Nine-Year Trend of Anticoagulation Use, Thromboembolic Events, and Major Bleeding in Patients With Non-Valvular Atrial Fibrillation　– Shinken Database Analysis –

Abstract

Background: Trends of oral anticoagulant (OAC) prescription and incidence of thromboembolism (TE) and/or major bleeding (MB) in patients with non-valvular atrial fibrillation (NVAF) in Japan are still unclear.

Methods and Results: We used data from Shinken Database 2004–2012, which included all new patients attending the Cardiovascular Institute between June 2004 and March 2013. Of them, 2,434 patients were diagnosed with NVAF. Patients were divided into 3 time periods according to the year of initial visit: 2004–2006 (n=681), 2007–2009 (n=833), and 2010–2012 (n=920). OAC prescription rate steadily increased from 2004–2006 to 2010–2012. Between 2004–2006 and 2007–2009, irrespective of increased warfarin usage, MB tended to decrease, presumably due to low-intensity therapy and avoidance of concomitant use of dual antiplatelets, but TE did not improve. In 2010–2012, direct OACs (DOAC), preferred in low-risk patients, may have contributed to not only decrease TE, but also increase MB, especially extracranial bleeds. In high-risk patients in that time period, mostly treated with warfarin, incidence of TE and MB did not improve.

Conclusions: The 9-year trend of stroke prevention indicated a steady increase of OAC prescription and a partial improvement of TE and MB. Even in the era of DOAC, TE prevention was insufficient in high-risk patients, and DOAC were associated with increased extracranial bleeding.

The concept and tools of anticoagulation therapy for preventing thromboembolism in non-valvular atrial fibrillation (NVAF) have grossly changed in the last 10 years through the accumulation of evidence from numerous large clinical trials: continuing anticoagulation therapy has been recommended even under successful rhythm control with anti-arrhythmic drugs,^1,2 and the role of antiplatelet as an alternative to warfarin has been minimized.^3,4 Moreover, direct oral anticoagulants (DOAC) were approved in Japan in 2011,^5–8 which increases the anticoagulation therapy options for NVAF patients.

These changes may influence the usage conditions of anticoagulants and further change the outcomes of NVAF, including thromboembolism and major bleeding. In Western countries, several reports on the temporal trends of anticoagulation therapy and patient outcomes in AF have noted an increase in oral anticoagulant (OAC) prescription rate and decrease of thromboembolism incidence.^9,10 In Japan, however, reports on long-term trends have been scarce.

Shinken Database is a single hospital-based cohort that started in 2004 and has continually accumulated registration and follow-up data of patients with cardiovascular disease, including AF, for more than 10 years.^11,12 Using the data from this database with 9-year registration and 3-year follow-up, we report on the trends of anticoagulants usage and incidence of thromboembolism and major bleeding in NVAF patients.

Methods

Subjects

The Shinken Database, which was established in June 2004, contains data on all new patients attending the Cardiovascular Institute Hospital in Tokyo, Japan (abbreviated in Japanese as ‘Shinken’), excluding foreign travelers and patients with active cancer. The principal aim of this hospital-based database is the monitoring of the prevalence and prognosis of cardiovascular diseases in urban areas of Japan. The registry started in June 2004, and thereafter patients have been continually registered to the database annually.

The data used in this study were derived from the records of 19,994 new patients between June 2004 and March 2013 (Shinken Database 2004–2012). Of them, 2,434 patients were diagnosed with NVAF at initial visit, and we separated them into 3 time periods according to the year of the initial visit: 2004–2006 (n=681), 2007–2009 (n=833), and 2010–2012 (n=920). Prescription rate of oral anti-thrombotic medications, including warfarin, DOAC (dabigatran and rivaroxaban) and antiplatelets, and incidence of thromboembolism and major bleeding, were compared between the 3 time periods.

Patient Follow-up

The health status and the incidence of cardiovascular events and mortality were maintained in the database by being linked to the medical records of the hospital, and by study documents of prognosis sent once per year to those who stopped hospital visits or who were referred to other hospitals. The end of follow-up in the present study was 30 June 2015.

In the present study, we excluded data >3 years after the initial visit to avoid imbalance of follow-up period among patients due to the different registered years (2004–2012).

Data Collection at Initial Visit

After electrocardiogram (ECG) and chest X-ray, cardiovascular status was evaluated using echocardiography, exercise test, 24-h Holter recording and blood laboratory data from the initial visit. In addition to gender, age, height and weight, we collected data on cardiovascular diseases, including heart failure (New York Heart Association class ≥2), valvular heart disease (moderate or severe stenosis or regurgitation on echocardiography), coronary artery disease (diagnosed on angiography or scintigraphy), hypertrophic and dilated cardiomyopathy (diagnosed on echocardiography or magnetic resonance imaging [MRI]), and history of disabling cerebral infarction or transient ischemic attack (TIA; diagnosed on computed tomography or MRI). Cardiovascular risk factors were defined as follows: hypertension, use of antihypertensive agents, systolic blood pressure ≥140 mmHg, or diastolic blood pressure ≥90 mmHg; diabetes mellitus, use of oral hypoglycemic agents or insulin, or glycosylated hemoglobin ≥6.5%; dyslipidemia, use of statin or drugs for lowering triglyceride, low-density lipoprotein ≥140 mg/dl, high-density lipoprotein <40 mg/dl, or triglyceride ≥150 mg/dl; and, chronic kidney disease, estimated glomerular filtration rate (eGFR), <60 ml/min/1.73 m². Body mass index was calculated as weight in kilograms divided by height in meters squared. GFR was estimated using the new Japanese coefficient for the modified isotope dilution mass spectrometry-traceable 4-variable Modification of Diet in Renal Disease (MDRD) study equation (GFR=194×SCr^–1.004×Age^0.287×0.739 [if female]).¹³

Definition of AF

AF at the initial visit was diagnosed on 12-lead surface ECG and 24-h Holter recordings. It was also diagnosed on any medical history of AF from referring physicians.

Definition of Thromboembolism and Major Bleeding

Thromboembolism was defined as the composite of ischemic stroke, TIA, and systemic arterial thromboembolism requiring hospitalization. Major bleeding was defined as bleeding that required emergency hospitalization, as defined in a previous study,¹⁴ which includes intracranial and extracranial hemorrhage.

Thromboembolic and Bleeding Risk Stratification

CHADS2 score was determined based on the patient characteristics at the initial visit,¹⁵ according to the definitions given in the previous section. HAS-BLED score was also determined according to the 2010 ESC Guidelines for the management of AF,¹⁶ and the definition in the present study for “H” (hypertension) was systolic blood pressure ≥160 mmHg at initial visit, or taking >3 types of hypertensive agents from among calcium blocker, angiotensin-converting enzyme inhibitor, angiotensin II receptor blocker, β-blocker, thiazide diuretic, and α-blocker (given as “uncontrolled hypertension” in Table 1). Data on non-steroidal anti-inflammatory drugs were not available in the database.

