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

Table 2 Variance in the estimation of treatment effect (b₁), for different values of the true treatment effect (β₁) and of sample size, with a strong confounder effect (β₂ = − 1): 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
	Treatment effect β₁ = 0, continuous confounder			Treatment effect β₁ = 0, binary confounder
2 × 50	0.17	0.18	0.18	0.17	0.18	0.18
2 × 100	0.08	0.09	0.09	0.08	0.09	0.09
2 × 200	0.04	0.04	0.04	0.04	0.04	0.04
2 × 500	0.02	0.02	0.02	0.02	0.02	0.02
2 × 1000	0.01	0.01	0.01	0.01	0.01	0.01
	Treatment effect β₁ = 0.5, continuous confounder			Treatment effect β₁ = 0.5, binary confounder
2 × 50	0.17	0.19	0.18	0.17	0.18	0.18
2 × 100	0.08	0.09	0.09	0.08	0.09	0.09
2 × 200	0.04	0.04	0.04	0.04	0.04	0.04
2 × 500	0.02	0.02	0.02	0.02	0.02	0.02
2 × 1000	0.01	0.01	0.01	0.01	0.01	0.01
	Treatment effect β₁ = 1, continuous confounder			Treatment effect β₁ = 1, binary confounder
2 × 50	0.19	0.21	0.20	0.18	0.20	0.19
2 × 100	0.09	0.10	0.10	0.09	0.10	0.09
2 × 200	0.05	0.05	0.05	0.04	0.05	0.05
2 × 500	0.02	0.02	0.02	0.02	0.02	0.02
2 × 1000	0.01	0.01	0.01	0.01	0.01	0.01

ISSN: 1745-6215