If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Contemporary trend of reduced-dose non-vitamin K anticoagulants in Japanese patients with atrial fibrillation: A cross-sectional analysis of a multicenter outpatient registry
In Japan, 66.8% atrial fibrillation patients were prescribed non-vitamin K oral anticoagulants (NOACs), well beyond the global average.
•
About 15% of patients received NOAC with a non-standardized dose reduction (NSDR).
•
Patients with CHA2DS2-VASc score ≥2 had the highest proportion of NSDRs.
•
The physician's apprehension for bleeding should be balanced with stroke risk.
Abstract
Background
Non-vitamin K antagonist oral anticoagulants (NOACs) have been widely used to prevent stroke in non-valvular atrial fibrillation (NVAF) patients. Stringent monitoring is not required for NOACs, albeit dose adjustments are needed based on specific patient factors, such as renal function, body weight and age, or concomitant medications. We investigated the NOAC dosing patterns and evaluated the predictors of the non-standardized dose reduction (NSDR).
Methods
A total of 2452 newly diagnosed NVAF patients were consecutively recruited from secondary- and tertiary-care hospitals between 2012 and 2017. The NOAC doses were classified as one of three: (1) full dose; (2) standardized dose reduction (SDR); or (3) NSDR, consistent with Japanese package inserts.
Results
Overall, 66.8% (N = 1637) of the NVAF patients (median age: 69 years, interquartile range [IQR]: 60–76; 70% male; median CHA2DS2-VASc score of 2, IQR: 1–3) received NOACs. NOAC use dramatically increased during the study period (51.2% in 2012–13 to 74.4% in 2016–17). The percentages of SDR and NSDR were 19.6% and 14.4%, respectively; a proportion of SDR and NSDR did not alter drastically. Older age, concomitant antiplatelet therapy, impaired renal function, and prior heart failure or left ventricular dysfunction were independently associated with NSDR. Of note, patients with a high risk (CHA2DS2-VASc score ≥2) had the highest proportion of NSDRs.
Conclusions
Nearly half of the NOAC dose reductions in our registry were deemed “non-standardized,” which were seen mostly in patients at significant risk for ischemic stroke. The physician's apprehension regarding excessive bleeding under NOAC use should be appropriately balanced with concern for an increased risk of embolic events.
]. Although dose-adjusted warfarin was the preferred therapy for many years, non-vitamin K antagonist oral anticoagulants (NOACs) have been developed to prevent thromboembolism in patients with non-valvular AF (NVAF). This class of agents has been demonstrated to be at least as safe and effective as dose-adjusted warfarin [
Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials.
Non-vitamin K antagonist oral anticoagulants have better efficacy and equivalent safety compared to warfarin in elderly patients with atrial fibrillation: a systematic review and meta-analysis.
Predictive factors for bleeding during treatment with rivaroxaban and warfarin in Japanese patients with atrial fibrillation – subgroup analysis of J-ROCKET AF.
]. Based on randomized controlled trials on NVAF, each NOAC has been approved at a specific dose with recommended adjustments that are based on selected patient factors, such as renal function, body weight and age, or concomitant medications.
Several real-world studies and large administrative database analyses have confirmed the low incidence of ischemic stroke and major bleeding previously observed in clinical trials [
Bleeding risk of apixaban, dabigatran, and low-dose rivaroxaban compared with warfarin in Japanese patients with non-valvular atrial fibrillation: a propensity matched analysis of administrative claims data.
]. However, as their use expands into the general population, whether the dose recommendations are adhered to (for example, with respect to appropriateness) in common practice remains a major concern, particularly in non-Western countries [
Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study.
]. Japan has the highest proportion of elderly individuals in the population in the world. NOACs have been prescribed widely in up to 70% of these patients [
]. The objective of this study was to investigate the patterns of NOAC dosing and evaluate the risk of reducing the non-standardized dose, using a multicenter, outpatient registry in Japan.
Methods
Ethics
The institutional review board at each participating hospital approved the study's protocol, and all participants provided written informed consent. Before the launch of the Keio Inter-hospital Cardiovascular Studies for AF (KiCS-AF) registry, information regarding the objectives of the study and its social significance were provided to the University Hospital Medical Information Network, who recognized KiCS-AF as an “acceptable registry,” for clinical trial registration (UMIN 000022229).
Data sources
We obtained data from the KiCS-AF multicenter registry from September 2012 to July 2017. The KiCS-AF registry is a prospective, multicenter registry that was designed to collect clinical variables and outcome data of consecutive patients newly diagnosed with AF or referred to an outpatient clinic at each participating hospital. Dedicated clinical research coordinators are assigned to each hospital, and data on approximately 150 variables (described below) are collected for each patient. The KiCS-AF registry ensures data traceability by tracking the staff who approve the data and data-entry personnel at the participating institutions. It also validates data consistency via inspections of the participating institutions. Additionally, the database administrators provide on- and off-site training systems to instruct the clinical research coordinators how to input data consistently. Furthermore, administrative office personnel answer all inquiries regarding data entry.