Table 1. Baseline NVAF Patient Characteristics vs. Year of Initial Visit

Variable	2004–2006 (n=681)	2007–2009 (n=833)	2010–2012 (n=920)	P-value
Age (years)	63.6±12.4	62.6±12.7	63.1±13.1	0.323
Age ≥65 years	326 (47.9)	386 (46.3)	454 (49.3)	0.496
BMI (kg/m2)	23.7±3.4	24.1±3.4	23.9±3.5	0.083
Type of AF
Paroxysmal	399 (58.6)	497 (59.7)	593 (64.5)	0.032
Persistent/Permanent	282 (41.4)	336 (40.3)	327 (35.5)	0.032
Male	497 (73.0)	648 (77.8)	700 (76.1)	0.091
Hypertension	256 (37.6)	388 (46.6)	490 (53.3)	<0.001
Uncontrolled hypertension†	10 (1.5)	8 (1.0)	51 (5.5)	<0.001
Diabetes mellitus	93 (13.7)	113 (13.6)	124 (13.5)	0.995
Dyslipidemia	127 (18.6)	220 (26.4)	242 (26.3)	<0.001
Heart failure	48 (7.0)	94 (11.3)	54 (5.9)	<0.001
Mitral regurgitation	35 (5.1)	37 (4.4)	20 (2.2)	0.004
Aortic stenosis	12 (1.8)	15 (1.8)	5 (0.5)	0.034
Aortic regurgitation	16 (2.3)	16 (1.9)	34 (3.7)	0.058
Tricuspid regurgitation	17 (2.5)	37 (4.4)	23 (2.5)	0.034
Coronary artery disease	40 (5.9)	77 (9.2)	67 (7.3)	0.044
Hypertrophic cardiomyopathy	8 (1.2)	18 (2.2)	22 (2.4)	0.199
Dilated cardiomyopathy	12 (1.8)	25 (3.0)	20 (2.2)	0.260
Cerebral infarction or TIA	28 (4.1)	55 (6.6)	48 (5.2)	0.098
History of major bleeding	2 (0.3)	10 (1.2)	6 (0.7)	0.512
Chronic kidney disease	401 (58.9)	526 (63.1)	568 (61.7)	0.231
Abnormal renal function‡	18 (2.6)	34 (4.1)	19 (2.1)	0.372
Abnormal liver function§	16 (2.3)	13 (1.6)	17 (1.8)	0.516
CHADS2 score
0	315 (46.3)	333 (40.0)	332 (36.1)	0.002
1	210 (30.8)	275 (33.0)	349 (37.9)
2	112 (16.4)	137 (16.4)	158 (17.2)
3	34 (5.0)	55 (6.6)	50 (5.4)
4	6 (0.9)	24 (2.9)	22 (2.4)
5	4 (0.6)	7 (0.8)	9 (1.0)
6	0 (0.0)	2 (0.2)	0 (0.0)
≥2	156 (22.9)	225 (27.0)	239 (26.0)	0.172
HAS-BLED score
0	194 (28.5)	293 (35.2)	348 (37.8)	0.044
1	300 (44.1)	320 (38.4)	346 (37.6)
2	145 (21.3)	156 (18.7)	158 (17.2)
3	38 (5.6)	47 (5.6)	53 (5.8)
4	4 (0.6)	14 (1.7)	12 (1.3)
5	0 (0.0)	2 (0.2)	3 (0.3)
6	0 (0.0)	1 (0.1)	0 (0.0)
≥3	42 (6.2)	64 (7.7)	68 (7.4)	0.384

Data given as mean±SD or n (%). ^†Defined as systolic blood pressure≥160 mmHg at initial visit, or taking >3 types of hypertensive agents from among calcium blocker, angiotensin-converting enzyme inhibitor, angiotensin II receptor blocker, β-blocker, thiazide diuretic, and α-blocker. ^‡Defined as chronic dialysis or renal transplantation or serum creatinine ≥2.0 mg/dl. ^§Defined as chronic hepatic disease (eg, cirrhosis) or biochemical evidence of significant hepatic derangement (eg, bilirubin ≥2.0 mg/dl, and aspartate aminotransferase and/or alanine aminotransferase ≥100 IU/L). AF, atrial fibrillation; BMI, body mass index; NVAF, non-valvular AF; TIA, transient ischemic attack.

Measurement of Prothrombin Time

Prothrombin time international normalized ratio (PT-INR) was measured in patients on warfarin (reagent: HemosIL RecombiPlasTin 2G, Instrumentation Laboratory, USA; analyzer: Coagtron-180, Kyowa Hakko Kirin, Japan; normal range, 9.7–12.9 s). For patients with ≥2 times of PT-INR measurement between 3 months and 12 months after warfarin was started, time in the therapeutic range (TTR)¹⁷ was calculated using PT-INR between 3 months and 12 months, and the number of PT-INR measurements for each patient ranged from 3 to 10 because patients visited the outpatient clinic at intervals of 1–3 months. The target PT-INR range was set as 1.6–2.6 irrespective of age (≥70 and <70 years), based on the recent Japanese report showing no significant interaction between age and anticoagulation intensity in the effect on thromboembolism and major bleeding.¹⁸

Statistical Analysis

Statistical analysis was carried out using SPSS version 19.0 (SPSS, Chicago, IL, USA). Statistical significance was set at P<0.05. Patients were divided into 3 time periods according to the year of initial visit to hospital (2004–2006, 2007–2009, and 2010–2012).

First, patient background was compared between the 3 time periods; categorical and consecutive data are presented as number (%) and mean±SD, and the temporal difference in categorical and continuous variables between the 3 time periods was tested using Cochran-Armitage trend test and Jonckheere-Terpstra trend test, respectively.

The incidence rates of ischemic stroke and major bleeding were calculated using the person-year method, and the difference in time periods was tested using Cochran-Armitage test for trend. Also, the relative risks of the 2 later time periods (2007–2009 and 2010–2012) with regard to the earliest period (2004–2006) were calculated with 95% CI.

Ethics

The ethics committee at the Cardiovascular Institute granted ethics approval for this study and all the patients gave written informed consent.

Results

Subject Characteristics

Table 1 lists baseline subject clinical characteristics according to time period. Although average age and the prevalence of diabetes were similar between the 3 time periods, temporal increase was observed in hypertension (P<0.001) and tendency of increase was observed in the history of cerebral infarction or TIA (P=0.098). Accordingly, prevalence of CHADS2 score significantly increased (P=0.018), especially in the category of CHADS2 1 point (31.0% in 2004–2006 to 35.9% in 2010–2012). Similarly, the prevalence of HAS-BLED score increased with marginally statistical significance (P=0.048), but the proportion of patients with HAS-BLED ≥3 points did not temporally change (6.2% in 2004–2006 to 7.4% in 2010–2012; P=0.384).

Table 2 lists the incidence rates of procedures that can potentially affect the usage of anti-thrombotic medications, and which were carried out within 1 year after the initial visit for treatment of various cardiovascular diseases. Of the procedures, pulmonary vein isolation drastically increased from 2.1% in 2004–2006 to 12.0% and 21.7% in 2007–2009 and 2010–2012, respectively (P<0.001).

Table 2. Potential Affectors of Anti-Thrombotic Medication^†

Variable	2004–2006 (n=681)	2007–2009 (n=833)	2010–2012 (n=920)	P-value
Electrical cardioversion	42 (6.2)	58 (7.0)	71 (7.7)	0.485
Pulmonary vein isolation	14 (2.1)	100 (12.0)	200 (21.7)	<0.001
PCI	18 (2.6)	28 (3.4)	24 (2.6)	0.586
CABG	0 (0.0)	7 (0.8)	9 (1.0)	0.041

Data given as n (%). ^†Carried out within 1 year of initial visit. CABG, coronary artery bypass graft; PCI, percutaneous coronary intervention.

Anti-Thrombotic Medication Prescription Rate

The prescription rate of anticoagulants significantly increased from 40.7% in 2004–2006, to 47.5% and 55.9% in 2007–2009 and 2010–2012, respectively (P<0.001; Table 3). Given that use of DOAC started in 2011, it accounted for 25.5% of 55.9% of patients with anticoagulants in the last time period. The prescription rate increased according to increasing CHADS2 score and reached 64.4% in patients with CHADS2 ≥2 points in the last time period. DOAC were more prevalent in patients with CHADS2 0–1 point compared with CHADS2 ≥2 points. In contrast, although the prescription rate of anticoagulants increased with increasing HAS-BLED score, the rate tended to decrease with time, from 88.1% in 2004–2006 to 73.4% and 76.5% in 2007–2009 and 2010–2012, respectively (P=0.212).

