Table 1

Baseline characteristics of included studies

	Bassiouny et al15	Bernard et al23	Haines et al16	Ichiki et al24	Kaseno et al17	Kim et al18	Lakkireddy et al9	Maddox et al.19	Mendoza et al25	Nin et al20	Pavaci et al26	Rowley et al27	Snipelisky et al21	Yamaji et al22
Source type	Journal article	Conference abstract	Journal article	Conference abstract	Journal article	Journal article	Journal article	Journal article	Conference abstract	Journal article	Conference abstract	Conference abstract	Journal article	Journal article
Study design	Prospective, monocentric registry	Retrospective	Multicentric, retrospective, case-match analysis	Prospective, non-randomised	Retrospective	Retrospective	Multicentric, prospective, non-randomised	Retrospective	Retrospective	Prospective, randomised	Retrospective, case-match analysis	Retrospective	Retrospective	Retrospective, non-randomised and randomised*
Paroxysmal AF	57% D 55% W	46% D 50% W	53% D 48% W	53%	83% D 55% W	53% D 48% W	57% D 57% W	63% D 57% W	NA	76% D 71% W	NA	NA	68% D 46% W	65% D 62% W
Age	58.6 D 62.7 W	62.7 D 67.3 W	60.2 D 59.7 W	n.s.	59 D 62 W	61 D 61 W	60.4 D 60.3 W	62.3 D 62.5 W	62.9 D 64.0 W	61 D 61 W	NA Case-matched	63±10 similar	60.6 D 64.6 W	60 D 61 W
♀ Gender	25% D 26.6% W	NA	26% D 31% W	NA	25% D 21% W	20% D 26% W	21% D 21% W	24% D 33% W	10% D 12% W	16% D 20% W	NA Case-matched	NA	19.4% D 25.6% W	25% D 24% W
Estimated thrombo-embolic risk	CHADS₂ Score=0 42.6% D 29.1% W Score=1 37.0% D 41.6% W Score ≥2 20.5% D 29.4% W	CHA₂DS₂-VASc 2.00 D 2.68 W	CHA₂DS₂-VASc 1.6±1.3 D 1.9±1.4 W	Values NA ns differences between treatment groups	CHADS₂ Score=0 61% D 45% W Score=1 31%D 41% W Score ≥2 8% D 14% W	CHA₂DS₂-VASc 1.6±1.3 D 1.7±1.3 W	CHADS₂ Score=0 35% D 40% W Score=1 43% D 41% W Score ≥2 23% D 19% W CHA₂DS₂-VASc 1.6±1.4 D 1.5±1.3 W	CHADS₂ 0.92±0.88 D 0.92±0.85 W CHA₂DS₂-VASc 1.73±1.45 D 1.69±1.33 W	CHADS₂ 1.32 D 1.29 W	CHADS₂ Score ≤1 82% D 80% W Score=1 11% D 13% W Score ≥2 4% D 11% W	NA	CHADS₂ 1.3±1	CHADS₂ 0.80 D 1.16 W	CHADS₂ Score=0 36% D 33% W Score=1 40% D 38% W Score ≥2 24% D 29% W CHA₂DS₂-VASc 1.8±1.6 D 1.7±1.6 W
Estimated bleeding risk	NA	NA	NA	NA	HAS-BLED 0.5±0.7 D 0.6±0.6 W	HAS-BLED 1.0±0.9 D 1.1±0.9 W	HAS-BLED 1.2±0.9 D 1.1±0.9 W	NA	HAS-BLED 1.47 D 1.63 W	NA	NA	NA	NA	NA
LV ejection fraction	55% D 55% W	60% D 58% W	56% D 57% W	NA	64% D 63% W	58% D 57% W	56% D 56% W	53% D 54% W	NA	61% D 62% W	NA	NA	NA	60% D 60% W
CKD/kidney function	Creat 0.9 mg/dL D 0.9 mg/dL W	Creat 1.0 mg/dL D 1.3 mg/dL W	Creat clearance 93 ml/min D 93 ml/min W	NA	Creat clearance 97 ml/min D 98 ml/min W	Creat 0.9 mg/dL D 0.9 mg/dLW	CKD 1% D 2% W	Creat 1.0 mg/dL D 1.1 mg/dL W	NA	Creat 0.9 mg/dL D 0.9 mg/dL W	NA	NA	NA	Exclusion of patients with renal clearance < 30 mL/min

The first 291 patients were assigned to interrupted and non-interrupted warfarin in a non-randomised fashion and the last 212 patients were randomised to either uninterrupted warfarin or dabigatran.
AF, atrial fibrillation; CKD, chronic kidney disease; Creat, creatinine; D, dabigatran; LV, left ventricle; NA, not available; W, warfarin.