Table 1.

Risk of Modeled Health Events According to Treatment

	Rate per 100 Patient Years		Hazard Ratio (95% CI) vs Apixaban
	Apixaban	No Treatment26	Warfarin	Dabigatran	Rivaroxaban
Strokea	1.359b	4.186	1.67 (1.54–1.79)7	1.08 (0.90–1.28)7	0.95 (0.87–1.03)7
Intracranial hemorrhage	0.249b	0	2.38 (2.08–2.70)7	1.02 (0.62–1.68)7	1.15 (0.99–1.33)7
Other major bleed	1.603c	0	2.33 (2.17–2.50)7	1.08 (0.93–1.24)7	1.49 (1.39–1.59)7
Clinically relevant non-major bleed	2.090d	0	1.433 (1.241–1.655)d	1.00 (0.900–1.100)34	1.52 (1.28–1.80)34
Myocardial infarction	1.654e	1.003	1.1493	1.460 (0.960–2.200)34	1.060 (0.730–1.520)34
Systemic embolisma	0.060f	0.956	1.67 (1.54–1.79)7	1.08 (0.90–1.28)7	0.95 (0.87–1.03)7
Other CV hospitalizationsg	10.46	16.506	1.00	1.00	1.00
Treatment discontinuation for non-event related reasons	13.107d		1.10 (1.04–1.16)d	1.50 (1.36–1.67)34	1.18 (1.08–1.61)34

Notes: ^aAssumed equal to stroke and systemic embolism. ^bTotal number of events among 16,741 Finnish AF patients using warfarin were 375 for stroke and 99 for ICH.³¹ Assuming a full follow-up of one year for each patient, these were converted to a rate per 100 patient-years: 2.27 and 0.59, respectively. ^cTotal number of bleeding events in 125,261 Finnish patients using warfarin was 5412,³² and the corresponding rate was 4.32 per 100 patient-years. By subtracting the ICH rate from total bleeds, the rate for other major bleeds was estimated at 3.73 per 100 patient-years. ^dSecondary analysis of ARISTOTLE-trial. ^e1.900 per 100 patient-years for warfarin.³³ f0.1 per 100 patient-years for warfarin.⁸ gSecondary analysis of AVERROES-trial, all treatments assumed equal.