Fig. 5
From: An informatics consult approach for generating clinical evidence for treatment decisions
New observational evidence through target trial emulation (intention-to-treat analysis) where eligibility and treatment assignment were aligned with time zero of follow-up, as is done in randomised controlled trials. A CONSORT diagram showing the selection of eligible individuals for the target trial emulation of anticoagulation therapy in patients with atrial fibrillation and cirrhosis. B Kaplan–Meier plots of the propensity-matched cohort for all-cause mortality and ischaemic stroke. Flow diagram depicts analysis design. P values from logrank tests were indicated. Hazard ratios from Cox proportional hazards regression analyses were indicated. Numbers in parentheses indicate the 95% confidence intervals