Table 3. Anti-Thrombotic Therapy at Baseline

Variable	2004–2006	2007–2009	2010–2012	P-value
Total	n=681	n=833	n=920
Anticoagulants	277 (40.7)	396 (47.5)	514 (55.9)	<0.001
Warfarin	277 (40.7)	396 (47.5)	279 (30.3)	<0.001
TTR (%)	57.3±30.6	56.2±32.9	54.0±33.5	0.583
<60%	95 (14.0)	97 (11.6)	112 (12.2)	0.328
PT-INR at 3 months	1.68±0.57	1.61±0.49	1.68±0.57	0.326
PT-INR at 12 months	1.83±0.52	1.81±0.51	1.84±0.62	0.851
DOAC	0 (0.0)	0 (0.0)	235 (25.5)	<0.001
Antiplatelet	225 (33.0)	242 (29.1)	177 (19.2)	<0.001
Aspirin	220 (32.3)	232 (27.9)	172 (18.7)	<0.001
Thienopyridine derivatives	34 (5.0)	58 (7.0)	41 (4.5)	0.057
Dual antiplatelet therapy	29 (4.3)	52 (6.2)	36 (3.9)	0.055
Triple therapy	21 (3.1)	24 (2.9)	21 (2.3)	0.580
CHADS2=0 points	n=315	n=333	n=332
Anticoagulants	97 (30.8)	119 (35.7)	141 (42.5)	0.008
Warfarin	97 (30.8)	119 (35.7)	59 (17.8)	<0.001
TTR (%)	61.0±29.0	55.2±34.0	56.5±36.6	0.584
<60%	22 (7.6)	27 (8.5)	19 (5.8)	0.371
PT-INR at 3 months	1.67±0.54	1.50±0.42	1.54±0.52	0.131
PT-INR at 12 months	1.86±0.56	1.77±0.47	1.68±0.43	0.175
DOAC	0 (0.0)	0 (0.0)	82 (24.7)	<0.001
Antiplatelet	108 (34.3)	67 (20.1)	40 (12.0)	<0.001
Aspirin	108 (34.3)	67 (20.1)	38 (11.4)	<0.001
Thienopyridine derivatives	3 (1.0)	4 (1.2)	2 (0.6)	0.719
Dual antiplatelet therapy	3 (1.0)	4 (1.2)	0 (0.0)	0.153
Triple therapy	0 (0.0)	2 (0.6)	0 (0.0)	0.143
CHADS2=1 point	n=210	n=275	n=349
Anticoagulants	95 (45.2)	139 (50.5)	219 (62.8)	<0.001
Warfarin	95 (45.2)	139 (50.5)	112 (32.1)	<0.001
TTR (%)	58.4±28.6	60.6±33.4	54.7±34.1	0.539
<60%	37 (17.5)	25 (9.4)	32 (9.7)	0.010
PT-INR at 3 months	1.73±0.58	1.66±0.51	1.71±0.46	0.695
PT-INR at 12 months	1.73±0.43	1.78±0.50	1.79±0.48	0.708
DOAC	0 (0.0)	0 (0.0)	107 (30.7)	<0.001
Antiplatelet	57 (27.1)	74 (26.9)	65 (18.6)	0.019
Aspirin	55 (26.2)	73 (26.5)	64 (18.3)	0.024
Thienopyridine derivatives	9 (4.3)	6 (2.2)	10 (2.9)	0.397
Dual antiplatelet therapy	7 (3.3)	5 (1.8)	9 (2.6)	0.570
Triple therapy	6 (2.9)	4 (1.5)	3 (0.9)	0.179
CHADS2 ≥2 points	n=156	n=225	n=239
Anticoagulants	85 (54.5)	138 (61.3)	154 (64.4)	0.138
Warfarin	85 (54.5)	138 (61.3)	108 (45.2)	0.002
TTR (%)	51.6±34.2	52.9±31.4	52.3±31.7	0.971
<60%	36 (19.9)	45 (18.1)	61 (23.2)	0.332
PT-INR at 3 months	1.63±0.58	1.64±0.52	1.72±0.66	0.571
PT-INR at 12 months	1.89±0.54	1.85±0.54	1.94±0.76	0.664
DOAC	0 (0.0)	0 (0.0)	46 (19.2)	<0.001
Antiplatelet	60 (38.5)	101 (44.9)	72 (30.1)	0.004
Aspirin	57 (36.5)	92 (40.9)	70 (29.3)	0.031
Thienopyridine derivatives	22 (14.1)	48 (21.3)	29 (12.1)	0.020
Dual antiplatelet therapy	19 (12.2)	43 (19.1)	27 (11.3)	0.038
Triple therapy	15 (9.6)	18 (8.0)	18 (7.5)	0.753
HAS-BLED 0–1 point	n=494	n=613	n=694
Anticoagulants	157 (31.8)	247 (40.3)	348 (50.1)	<0.001
Warfarin	157 (31.8)	247 (40.3)	147 (21.2)	<0.001
TTR (%)	71.0±25.2	72.4±26.9	71.8±26.4	0.937
<60%	20 (4.0)	20 (3.3)	18 (2.6)	0.162
PT-INR at 3 months	1.84±0.59	1.66±0.44	1.75±0.47	0.080
PT-INR at 12 months	1.89±0.49	1.86±0.45	1.89±0.47	0.902
DOAC	0 (0.0)	0 (0.0)	201 (29.0)	<0.001
Antiplatelet	100 (20.2)	86 (14.0)	39 (5.6)	<0.001
Aspirin	99 (20.0)	85 (13.9)	38 (5.5)	<0.001
Thienopyridine derivatives	7 (1.4)	8 (1.3)	2 (0.3)	0.037
Dual antiplatelet therapy	6 (1.2)	7 (1.1)	1 (0.1)	0.029
Triple therapy	2 (0.4)	2 (0.3)	0 (0.0)	0.139
HAS-BLED=2 points	n=145	n=156	n=158
Anticoagulants	83 (57.2)	102 (65.4)	114 (72.2)	0.007
Warfarin	83 (57.2)	102 (65.4)	85 (53.8)	0.510
TTR (%)	49.5±31.4	46.4±33.4	46.4±32.3	0.806
<60%	47 (32.4)	41 (26.3)	55 (34.8)	0.624
PT-INR at 3 months	1.60±0.55	1.49±0.40	1.65±0.64	0.222
PT-INR at 12 months	1.80±0.54	1.77±0.52	1.83±0.69	0.862
DOAC	0 (0.0)	0 (0.0)	29 (18.4)	<0.001
Antiplatelet	87 (60.0)	102 (65.4)	82 (51.9)	0.139
Aspirin	83 (57.2)	97 (62.2)	78 (49.4)	0.154
Thienopyridine derivatives	11 (7.6)	29 (18.6)	19 (12.0)	0.278
Dual antiplatelet therapy	7 (4.8)	26 (16.7)	15 (9.5)	0.213
Triple therapy	3 (2.1)	11 (7.1)	7 (4.4)	0.105
HAS-BLED ≥3 points	n=42	n=64	n=68
Anticoagulants	37 (88.1)	47 (73.4)	52 (76.5)	0.212
Warfarin	37 (88.1)	47 (73.4)	47 (69.1)	0.032
TTR (%)	37.1±25.6	38.3±28.0	34.1±31.5	0.780
<60%	28 (66.7)	36 (56.3)	39 (57.4)	0.387
PT-INR at 3 months	1.42±0.41	1.67±0.68	1.61±0.61	0.160
PT-INR at 12 months	1.70±0.52	1.74±0.60	1.76±0.72	0.926
DOAC	0 (0.0)	0 (0.0)	5 (7.4)	0.014
Antiplatelet	38 (90.5)	54 (84.4)	56 (82.4)	0.264
Aspirin	38 (90.5)	50 (78.1)	56 (82.4)	0.367
Thienopyridine derivatives	16 (38.1)	21 (32.8)	20 (29.4)	0.352
Dual antiplatelet therapy	16 (38.1)	19 (29.7)	20 (29.4)	0.380
Triple therapy	16 (38.1)	11 (17.2)	14 (20.6)	0.043

Data given as n (%) or mean±SD. DOAC, direct oral anticoagulants; PT-INR, prothrombin time international normalized ratio; TTR, time in therapeutic range. Other abbreviations as in Table 1.

