A Proposed Approach for Joint Modeling of the Longitudinal and Time-To-Event Data in Heterogeneous Populations: An Application to HIV/AIDS’s Disease
Table 2
Comparison of the estimation of parameters in the three approaches for different sample sizes, .
Models
Class
Parameter
True value
Bias
SE
CP
Bias
SE
CP
Bias
SE
CP
JLCM
Class 1
Intercept
0.5
0.123
0.189
0.935
0.111
0.171
0.957
0.099
0.106
0.917
Class 2
Intercept
−0.5
0.104
0.311
0.928
0.098
0.149
0.942
0.086
0.115
0.973
PA
Class 1
Intercept
0.5
0.063
0.219
0.949
0.031
0.148
0.943
0.013
0.121
0.969
0.2
−0.022
0.204
0.951
−0.018
0.119
0.950
−0.001
0.101
0.952
Class 2
Intercept
−0.5
0.047
0.214
0.933
0.031
0.138
0.952
0.017
0.127
0.988
0.2
0.024
0.118
0.918
0.011
0.121
0.954
−0.001
0.111
0.979
SA
Class 1
Intercept
0.5
0.071
0.278
0.826
0.043
0.158
0.831
0.101
0.143
0.689
0.2
−0.151
0.354
0.911
−0.121
0.131
0.829
−0.103
0.112
0.634
Class 2
Intercept
−0.5
0.064
0.246
0.816
0.031
0.141
0.821
0.097
0.138
0.614
0.2
−0.141
0.297
0.721
−0.119
0.117
0.832
−0.097
0.114
0.633
JLCM: joint latent class models; PA: proposed approach; SA: separate approach; SE: standard error; CP: coverage probability of the 95% confidence interval.