Fig. 4
From: Interval-cohort designs and bias in the estimation of per-protocol effects: a simulation study
Naive versus IP weighted estimates under strong confounding but complete measurement of covariate history
Fig. 4 | Trials
From: Interval-cohort designs and bias in the estimation of per-protocol effects: a simulation study
Naive versus IP weighted estimates under strong confounding but complete measurement of covariate history
Trials