PT-INR at 3 months and 12 months after warfarin was started was constantly at a lower level of the therapeutic range in all time periods (average PT-INR, approximately 1.6 and 1.8, respectively). Average TTR ranged from 54.0% to 57.3% in the 3 time periods, showing no significant time period-dependent differences. Similar findings were also observed for the different CHADS2 and HAS-BLED score levels.

The prescription rate of antiplatelet significantly decreased from 34.3% in 2004–2006, to 29.1% and 19.2% in 2007–2009 and 2010–2012, respectively (P<0.001), and the decrease was especially obvious for CHADS2 0 point (34.3%, 20.1%, and 11.4%, respectively; P<0.001), whereas for CHADS2 ≥2 points, the prescription of antiplatelet increased in 2007–2009 and decreased thereafter in 2010–2012 (36.5%, 40.9%, and 29.3%, respectively; P=0.031). Similarly, the prescription rate of antiplatelet significantly decreased in HAS-BLED 0–1 point (20.2%, 14.0%, and 5.6% in 2004–2006, 2007–2009, and 2010–2012, respectively; P<0.001).

In HAS-BLED 2 and ≥3 points, the prescription rate of antiplatelet was extremely high and did not significantly decrease: 60.0%/80.3%, 65.4%/77.6%, and 51.9%/73.6% in 2004–2006, 2007–2009, and 2010–2012 (P=0.139/0.328), respectively. For HAS-BLED ≥3 points, the rate of dual antiplatelet therapy was constantly high (38.1%, 29.7%, and 29.4%, respectively; P=0.380), but that of triple therapy significantly decreased with time (38.1%, 17.2%, and 20.6%, respectively; P=0.043).

Thromboembolism and Major Bleeding

The incidence rates of thromboembolism were 12.6 (95% CI: 8.3–19.1) and 12.0 (95% CI: 8.0–18.1) per 1,000 person-years in 2004–2006 and 2007–2009, respectively, which decreased to 6.5 (95% CI: 3.5–12.0) per 1,000 person-years in 2010–2012 (Tables 4,5). The relative risks of 2007–2009 and 2010–2012 compared with 2004–2006 were 0.95 (95% CI: 0.53–1.71) and 0.52 (95% CI: 0.24–1.09), respectively. Ischemic stroke comprised 20 of 21, 21 of 23, and 8 of 10 thromboembolism events in 2004–2006, 2007–2009, and 2010–2012, respectively, accounting for a similar percentage over time (80–90%). The proportion of thromboembolisms without OAC decreased from 71.4% in 2004–2006 to 60.9% and 50.0% in 2007–2009 and 2010–2012, respectively.

Table 4. Incidence of Thromboembolism and Major Bleeding vs. Time Period

Time period	(A) 2004–2006				(B) 2007–2009				(C) 2010–2012				P for trend	Relative risk
	No. events	Person- years	Incidence rate†	95% CI	No. events	Person- years	Incidence rate†	95% CI	No. events	Person- years	Incidence rate†	95% CI		(B) vs. (A)	95% CI	(C) vs. (A)	95% CI
Total
Thromboembolism	21	1,751	12.0	7.8–18.3	23	1,912	12.0	8.0–18.1	10	1,531	6.5	3.5–12.0	0.207	1.00	0.55–1.81	0.54	0.26–1.15
Ischemic stroke	20	1,751	11.4	7.4–17.6	21	1,916	11.0	7.2–16.8	8	1,531	5.2	2.6–10.3
TIA	2	1,779	1.1	0.3–4.1	2	1,933	1.0	0.3–3.8	1	1,540	0.6	0.1–3.7
Systemic embolism	0	1,780	0.0	0.0–2.2	0	1,936	0.0	0.0–2.0	1	1,540	0.6	0.1–3.7
Major bleeding	10	1,763	5.7	3.1–10.4	3	1,931	1.6	0.5–4.6	10	1,530	6.5	3.5–12.0	0.055	0.28	0.08–1.02	1.14	0.47–2.74
Intracranial hemorrhage	6	1,771	3.4	1.6–7.4	2	1,934	1.0	0.3–3.8	6	1,531	3.9	1.8–8.5
Extracranial hemorrhage	5	1,771	2.8	1.2–6.6	1	1,934	0.5	0.1–2.9	4	1,540	2.6	1.0–6.7
CHADS2=0 points
Thromboembolism	3	759	4.0	1.3–11.6	9	749	12.0	6.3–22.9	0	514	0.0	0.0–7.5	0.016	3.00	0.81–11.08	–	–
Ischemic stroke	2	760	2.6	0.7–9.6	7	752	9.3	4.5–19.2	0	514	0.0	0.0–7.5
TIA	1	763	1.3	0.2–7.4	2	755	2.7	0.7–9.7	0	514	0.0	0.0–7.5
Systemic embolism	0	763	0.0	0.0–5.0	0	758	0.0	0.0–5.1	0	514	0.0	0.0–7.5
Major bleeding	1	763	1.3	0.2–7.4	0	758	0.0	0.0–5.1	1	513	1.9	0.3–11.0	0.518	–	–	1.46	0.09–23.37
Intracranial hemorrhage	0	763	0.0	0.0–5.0	0	758	0.0	0.0–5.1	1	513	1.9	0.3–11.0
Extracranial hemorrhage	1	763	1.3	0.2–7.4	0	758	0.0	0.0–5.1	0	514	0.0	0.0–7.5
CHADS2=1 point
Thromboembolism	8	539	14.8	7.5–29.3	2	603	3.3	0.9–12.1	2	581	3.4	0.9–12.6	0.030	0.22	0.05–1.05	0.23	0.05–1.08
Ischemic stroke	8	539	14.8	7.5–29.3	2	603	3.3	0.9–12.1	2	581	3.4	0.9–12.6
TIA	1	554	1.8	0.3–10.2	0	606	0.0	0.0–6.3	0	584	0.0	0.0–6.6
Systemic embolism	0	554	0.0	0.0–6.9	0	606	0.0	0.0–6.3	0	584	0.0	0.0–6.6
Major bleeding	2	550	3.6	1.0–13.3	0	606	0.0	0.0–6.3	1	584	1.7	0.3–9.7	0.330	–	–	0.47	0.04–5.21
Intracranial hemorrhage	2	550	3.6	1.0–13.3	0	606	0.0	0.0–6.3	0	584	0.0	0.0–6.6
Extracranial hemorrhage	0	554	0.0	0.0–6.9	0	606	0.0	0.0–6.3	1	584	1.7	0.3–9.7
CHADS2 ≥2 points
Thromboembolism	10	452	22.1	12.0–40.7	12	560	21.4	12.3–37.4	8	435	18.4	9.3–36.3	0.917	0.97	0.42–2.24	0.83	0.33–2.11
Ischemic stroke	10	452	22.1	12.0–40.7	12	560	21.4	12.3–37.4	6	436	13.8	6.3–30.0
TIA	0	463	0.0	0.0–8.3	0	573	0.0	0.0–6.7	1	442	2.3	0.4–12.8
Systemic embolism	0	463	0.0	0.0–8.3	0	573	0.0	0.0–6.7	1	442	2.3	0.4–12.8
Major bleeding	7	450	15.6	7.5–32.1	3	568	5.3	1.8–15.5	8	434	18.4	9.3–36.4	0.135	0.34	0.09–1.31	1.18	0.43–3.25
Intracranial hemorrhage	3	458	6.5	2.2–19.2	2	571	3.5	1.0–12.8	5	434	11.5	4.9–26.9
Extracranial hemorrhage	5	454	11.0	4.7–25.8	1	570	1.8	0.3–9.9	3	443	6.8	2.3–19.9
HAS-BLED=0–1 point
Thromboembolism	11	1,281	8.6	4.8–15.4	15	1,395	10.8	6.5–17.7	6	1,095	5.5	2.5–12.0	0.362	1.26	0.58–2.73	0.64	0.24–1.73
Ischemic stroke	10	1,281	7.8	4.2–14.4	13	1,399	9.3	5.4–15.9	5	1,095	4.6	1.9–10.7
TIA	2	1,294	1.5	0.4–5.6	2	1,412	1.4	0.4–5.2	1	1,102	0.9	0.2–5.1
Systemic embolism	0	1,295	0.0	0.0–3.0	0	1,415	0.0	0.0–2.7	0	1,102	0.0	0.0–3.5
Major bleeding	2	1,292	1.5	0.4–5.6	0	1,415	0.0	0.0–2.7	5	1,100	4.5	1.9–10.6	0.029	–	–	3.00	0.58–15.46
Intracranial hemorrhage	1	1,294	0.8	0.1–4.4	0	1,415	0.0	0.0–2.7	1	1,101	0.9	0.2–5.1
Extracranial hemorrhage	2	1,292	1.5	0.4–5.6	0	1,415	0.0	0.0–2.7	4	1,103	3.6	1.4–9.3
HAS-BLED=2 points
Thromboembolism	5	382	13.1	5.6–30.6	4	377	10.6	4.1–27.3	1	312	3.2	0.6–18.2	0.384	0.81	0.22–3.01	0.24	0.03–2.09
Ischemic stroke	5	382	13.1	5.6–30.6	4	377	10.6	4.1–27.3	1	312	3.2	0.6–18.2
TIA	0	393	0.0	0.0–9.8	0	378	0.0	0.0–10.2	0	313	0.0	0.0–12.3
Systemic embolism	0	393	0.0	0.0–9.8	0	378	0.0	0.0–10.2	0	313	0.0	0.0–12.3
Major bleeding	7	379	18.5	9.0–38.2	3	373	8.0	2.7–23.6	3	309	9.7	3.3–28.6	0.383	0.43	0.11–1.67	0.52	0.14–2.03
Intracranial hemorrhage	4	385	10.4	4.0–26.7	2	376	5.3	1.5–19.4	3	309	9.7	3.3–28.6
Extracranial hemorrhage	3	387	7.8	2.6–22.8	1	375	2.7	0.5–15.1	0	313	0.0	0.0–12.3
HAS-BLED ≥3 points
Thromboembolism	5	88	56.9	24.3–133.1	4	140	28.6	11.1–73.6	3	124	24.3	8.2–71.3	0.391	0.50	0.13–1.87	0.43	0.10–1.79
Ischemic stroke	5	88	56.9	24.3–133.1	4	140	28.6	11.1–73.6	2	124	16.1	4.4–58.6
TIA	0	92	0.0	0.0–41.6	0	144	0.0	0.0–26.8	0	125	0.0	0.0–30.6
Systemic embolism	0	92	0.0	0.0–41.6	0	144	0.0	0.0–26.8	1	125	8.0	1.4–45.4
Major bleeding	1	92	10.9	1.9–61.5	0	144	0.0	0.0–26.8	2	121	16.5	4.5–60.1	0.328	–	–	1.51	0.14–16.69
Intracranial hemorrhage	0	92	0.0	0.0–41.6	0	144	0.0	0.0–26.8	2	121	16.5	4.5–60.1
Extracranial hemorrhage	1	92	10.9	1.9–61.5	0	144	0.0	0.0–26.8	0	125	0.0	0.0–30.6

