. 2021 Jun 14;23(10):1511–1527. doi: 10.1093/europace/euab092

Table 1.

Summary of the characteristics of the included studies

First author (year), country	Study cohort	Cohort description	Selection criteria	Length of follow-up, mean ± SD or median (IQR)	Outcomes
Prospective
Domek (2020), Middle East¹⁵	Gulf Survey of Atrial Fibrillation Events (SAFE) Registry	Consecutive patients admitted to ED in 23 hospitals in 6 Middle East countries independently from the primary reason for admission, 603,^a 63.42 ± 11.75,^b 315 (52.2%),^c not reported,^d 3.69 ± 1.58,^e 1.56 ± 1.07^f	Inclusion criteria: ≥ 18 years old, >30 s AF on 12-lead resting ECG, diabetes	12 months	Primary: ACM, composite: stroke/systemic embolism, ACM, CV hospitalization
Gumprecht (2020), Middle East¹⁶	Gulf Survey of Atrial Fibrillation Events (SAFE) Registry	Consecutive patients admitted to ED in 23 hospitals in 6 Middle East countries independently from the primary reason for admission, 2021,^a 56.74 ± 16.47,^b 968 (47.9%),^c not reported,^d 2.34 ± 1.78,^e 1.13 ± 1.065^f	Inclusion criteria: ≥18 years old, >30 s AF on 12-lead resting electrocardiogram. Exclusion criteria: insufficient data for calculating CHA₂DS₂-VASc score	1 year	Primary: ACM, composite of ischaemic stroke or systemic embolism/all-cause mortality and CV hospitalization
Guo (2020) 1 year, China¹⁸	mAFA II trial	2 arm cluster-RCT. Clusters were 40 Chinese hospitals, 3324,^a mAFA: 67.0 ± 15.0 UC: 70.0 ± 12.0,^b mAFA: 625 (38.0%) UC: 637 (38.0%),^c not reported,^d mAFA: 3 (2–4) UC: 3 (2–4),^e mAFA: 1 (1–2) UC: 1 (1–2)^f	Inclusion criteria: ≥18 years old, AF confirmed by ECG or 24-h Holter, CHA₂DS₂-VASc ≥2. Exclusion criteria: mechanical prosthetic value or moderate/severe mitral stenosis, unable to provide informed consent, unable to be followed up for 1 year for any reason	12 months	Primary: composite: stroke/thromboembolism, ACM, and re-hospitalization
Guo (2020) extension, China¹⁷	mAFA II trial	2 arm cluster-RCT. Clusters were 40 Chinese hospitals, 2473,^a mAFA: 67.8 ± 15.4 UC: 70.1 ± 12.0,^b mAFA: 430 (34.1%) UC: 511 (42.1%),^c not reported,^d mAFA: 3 (2–4) UC: 3 (2–4),^e mAFA: 2 (1–3) UC: 2 (1–3)^f	Inclusion criteria: ≥18 years old, AF confirmed by ECG or 24-h Holter, CHA₂DS₂-VASc ≥2, Over 1 year of follow-up. Exclusion criteria: mechanical prosthetic value or moderate/severe mitral stenosis, unable to provide informed consent	mAFA: 687 ± 191; 701 (489–841) days, usual care: 514 ± 167; 546 (394–632) days	Primary: composite: stroke/thromboembolism, ACM, and re-hospitalization. Secondary: ischaemic stroke, other thromboembolism, intracranial bleeding, extracranial bleeding, recurrent AF or AF symptom, heart failure, ACM
Koziel (2020), Balkans¹²	BALKAN-AF survey	Consecutive patients managed in hospitals and outpatient settings; 8 Balkan countries; 49 centres; 14-week observational survey recorded prospectively, 2712,^a ABC: 49 (41, 57) non-ABC: 64 (55, 71),^b ABC: 485 (47.9%) non-ABC: 557 (42.9%),^c not reported,^d ABC: 3.4 ± 1.8 non-ABC: 3.4 ± 1.9,^e ABC: 1.94 ± 1.2 non-ABC: 1.99 ± 1.2^f	Inclusion criteria: ≥18 years old. Exclusion criteria: prosthetic mechanical heart valves, moderate or severe mitral valve stenosis or any significant heart valve disease with indications for surgical treatment	None	Primary: ABC adherence
Retrospective—post hoc
Proietti (2018, 2020), USA and Canada²¹^,²²	AFFIRM	Retrospective analysis of RCT comparing rate vs. rhythm control and OAC.; 200 sites in USA and Canada, 3169,^a 70 (65–76),^b 1237 (39.0%),^c NR,^d 3 (2–4),^e not reported^f	Inclusion criteria: on VKA—warfarin, documented AF within last 6 weeks, aged ≥65 years, or <65 years with ≥1 risk factor for stroke, AF episodes in last 6 months totalling ≥6 h, unless cardioversion within 6 h, continuous AF <6 months, unless SR restored and maintained ≥24 h, eligible for rate and rhythm control, eligible for ≥2 AADs (or 2 dose levels of amiodarone) and ≥2 rate-control drugs	3.7 (2.8–4.6)	Primary: ACM, composite: stroke/major bleeding/CV mortality, hospitalization. Secondary: stroke, major bleeding, CV mortality, CV hospitalization, recurrent hospitalization, total hospitalizations, length of stay for first hospitalization, total length of stay
