Additive Hazards Model
Table 1
Simulation results when .
Cutpoints
Method
Bias
SD
SE
CP
Bias
SD
SE
CP
(0.3, 0.7)
Oracle
0.1198
0.1149
0.931
0.0052
0.0991
0.0948
0.941
Naive
0.1202
0.1151
0.935
0.1154
0.0697
0.0631
0.523
Proposed
0.1200
0.1149
0.932
0.0014
0.1024
0.0980
0.943
(0.2, 0.8)
Oracle
0.0996
0.0962
0.936
0.0014
0.0801
0.0800
0.942
Naive
0.0994
0.0962
0.935
0.1119
0.0544
0.0531
0.436
Proposed
0.0996
0.0962
0.937
0.0828
0.0830
0.945
(0.3, 0.7)
Oracle
0.0959
0.0978
0.944
0.0030
0.0797
0.0792
0.950
Naive
0.0958
0.0977
0.947
0.1138
0.0544
0.0537
0.428
Proposed
0.0961
0.0978
0.942
0.0015
0.0828
0.0822
0.955
(0.2, 0.8)
Oracle
0.0871
0.0880
0.953
0.0026
0.0723
0.0715
0.951
Naive
0.0869
0.0880
0.956
0.1109
0.0510
0.0485
0.351
Proposed
0.0869
0.0880
0.955
0.0753
0.0745
0.951
(0.3, 0.7)
Oracle
0.0940
0.0951
0.952
0.0168
0.0777
0.0752
0.957
Naive
0.0934
0.0951
0.950
0.1184
0.0533
0.0501
0.308
Proposed
0.0939
0.0951
0.953
0.0155
0.0817
0.0783
0.959
(0.2, 0.8)
Oracle
0.0941
0.0927
0.941
0.0163
0.0717
0.0725
0.949
Naive
0.0934
0.0926
0.941
0.1157
0.0507
0.0486
0.304
Proposed
0.0941
0.0927
0.942
0.0146
0.0750
0.0755
0.952
“Oracle” denotes the oracle estimator calculated based on the true covariate. “Naive” denotes the naive estimator calculated based on the unadjusted covariate. “Proposed” denotes the proposed estimator calculated based on the adjusted covariate.