Multiple imputation for handling missing outcome data when estimating the relative risk

Table 2 Results for X ₁ and X ₂ Continuous, λ = 2, Corr(X ₁, X ₂) = 0.70 and β ₁ = β ₂ = log(3)

Simulation scenario	Method	β ₁					β ₂
Simulation scenario	Method	Bias	Avg SE	Emp SE	Coverage	MSE	Bias	Avg SE	Emp SE	Coverage	MSE
Coordinated, prevalence =0.10	MVNI	−0.56	0.48	0.39	0.838	0.47	−0.22	0.50	0.43	0.958	0.24
	MVNI + deletion	0.01	0.58	0.55	0.965	0.30	−0.03	0.55	0.53	0.959	0.28
	FCS	−0.08	0.51	0.49	0.961	0.25	−0.14	0.50	0.48	0.950	0.25
	FCS + deletion	0.02	0.58	0.55	0.964	0.30	−0.04	0.55	0.53	0.961	0.28
	CCA	0.01	0.66	0.67	0.943	0.45	0.01	0.53	0.55	0.936	0.30
Coordinated, prevalence =0.30	MVNI	−0.26	0.26	0.20	0.890	0.11	−0.27	0.25	0.20	0.859	0.11
	MVNI + deletion	0.01	0.31	0.26	0.978	0.07	−0.11	0.28	0.23	0.963	0.07
	FCS	−0.09	0.26	0.22	0.962	0.06	−0.24	0.24	0.21	0.878	0.10
	FCS + deletion	0.02	0.31	0.26	0.980	0.07	−0.12	0.28	0.23	0.963	0.07
	CCA	0.02	0.32	0.32	0.951	0.11	0.00	0.25	0.26	0.950	0.07
Opposite, prevalence =0.10	MVNI	−0.58	0.47	0.37	0.830	0.47	−0.17	0.48	0.42	0.961	0.21
	MVNI + deletion	0.00	0.56	0.52	0.966	0.28	0.02	0.53	0.51	0.959	0.26
	FCS	−0.08	0.48	0.46	0.961	0.22	−0.07	0.49	0.47	0.959	0.22
	FCS + deletion	0.01	0.56	0.52	0.971	0.27	0.01	0.53	0.51	0.962	0.26
	CCA	0.00	0.60	0.62	0.939	0.39	0.01	0.48	0.50	0.938	0.25
Opposite, prevalence =0.30	MVNI	−0.25	0.24	0.19	0.886	0.10	−0.07	0.26	0.20	0.981	0.05
	MVNI + deletion	0.01	0.30	0.25	0.983	0.06	0.06	0.29	0.23	0.980	0.06
	FCS	−0.07	0.24	0.21	0.974	0.05	−0.02	0.26	0.22	0.980	0.05
	FCS + deletion	0.02	0.29	0.25	0.983	0.06	0.05	0.29	0.23	0.982	0.06
	CCA	0.00	0.29	0.29	0.945	0.08	0.01	0.23	0.22	0.949	0.05

Abbreviations: MVNI multivariate normal imputation, FCS fully conditional specification, CCA complete case analysis, Avg SE average standard error, Emp SE empirical standard error, MSE mean square error

ISSN: 1471-2288