^†Per 1,000 person-years. Abbreviations as in Table 1.

Table 5. Thromboembolism, Major Bleeding, and Anticoagulants

Time period	(A) 2004–2006				(B) 2007–2009				(C) 2010–2012
	Total	NonOAC	Warfarin	DOAC	Total	NonOAC	Warfarin	DOAC	Total	NonOAC	Warfarin	DOAC
Total
Thromboembolism	21 (100.0)	15 (71.4)	6 (28.6)	–	23 (100.0)	14 (60.9)	9 (39.1)	–	10 (100.0)	5 (50.0)	3 (30.0)	2 (20.0)
Ischemic stroke	18 (100.0)	13 (72.2)	5 (28.8)	–	21 (100.0)	16 (76.2)	5 (23.8)	–	8 (100.0)	5 (62.5)	2 (25.0)	1 (12.5)
TIA	2 (100.0)	1 (50.0)	1 (50.0)	–	2 (100.0)	0 (0.0)	2 (100.0)	–	1 (100.0)	0 (0.0)	0 (0.0)	1 (100.0)
Systemic embolism	0 (–)	0 (–)	0 (–)	–	0 (–)	0 (–)	0 (–)	–	1 (100.0)	0 (0.0)	1 (100)	0 (0.0)
Major bleeding	10 (100.0)	3 (30.0)	7 (70.0)	–	3 (100.0)	0 (0.0)	3 (100.0)	–	10 (100.0)	1 (10.0)	4 (40.0)	5 (50.0)
Intracranial hemorrhage	6 (100.0)	2 (33.3)	4 (66.7)	–	2 (100.0)	0 (0.0)	2 (100.0)	–	6 (100.0)	1 (16.7)	3 (50.0)	2 (33.3)
Extracranial hemorrhage	5 (100.0)	1 (20.0)	4 (80.0)	–	1 (100.0)	0 (0.0)	1 (100.0)	–	4 (100.0)	0 (0.0)	1 (25.0)	3 (75.0)

Data given as n (%). OAC, oral anticoagulant. Other abbreviations as in Tables 1,3.

The incidence rate of major bleeding was 5.7 (95% CI: 3.1–10.4) in 2004–2006, which decreased to 1.6 (95% CI: 0.5–4.6) in 2007–2009, and increased again to 6.5 (95% CI: 1.0–6.7) in 2010–2012. The relative risks of 2007–2009 and 2010–2012 with regard to 2004–2006 were 0.18 (95% CI: 0.02–1.53) and 0.93 (95% CI: 0.25–3.46), respectively. Intracranial hemorrhage comprised 6 of 10, 2 of 3, and 6 of 10 major bleeds in 2004–2006, 2007–2009, and 2010–2012, respectively, accounting for a similar percentage over time (60–70%). Major bleeds in patients with OAC accounted for 70%, 100%, and 90% of the total in 2004–2006, 2007–2009, and 2010–2012, respectively, and that in patients with DOAC accounted for 50% in 2010–2012.

Discussion

Major Findings

In the present study, we have described the 9-year trend of anti-thrombotic therapy and the incidence rates of thromboembolism and major bleeding using a single-center cohort from a cardiovascular hospital. The prescription rate of OAC increased from 40.7% in 2004–2006 to 47.5% and 55.9% in 2007–2009 and 2010–2012, respectively (P<0.001), and DOAC accounted for 25.5% in the last time period. Although no significant change was observed in thromboembolism or in major bleeding, tendency of reduction in the former and increase in the latter was observed in 2010–2012, presumably due to the use of DOAC.

