Adjustment for baseline characteristics in randomized trials using logistic regression: sample-based model versus true model

Table 4 Proportion of type 1 errors, for different values of the confounder effect (β₂) and of sample size, for 3 logistic regression models: unadjusted for confounder, adjusted in sample-based model, and adjusted in true model

Sample size	Unadjusted analysis	Adjusted, sample-based model	Adjusted, true model	Unadjusted analysis	Adjusted, sample-based model	Adjusted, true model
	Continuous confounder effect β₂ = 0			Binary confounder effect β₂ = 0
2 × 50	0.056	0.053	0.056	0.056	0.052	0.056
2 × 100	0.056	0.053	0.056	0.055	0.051	0.055
2 × 200	0.050	0.050	0.050	0.052	0.051	0.052
2 × 500	0.049	0.050	0.049	0.050	0.051	0.050
2 × 1000	0.050	0.049	0.050	0.050	0.049	0.050
	Continuous confounder effect β₂ = − 0.5			Binary confounder effect β₂ = − 0.5
2 × 50	0.057	0.052	0.050	0.056	0.052	0.050
2 × 100	0.055	0.050	0.049	0.056	0.052	0.050
2 × 200	0.051	0.050	0.050	0.050	0.049	0.048
2 × 500	0.048	0.048	0.048	0.052	0.050	0.050
2 × 1000	0.053	0.051	0.051	0.051	0.051	0.050
	Continuous confounder effect β₂ = − 1			Binary confounder effect β₂ = − 1
2 × 50	0.059	0.051	0.050	0.056	0.051	0.050
2 × 100	0.057	0.051	0.051	0.053	0.049	0.048
2 × 200	0.052	0.051	0.051	0.049	0.050	0.050
2 × 500	0.049	0.051	0.050	0.050	0.048	0.048
2 × 1000	0.053	0.051	0.051	0.050	0.049	0.049

ISSN: 1745-6215