Application of Seemingly Unrelated Regression in Medical Data with Intermittently Observed Time-Dependent Covariates

Table 4

Simulation results (summary of estimated parameters, bias, and MSE) for intermittent observation of time-dependent covariates (situation 4 from Table 1) with different sample sizes (𝑛=20,60,120, and 240).

Parameter; (True value)	Method	Sample size = 20			Sample size = 60			Sample size = 120			Sample size = 240
		^↼𝛽	Bias	MSE	^↼𝛽	Bias	MSE	^↼𝛽	Bias	MSE	^↼𝛽	Bias	MSE
𝛽2; (−1.13)	SURCC	−1.002	.13	7.13	−1.14	−.01	1.87	−1.12	.01	.83	−1.13	−.005	.42
	MRMCC	−1.01	.13	6.75	−1.17	−.04	1.94	−1.13	.002	.87	−1.14	−.007	.44
	SURBOCF	−1.02	.11	7.42	−1.14	−.01	1.88	−1.12	.01	.83	−1.14	−.006	.42
	MRMBOCF	−1.03	.10	6.22	−1.16	−.03	1.82	−1.12	.01	.83	−1.14	−.009	.42
	SURLOCF	−1.02	.11	7.42	−1.14	−.01	1.88	−1.12	.01	.83	−1.14	−.006	.42
	MRMLOCF	−1.03	.10	6.22	−1.16	−.03	1.82	−1.12	.01	.83	−1.14	−.009	.42
𝛽3; (.23)	SURCC	.18	−.05	7.17	.14	−.09	1.88	.21	−.02	.86	.24	.01	.43
	MRMCC	.16	−.07	6.32	.12	−.11	1.94	.19	−.03	.89	.23	.004	.44
	SURBOCF	.18	−.05	7.59	.14	−.08	1.89	.21	−.02	.86	.25	.01	.42
	MRMBOCF	.15	−.08	5.97	.13	−.10	1.81	.20	−.03	.85	.24	.01	.42
	SURLOCF	.18	−.05	7.59	.14	−.08	1.89	.21	−.02	.86	.25	.01	.42
	MRMLOCF	.15	−.08	5.97	.13	−.10	1.81	.20	−.03	.85	.24	.01	.42
𝛽4; (−.01)	SURCC	−.08	−.07	1.79	−.02	−.009	.47	−.02	−.01	.23	−.01	−.004	.10
	MRMCC	−.02	−.01	3.36	.01	.02	.93	−.01	−.004	.45	−.02	−.008	.19
	SURBOCF	.04	.05	3.62	−.02	.02	.81	−.01	−.0004	.41	−.02	−.008	.18
	MRMBOCF	−.02	−.01	2.76	.009	.02	.76	−.01	−.004	.39	−.02	−.008	.17
	SURLOCF	.04	.05	3.62	−.02	.02	.81	−.01	−.0004	.41	−.02	−.008	.18
	MRMLOCF	−.02	−.01	2.76	.009	.02	.76	−.01	−.004	.39	−.02	−.008	.17
𝛽5; (−1.12)	SURCC	−1.10	.02	2.27	−1.12	−.005	.57	−1.08	.03	.27	−1.10	.02	.13
	MRMCC	−1.13	−.01	2.40	−1.15	−.03	.64	−1.09	.03	.32	−1.10	.02	.15
	SURBOCF	−1.11	.006	2.39	−1.12	−.004	.57	−1.08	.03	.27	−1.10	.02	.13
	MRMBOCF	−1.14	−.02	1.87	−1.13	−.006	.54	−1.08	.04	.26	−1.10	.02	.13
	SURLOCF	−1.11	.006	2.39	−1.12	−.004	.57	−1.08	.03	.27	−1.10	.02	.13
	MRMLOCF	−1.14	−.02	1.87	−1.13	−.006	.54	−1.08	.04	.26	−1.10	.02	.13

*Note: SUR_CC,MRM_CC,SUR_BOCF,MRM_BOCF,SUR_LOCF,  and MRM_LOCF represent the estimated parameters from MRM and SUR models under complete case (CC) analysis, baseline observation carried forward (BOCF), and last observation carried forward (LOCF) method, respectively.