Anti-Thrombotic Medication Prescription Rate

In Japan, several multicenter registries have reported the prescription rate of anti-thrombotic medications in different time periods. In the Hokkaido Study, a multicenter community-based registry in 1995 (NVAF, n=2,173), the prescription rate of warfarin was only 8%.¹⁹ In the Fushimi AF registry, however, a recent multicenter community-based registry started in 2011 (NVAF, n=3,282), the prescription rate of warfarin increased to 53%.²⁰ A similar increase in prescription rate can be observed in various cardiovascular hospital multicenter registries. Although the prescription rate of warfarin was 52% in a multicenter cohort by Inoue et al in 1999 (NVAF, n=509),²¹ it increased to 87% in the largest nationwide multicenter cohort of the J-RHYTHM registry in 2010 (AF, n=7,937).²² Therefore, it is understandable that, in the present cohort, the prescription rate of anticoagulants significantly increased over time, reflecting the overall trend in Japan.

There are several reasons for the increase in anticoagulant prescription rate, as reported in several countries.^9,10,23,24 The main reason may be the AFFIRM and RACE study, which demonstrated the limitation of rhythm control, especially the concept of “complete rhythm control can prevent ischemic stroke without anticoagulation”,^25,26 and re-acknowledged the significance of continuing anticoagulation therapy even for paroxysmal and/or asymptomatic AF patients.^1,2 Also, similar results have been confirmed in different populations.^27,28 Furthermore, in Japan, a randomized controlled trial by Sato et al showed that aspirin cannot decrease ischemic stroke in low-risk NVAF patients,³ which influenced, at least in part, the revised 2008 guideline by the Japanese Circulation Society to delete antiplatelet therapy from the recommended options for prevention of thromboembolism in NVAF patients and to recommend warfarin as the only option.²⁹ These are the possible reasons for the increase in warfarin usage for stroke prevention in NVAF patients in Japan and also in the present cohort between 2004–2006 and 2007–2009.

Since DOAC were approved and launched, an improvement in the underuse of anticoagulant therapy has been expected because DOAC are much more accessible compared with warfarin. In the present patients, the prescription rate of anticoagulants increased from 47.5% in 2007–2009 to 55.9% in 2010–2012. The change was observed in patients with low-moderate risk (CHADS2 0–1 point), for whom the prescription rate increased from 35.7%/50.5% (CHADS2 0/1 point, respectively) in 2007–2009 to 42.5%/62.8% in 2010–2012. Notably, DOAC accounted for half of the total anticoagulants in the low-moderate risk categories in 2010–2012. One of the reasons for the increase in OAC usage even in the patients with CHADS2 0 point may be due to the remarkable increase in catheter ablation for NVAF, which increased from 2.1% in 2004–2006 to 21.7% in 2010–2012 in the present patients. Of 200 patients who underwent catheter ablation in 2010–2012, 74 and 103 patients had CHADS2 0 and 1, respectively (data not shown). Given that, currently, catheter ablation, which requires periprocedural anticoagulation therapy, is indicated mainly for patients with neither severe left atrial enlargement, severe left ventricular dysfunction, nor severe pulmonary disease,^30,31 the patients undergoing catheter ablation inevitably tend to have low CHADS2 score,³² and thus, the association between increased catheter ablation for NVAF and increasing anticoagulation use for low CHADS2 score is understandable. In contrast, in patients with high risk (CHADS2 ≥2 points) the increase in prescription rate was limited: 61.3% in 2007–2009 to 64.4% in 2010–2012. In 2010–2012, the prescription rate of DOAC was 19.2%, accounting for only one-third of the total anticoagulants. A possible explanation for the limited increase of DOAC usage in high-risk patients may be the blue letter (safety alarm) on dabigatran issued by the Ministry of Health, Labour and Welfare in Japan in August 2011, which reported 5 fatal cases of severe bleeding.³³

As a whole, usage of antiplatelet gradually decreased from 33.0% in 2004–2006 to 29.1% and 19.2% in 2007–2009 and 2010–2012, respectively (P<0.001). Especially in patients with low risk (CHADS2 0 point), the prescription rate of aspirin drastically decreased from 34.3% in 2004–2006 to 20.1% and 12.0% in 2007–2009 and 2010–2012, respectively. It seems that the trend simply reflects the recognition of the evidence demonstrating the limited prophylactic effect on thromboembolism³ and the consequent revision of the Japanese guidelines.²⁹

Change in Thromboembolism and Major Bleeding

From 2004–2006 to 2007–2009 In the present study, although the incidence rate of thromboembolism did not change significantly (from 12.6 per 1,000 person-years in 2004–2006 to 12.0 per 1,000 person-years in 2007–2009), that of major bleeding tended to decrease from 5.7 per 1,000 person-years to 1.6 per 1,000 person-years, respectively. This trend is not easy to understand because the prescription rate increased and the intensity of anticoagulation did not change between the 2 time periods. When we stratified the patients according to HAS-BLED score, however, the prescription rate of anticoagulants reduced by approximately 10% for high HAS-BLED score (≥3 points) between 2004–2006 and 2007–2009, which may be associated with the approximately 21% decrease of triple therapy. In 2004–2006, all of the 16 patients with dual antiplatelet therapy and HAS-BLED ≥3 points received anticoagulants (consequently had triple therapy), whereas in 2007–2009, only half of 19 patients with this in the similar risk category received anticoagulants. Therefore, it can be assumed that the decision to avoid anticoagulant usage under dual antiplatelet therapy (avoidance of triple therapy) may, at least in part, have contributed to the decrease in major bleeding from 2004–2006 to 2007–2009.

Although the incidence rate of thromboembolism did not change between 2004–2006 and 2007–2009 in the whole patient group, it tended to increase in patients with CHADS2 0 point from 4.0 to 12.0 per 1,000 person-years. Among the 12 patients with CHADS2 0 point who had thromboembolism in the present study, 6 were aged in the range 65–74 years, and 1 of 2 in 2004–2006 and 3 of 4 in 2007–2009, respectively, were not anticoagulated (data not shown). Although the age range 65–74 years had been recognized as a thromboembolic risk in the Japanese guidelines,^29,34 such recognition may have weakened in the period 2007–2009, when the use of CHADS2 score, in which age ≥75 years is a thromboembolic risk, was widespread.

From 2007–2009 to 2010–2012 Compared with the trend between 2004–2006 and 2007–2009, the trend was different between 2007–2009 and 2010–2012: the incidence rate of thromboembolism tended to decrease from 12.0 to 6.5 per 1,000 person-years, whereas that of major bleeding tended to increase from 1.6 to 6.5 per 1,000 person-years. When stratified according to CHADS2 risk, the decrease of thromboembolism was obvious in CHADS2 0 point, but not so evident in CHADS2 1 and ≥2 points. The possible reasons for the decrease in thromboembolism for CHADS2 0 point may be due to the recent changes in the concept and tools. With regard to the concept, the recent emerging of CHA2DS2-VASc score, which additionally regards age 65–74 years as a thromboembolic risk,³⁵ may play a significant role, through which the significance of the age category may have been recognized again. In the present study, no thromboembolism occurred in patients with CHADS2 0 point and aged 65–74 years in 2010–2012 (data not shown), presumably reflecting the change in concept. With regard to the tools, it is assumed that DOAC enabled the active administration of anticoagulation therapy for relatively low-risk patients. In the net clinical benefit, in which the total balance of the reduction of thromboembolic and bleeding risk is evaluated,^36,37 DOAC including dabigatran and rivaroxaban were found to have a significant advantage, even in the low-risk NVAF patients,³⁷ whereas warfarin was found to have no advantage.³⁶ Additionally, due to the recognition that DOAC, including dabigatran and rivaroxaban, had a disadvantage in safety in relatively high-risk patients,^38–42 they were therefore selected for relatively low-risk patients.³³ In contrast, no obvious improvement was observed in high-risk patients (CHADS2 ≥2 points), irrespective of the increase in the prescription rate of anticoagulants. Low intensity of warfarin may be a possible reason, and using DOAC with a good safety profile even in high-risk patients⁴³ may reduce the incidence of thromboembolism, as was the case with the low-risk patients.