Pastori (2019), Italy¹⁹	ATHERO-AF	Single-centre cohort study in Rome, February 2008 to December 2016; Retrospective analysis on prospective observational study, 882,^a 73.1 ± 8.5,^b 40.8%,^c not reported,^d 3.50 ± 1.5,^e not reported^f	Inclusion criteria: ≥18 years old, AF, all patients on warfarin after risk stratification: CHA₂DS₂-VASc for men/women: 0/1—maybe aspirin but no OAC, 1/2 maybe aspirin but preferably OAC, 2+/3+ OAC. Exclusion criteria: prosthetic heart valves or severe valvulopathies, severe cognitive impairment, chronic infections (HIV, hepatitis B or C), systemic autoimmune disease, active cancer, liver insufficiency (e.g. cirrhosis)	36.9 (20.0–57.5) months	Primary: CV events
Retrospective—Registry or Electronic health records
Yoon (2019), South Korea¹⁴	Korea National Health Insurance Service database	National cohort; data from 2005 to 2015; retrospective analysis, 204842,^a ABC: 52.9 ± 12.2 non-ABC: 64.9 ± 10.8,^b ABC: 10129 (32.0%) non-ABC: 66778 (38.6%),^c not reported,^d ABC: 0.91 ± 1.39 non-ABC: 2.97 ± 1.80,^e not reported^f	Inclusion criteria: adult, non-valvular AF, baseline health check-up data within the year before enrolment, AF outpatient clinic visit during the follow-up period	6.2 ± 3.5 years	Primary: ACM, ischaemic stroke, major bleeding, myocardial infarction, composite of other 4 outcomes
Proietti (2020) ESC-EHRA, Europe²⁰	ESC-EORP Atrial Fibrillation General Long-Term Registry	Multicentre observational registry held by the ESC and endorsed by the European Heart Rhythm Association (EHRA), 9663,^a ABC: 70 (61–76) non-ABC: 69 (61–76),^b ABC: 741 (37.1%), non-ABC: 1926 (41.4%),^c not reported,^c ABC: 2.68 ± 1.57; 3 (2–4) non-ABC: 3.07 ± 1.90; 3 (2–4),^e ABC: 1.58 ± 1.12; 2 (1–2) non-ABC: 1.26 ± 0.93; 1 (1–2)^f	Inclusion criteria: ≥ 18 years old, AF documented within 12 months before enrolment based on objective electrocardiographic evaluation	12 months	Primary: composite: TE, ACS, CV mortality, CV mortality, ACM, Stroke, Any TE, bleeding events, ICH, any readmission, any AF readmission, any CV readmission, ACS
Yang (2020) dementia, South Korea²³	Korea National Health Insurance Service database	National cohort; data from 2005 to 2015, 228026,^a ABC: 68.8 ± 10.2 non-ABC: 69.7 ± 11.6,^b ABC: 18016 (39.2%) non-ABC: 70218 (38.6%),^c not reported,^d ABC: 0 (0–1) non-ABC: 2 (1–3),^e ABC: 0 (0–1) non-ABC: 2 (1–3)^f	Inclusion criteria: ≥18 years old, non-valvular AF, have baseline health check-up data within the year before enrolment. Exclusion criteria: patients who had an ischaemic stroke, patients with a history of dementia, patients with an ischaemic stroke during the follow-up period	6.0 (3.3–9.5) years	Primary: dementia. Secondary: Alzheimer’s disease, vascular dementia
Yang (2020) frailty, South Korea¹³	Korea National Health Insurance Service database	National cohort; data from 2005 to 2015, 262 987,^a ABC: 50 (41, 58) non-ABC: 65 (56,72),^b ABC: 39.4%, non-ABC: 38.6%,^c not reported,^d ABC: 0 (0–1), non-ABC: 2 (1–3),^e ABC: 0 (0–1), non-ABC: 2 (1–3)^f	Inclusion criteria: ≥18 years old, non-valvular AF, Have baseline health check-up data within the year before enrolment. Exclusion criteria: patients who had an ischaemic stroke	5.9 (3.2, 9.4)	Primary: ACM, ischaemic stroke, heart failure admission, myocardial infarction, major bleeding, composite of other 5 outcomes

AADs, anti-arrhythmic drugs; ABC, Atrial Fibrillation Better Care; ACM, all-cause mortality; ACS, acute coronary syndrome; AF, atrial fibrillation; CV, cardiovascular; ED, emergency department; EHRA, European Heart Rhythm Association; ESC, European Society of Cardiology; ICH, intra-cranial haemorrhage; mAFA, mobile AF-App; RCT, randomized controlled trial; TE, thromboembolism; UC, usual care.

Mean ± SD or median (IQR) age.

N (%) female.

Ethnicity.

Mean ± SD or median (IQR) CHA₂DS₂-VASc score.

Mean ± SD or median (IQR) HAS-BLED score.