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 4

Comparison of the results for the three approaches on HIV/AIDS patients.

	Parameter name	PA			SA			JLCM
		Estimates	SE	value	Estimates	SE	value	Estimates	SE	value
Class 1	Longitudinal submodel
	Intercept	0.626	0.054	<0.001	0.612	0.052	<0.001	1.057	0.134	<0.001
	Time	−0.011	0.002	<0.001	−0.006	0.002	0.002	−0.007	0.002	<0.001
	Treatment (ddC)	0.029	0.018	0.106	0.0282	0.018	0.124	−0.016	0.039	0.681
	Hgb	0.008	0.004	0.048	0.007	0.004	0.049	−0.001	0.007	0.815
	Survival submodel
	Treatment (ddC)	0.055	0.120	0.646	0.151	0.170	0.373	−0.259	0.955	0.785
	Hgb	0.034	0.025	0.174	0.006	1.006	0.848	0.308	0.096	0.001
		−0.508	0.644	<0.001	−0.480	0.296	<0.001	-	-	-
Class 2	Longitudinal submodel
	Intercept	1.163	0.086	<0.001	1.152	0.086	<0.001	0.708	0.098	<0.001
	Time	−0.006	0.002	0.003	−0.004	0.002	0.055	−0.011	0.002	<0.001
	Treatment (ddC)	0.021	0.022	0.345	0.022	0.022	0.329	0.004	0.028	0.877
	Hgb	−0.018	0.005	<0.001	−0.018	0.005	0.001	−0.002	0.006	0.709
	Survival submodel
	Treatment (ddC)	−0.762	0.221	<0.001	−0.763	0.321	0.017	−0.558	0.263	0.034
	Hgb	0.059	0.059	0.317	0.031	0.082	0.702	−0.006	0.060	0.917
		−0.488	0.719	<0.001	−0.430	0.494	<0.001	-	-	-

JLCM: joint latent class models; PA: proposed approach; SA: separate approach; SE: standard error; Hgb: baseline hemoglobin values; ddC: Zalcitabine treatment.