An increasing tendency of major bleeding was observed from 2007–2009 to 2010–2012, and the increase was extremely large for CHADS2 ≥2 points between the 2 time periods (from 5.3 to 18.4 per 1,000 person-years). Interestingly, when stratified according to HAS-BLED risk, tendency of increase in major bleeding was observed not only in the high HAS-BLED category (≥3 points: from 0.0 to 16.5 per 1,000 person-years), but also in the low HAS-BLED category (0–1 point: from 0.0 to 4.5 per 1,000 person-years). Major bleeding in the low HAS-BLED score category in 2010–2012 mainly consisted of extracranial bleeding (Table 4), which was mainly due to DOAC (Table 5). Notably, among 9 cases of major bleeding in 2010–2012 under oral anticoagulation therapy, 4 of 4 in the warfarin group and 4 of 5 in the DOAC group occurred in patients aged ≥65 years (data not shown), indicating the significance of age as a bleeding risk under anticoagulation therapy. The increase in major bleeding, especially extracranial bleeding due to the use of DOAC, is the remaining task to be solved in the era of DOAC.

Study Limitations

The present data derive from a single-center database from a cardiovascular hospital. Therefore, caution is necessary when comparing the results between the present patients and other cohorts with different indications. Although the present prescription rates were lower than that in multicenter registries from cardiovascular hospitals at similar time periods,^21,22 the multicenter registries may have recruited patients who had been visiting in a stable condition for a certain period of time (ie, >1 year), which may exclude patients with a worse status and/or reluctant to undergo treatment including anticoagulation therapy. Therefore, we believe that the present data represent the real-world status of anticoagulation therapy as derived from a cardiovascular hospital without strict inclusion criteria.

Conclusions

The present study has reported on the trends in anticoagulation therapy for Japanese NVAF patients in a cardiovascular hospital during a 9-year time period, including the use of warfarin and DOAC (including dabigatran and rivaroxaban). According to the 9-year trend, use of anticoagulation steadily increased, and low-intensity warfarin with avoidance of concomitant use of dual antiplatelet seems effectively to have decreased major bleeding, but yielded no improvement in thromboembolism, especially in high-risk patients. DOAC including dabigatran and rivaroxaban were preferred in low-risk patients, which contributed, at least in part, to decrease the thromboembolism, but which may be associated with the increase in major bleeding, especially of extracranial bleeding. Thus, the present data regarding the 9-year trend of OAC for NVAF patients indicate that, even in the era of DOAC, insufficient improvement in the incidence of thromboembolism in high-risk patients and increase of extracranial bleeding due to DOAC are the remaining tasks to be solved.

Acknowledgments

We thank Shiro Ueda and Nobuko Ueda at Medical Edge for assembling the database using the Clinical Study Supporting System and Ineko Hayakawa, Hiroaki Arai, and Hirokazu Aoki for data management and system administration.

Conflicts of Interest

S.S. received research funding and remuneration from Boehringer Ingelheim. T.Y. received research funding from Boehringer Ingelheim and Daiichi-Sankyo, and remuneration from Boehringer Ingelheim, Daiichi-Sankyo, Bayer Healthcare, Pfizer, Bristol-Myers Squibb, Eisai and Ono Pharmaceutical.

Funding

This study is partially supported by the Practical Research Project for Life-Style related Diseases including Cardiovascular Diseases and Diabetes Mellitus from Japan Agency for Medical Research and Development, AMED (15656344).

References

• 1.    Colonna  P,  Sorino  M,  de Luca  L,  Bovenzi  F,  de Luca  I. Anti-thrombotic therapy in atrial fibrillation: Beyond the AFFIRM study. J Cardiovasc Med (Hagerstown) 2006; 7: 505–513.
• 2.    Flaker  GC,  Belew  K,  Beckman  K,  Vidaillet  H,  Kron  J,  Safford  R, et al. Asymptomatic atrial fibrillation: Demographic features and prognostic information from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Am Heart J 2005; 149: 657–663.
• 3.    Sato  H,  Ishikawa  K,  Kitabatake  A,  Ogawa  S,  Maruyama  Y,  Yokota  Y, et al. Low-dose aspirin for prevention of stroke in low-risk patients with atrial fibrillation: Japan Atrial Fibrillation Stroke Trial. Stroke 2006; 37: 447–451.
• 4.   Active Writing Group of the ACTIVE Investigators,  Connolly  S,  Pogue  J,  Hart  R,  Pfeffer  M,  Hohnloser  S,  Chrolavicius  S, et al. Clopidogrel plus aspirin versus oral anticoagulation for atrial fibrillation in the Atrial fibrillation Clopidogrel Trial with Irbesartan for prevention of Vascular Events (ACTIVE W): A randomised controlled trial. Lancet 2006; 367: 1903–1912.
• 5.    Connolly  SJ,  Ezekowitz  MD,  Yusuf  S,  Eikelboom  J,  Oldgren  J,  Parekh  A, et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009; 361: 1139–1151.
• 6.    Patel  MR,  Mahaffey  KW,  Garg  J,  Pan  G,  Singer  DE,  Hacke  W, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med 2011; 365: 883–891.
• 7.    Granger  CB,  Alexander  JH,  McMurray  JJ,  Lopes  RD,  Hylek  EM,  Hanna  M, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2011; 365: 981–992.
• 8.    Giugliano  RP,  Ruff  CT,  Braunwald  E,  Murphy  SA,  Wiviott  SD,  Halperin  JL, et al. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2013; 369: 2093–2104.
• 9.    Lakshminarayan  K,  Solid  CA,  Collins  AJ,  Anderson  DC,  Herzog  CA. Atrial fibrillation and stroke in the general medicare population: A 10-year perspective (1992 to 2002). Stroke 2006; 37: 1969–1974.
• 10.    Shroff  G,  Solid  C,  Herzog  C. Temporal trends in ischemic stroke and anticoagulation therapy among Medicare patients with atrial fibrillation: A 15-year perspective (1992–2007). JAMA Intern Med 2013; 173: 159–160.
• 11.    Suzuki  S,  Yamashita  T,  Ohtsuka  T,  Sagara  K,  Uejima  T,  Oikawa  Y, et al. Prevalence and prognosis of patients with atrial fibrillation in Japan: A prospective cohort of Shinken Database 2004. Circ J 2008; 72: 914–920.
• 12.    Suzuki  S,  Yamashita  T,  Otsuka  T,  Sagara  K,  Uejima  T,  Oikawa  Y, et al. Recent mortality of Japanese patients with atrial fibrillation in an urban city of Tokyo. J Cardiol 2011; 58: 116–123.
• 13.    Matsuo  S,  Imai  E,  Horio  M,  Yasuda  Y,  Tomita  K,  Nitta  K, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 2009; 53: 982–992.
• 14.    Suzuki  S,  Yamashita  T,  Kato  T,  Fujino  T,  Sagara  K,  Sawada  H, et al. Incidence of major bleeding complication of warfarin therapy in Japanese patients with atrial fibrillation. Circ J 2007; 71: 761–765.
• 15.    Gage  BF,  Waterman  AD,  Shannon  W,  Boechler  M,  Rich  MW,  Radford  MJ. Validation of clinical classification schemes for predicting stroke: Results from the National Registry of Atrial Fibrillation. JAMA 2001; 285: 2864–2870.
• 16.   European Heart Rhythm Association; European Association for Cardio-Thoracic Surgery,  Camm  AJ,  Kirchhof  P,  Lip  GY,  Schotten  U,  Savelieva  I,  Ernst  S, et al. Guidelines for the management of atrial fibrillation: The Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J 2010; 31: 2369–2429.
• 17.    Rosendaal  FR,  Cannegieter  SC,  van der Meer  FJ,  Briet  E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993; 69: 236–239.
• 18.    Yamashita  T,  Inoue  H,  Okumura  K,  Atarashi  H,  Origasa  H; J-Rhythm Registry Investigators. Warfarin anticoagulation intensity in Japanese nonvalvular atrial fibrillation patients: A J-RHYTHM Registry analysis. J Cardiol 2015; 65: 175–177.
• 19.    Tomita  F,  Kohya  T,  Sakurai  M,  Kaji  T,  Yokoshiki  H,  Sato  M, et al. Prevalence and clinical characteristics of patients with atrial fibrillation: Analysis of 20,000 cases in Japan. Jpn Circ J 2000; 64: 653–658.
• 20.    Akao  M,  Chun  YH,  Esato  M,  Abe  M,  Tsuji  H,  Wada  H, et al. Inappropriate use of oral anticoagulants for patients with atrial fibrillation. Circ J 2014; 78: 2166–2172.
• 21.    Inoue  H,  Nozawa  T,  Hirai  T,  Iwasa  A,  Okumura  K,  Lee  JD, et al. Accumulation of risk factors increases risk of thromboembolic events in patients with nonvalvular atrial fibrillation. Circ J 2006; 70: 651–656.
• 22.    Atarashi  H,  Inoue  H,  Okumura  K,  Yamashita  T,  Kumagai  N,  Origasa  H, et al. Present status of anticoagulation treatment in Japanese patients with atrial fibrillation: A report from the J-RHYTHM Registry. Circ J 2011; 75: 1328–1333.
• 23.    Robson  J,  Dostal  I,  Mathur  R,  Sohanpal  R,  Hull  S,  Antoniou  S, et al. Improving anticoagulation in atrial fibrillation: Observational study in three primary care trusts. Br J Gen Pract 2014; 64: e275–e281, doi:10.3399/bjgp14X679705.
• 24.    Weitz  JI,  Semchuk  W,  Turpie  AG,  Fisher  WD,  Kong  C,  Ciaccia  A, et al. Trends in prescribing oral anticoagulants in Canada, 2008–2014. Clin Ther 2015; 37: 2506–2514.e4, doi:10.1016/j.clinthera.2015.1009.1008.
• 25.    Wyse  DG,  Waldo  AL,  DiMarco  JP,  Domanski  MJ,  Rosenberg  Y,  Schron  EB, et al. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002; 347: 1825–1833.
• 26.    Van Gelder  IC,  Hagens  VE,  Bosker  HA,  Kingma  JH,  Kamp  O,  Kingma  T, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med 2002; 347: 1834–1840.
• 27.    Roy  D,  Talajic  M,  Nattel  S,  Wyse  DG,  Dorian  P,  Lee  KL, et al. Rhythm control versus rate control for atrial fibrillation and heart failure. N Engl J Med 2008; 358: 2667–2677.
• 28.    Ogawa  S,  Yamashita  T,  Yamazaki  T,  Aizawa  Y,  Atarashi  H,  Inoue  H, et al. Optimal treatment strategy for patients with paroxysmal atrial fibrillation: J-RHYTHM Study. Circ J 2009; 73: 242–248.
• 29.   JCS Joint Working Group. Guidelines for pharmacotherapy of atrial fibrillation (JCS 2008): Digest version. Circ J 2010; 74: 2479–2500.
• 30.   JCS Joint Working Group. Guidelines for non-pharmacotherapy of cardiac arrhythmias (JCS 2011): Digest version. Circ J 2013; 77: 249–274.
• 31.   JCS Joint Working Group. Guidelines for clinical cardiac electrophysiologic studies (JCS 2011): Digest version. Circ J 2013; 77: 497–518.
• 32.    Hrynkiewicz-Szymanska  A,  Dluzniewski  M,  Platek  AE,  Szymanski  FM,  Syska-Suminska  J,  Klos-Szadryn  A, et al. Association of the CHADS2 and CHA2DS2-VASc scores with left atrial enlargement: A prospective cohort study of unselected atrial fibrillation patients. J Thromb Thrombolysis 2015; 40: 240–247.
• 33.    Suzuki  S,  Sagara  K,  Otsuka  T,  Kano  H,  Matsuno  S,  Takai  H, et al. “Blue letter effects”: Changes in physicians’ attitudes toward dabigatran after a safety advisory in a specialized hospital for cardiovascular care in Japan. J Cardiol 2013; 62: 366–373.
• 34.   JCS Joint Working Group. Guidelines for pharmacotherapy of atrial fibrillation (JCS 2013): Digest version. Circ J 2014; 78: 1997–2021.
• 35.    Lip  GY,  Nieuwlaat  R,  Pisters  R,  Lane  DA,  Crijns  HJ. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: The Euro Heart Survey on atrial fibrillation. Chest 2010; 137: 263–272.
• 36.    Singer  DE,  Chang  Y,  Fang  MC,  Borowsky  LH,  Pomernacki  NK,  Udaltsova  N, et al. The net clinical benefit of warfarin anticoagulation in atrial fibrillation. Ann Intern Med 2009; 151: 297–305.
• 37.    Banerjee  A,  Lane  DA,  Torp-Pedersen  C,  Lip  GY. Net clinical benefit of new oral anticoagulants (dabigatran, rivaroxaban, apixaban) versus no treatment in a ‘real world’ atrial fibrillation population: A modelling analysis based on a nationwide cohort study. Thromb Haemost 2012; 107: 584–589.
• 38.    Abraham  NS,  Singh  S,  Alexander  GC,  Heien  H,  Haas  LR,  Crown  W, et al. Comparative risk of gastrointestinal bleeding with dabigatran, rivaroxaban, and warfarin: Population based cohort study. BMJ 2015; 350: h1857.
• 39.    Oldgren  J,  Alings  M,  Darius  H,  Diener  HC,  Eikelboom  J,  Ezekowitz  MD, et al. Risks for stroke, bleeding, and death in patients with atrial fibrillation receiving dabigatran or warfarin in relation to the CHADS2 score: A subgroup analysis of the RE-LY trial. Ann Intern Med 2011; 155: 660–667, W204.
• 40.    Hijazi  Z,  Hohnloser  SH,  Oldgren  J,  Andersson  U,  Connolly  SJ,  Eikelboom  JW, et al. Efficacy and safety of dabigatran compared with warfarin in relation to baseline renal function in patients with atrial fibrillation: A RE-LY (Randomized Evaluation of Long-term Anticoagulation Therapy) trial analysis. Circulation 2014; 129: 961–970.
• 41.    Hori  M,  Matsumoto  M,  Tanahashi  N,  Momomura  S,  Uchiyama  S,  Goto  S, et al. Rivaroxaban vs. warfarin in Japanese patients with non-valvular atrial fibrillation in relation to age. Circ J 2014; 78: 1349–1356.
• 42.    Hori  M,  Fukaya  T,  Kleine  E,  Reilly  PA,  Ezekowitz  MD,  Connolly  SJ, et al. Efficacy and safety of dabigatran etexilate vs. warfarin in Asian RE-LY patients according to baseline renal function or CHADS₂ score. Circ J 2015; 79: 2138–2147.
• 43.    Lopes  RD,  Al-Khatib  SM,  Wallentin  L,  Yang  H,  Ansell  J,  Bahit  MC, et al. Efficacy and safety of apixaban compared with warfarin according to patient risk of stroke and of bleeding in atrial fibrillation: A secondary analysis of a randomised controlled trial. Lancet 2012; 380: 1749–1758.