Clinical assessment
To recruit treatment-naïve patients, we enrolled only those with a diagnostic code for AF within the previous 6 months. Data regarding the patients’ background, symptoms, prior and current drug use [including oral anticoagulants (OACs)], electrocardiography and echocardiography results, and blood sampling test results were collected from medical records. The patients’ background information included parameters for calculating the risk scores for stroke {congestive heart failure or left ventricular dysfunction, hypertension, age ≥75 years, diabetes, stroke including vascular disease, age 65–74 years, and sex category (female) according to the CHA2DS2-VASc], and bleeding events [hypertension (uncontrolled or systolic blood pressure > 160 mmHg), renal impairment (dialysis, transplant, or creatinine > 2.26 mg/dL), liver impairment (aspartate aminotransferase, alanine aminotransferase more than threefold the upper limit), stroke history, prior major bleeding or predisposition to bleeding, labile international normalized ratio, elderly (> 65 years), drugs/alcohol concomitantly according to the HAS-BLED score [
]. OACs included warfarin and NOACs (dabigatran, rivaroxaban, apixaban, and edoxaban).
For the present analysis, all NVAF patients who were consecutively registered in the KiCS-AF registry (n = 2452) were included. We excluded 37 patients (2.3%) prescribed a standard-dose NOAC who had a renal indication for dose reduction (potential overdosing or a patient with chronic kidney disease who was on hemodialysis), four patients (0.2%) with missing data on the NOAC dose, and 68 patients (4.2%) with missing information on renal function (Fig. 1). The included patients were categorized into three groups: (1) the warfarin group, defined as patients who were prescribed warfarin at the time of registration; (2) the NOAC group, defined as patients who were prescribed any one of the NOACs at the time of registration; and (3) patients without OACs (Fig. 1). Further, within the NOAC group, patients were divided into three sub-groups according to the NOAC dose at registration: (1) full dose: prescription of the maximum dose according to Japanese package inserts (i.e. dabigatran: 300 mg/day; rivaroxaban: 15 mg/day; apixaban: 10 mg/day; and edoxaban: 60 mg/day), (2) standardized dose reduction (SDR): prescription of a reduced dose according to the standard criteria, and (3) non-standardized dose reduction (NSDR): a reduced dose that was not prescribed according to the standard criteria (Fig. 1).
Fig. 1Study flow chart. AF, atrial fibrillation; OAC, oral anticoagulant; NOAC, non-vitamin K oral anticoagulant.
The criteria for the approved dose reduction for each NOAC in Japan were as follows: dabigatran 220 mg/day for patients with creatinine clearance (CrCl) 30–50 ml/min, or concomitant use of P-glycoprotein inhibitors; rivaroxaban 10 mg/day for patients with CrCl 15–49 ml/min; apixaban 5 mg/day for patients with more than two of the following: age ≥80 years, body weight ≤60 kg, or serum creatinine ≥1.5 mg/dl; edoxaban 30 mg/day for patients with body weight ≤60 kg, CrCl ≥15–50 ml/min or concomitant use of P-glycoprotein inhibitors. In this study, we defined patients with 220 mg/day of dabigatran regardless of calculated CrCl as SDR, and with 150 mg/day or less of dabigatran as NSDR (Table 1) [
We evaluated the extent of OAC use as reported in the KiCS-AF registry and the frequency of SDR and NSDR in each subgroup within the NOAC group. These descriptive analyses were applied to pre-specified sub-groups for which the embolic risk was classified using the CHA2DS2-VASc score (low: 0; intermediate: 1; high: ≥2) [
]. Then, we examined temporal trends in the proportion of the OAC and NOAC dose categories (full dose, SDR, and NSDR). Finally, within the NOAC sub-groups, we compared baseline characteristics, including demographics, general medical history, cardiovascular history, vital signs, echocardiographic data, and laboratory studies. In addition, we investigated the independent predictors of the use of NOACs with NSDR.
Continuous variables are presented as a median and interquartile range (IQR); categorical variables are presented as numbers and percentages. Group differences were evaluated using the chi-squared test for categorical variables and the Wilcoxon rank-sum test for continuous variables. To analyze trends, we used Mantel-Haenszel's chi-square test of linear association for categorical variables and a linear regression analysis for continuous variables. After excluding the SDR group, we performed multivariable logistic regression analyses to estimate the adjusted odds ratios and 95% confidence intervals for the likelihoods of NSDR among the NOAC sub-groups, after adjusting for, body mass index, chronic kidney disease [estimated glomerular filtration rate (eGFR) <60 ml/min], prior major bleeding, concomitant antiplatelet therapy, and components of the CHA2DS2-VASc score [i.e. prior history of congestive heart failure or left ventricular dysfunction (left ventricular ejection fraction < 40%), hypertension, age ≥65 years, diabetes, stroke including vascular disease, and sex category (female)]. These variables were selected based on previous studies [
] addressing a similar issue and with a similar clinical perspective. All p-values were two-sided with a significance threshold of p < 0.01 for group differences and the trend analysis (a lower-than-usual p-value was selected to correct for the inflation of type I errors because of repeated testing for a large number of variables). The data were analyzed using SPSS, version 23.0 (IBM Corp., Armonk, NY, USA).
Results
Overall use of OACs and the prevalence of NSDR
We identified 2452 patients in our registry. The median age was 69 years (IQR: 60 to 76 years), 70% of the patients were male, and the median CHA2DS2-VASc score was 2 (IQR: 1 to 3) in the entire cohort. Of the study population 1637 patients (66.8%) were prescribed NOACs and 384 were administered warfarin (15.7%; Fig. 2). Among patients who were treated with NOACs, 249 were prescribed dabigatran (16.3%), 577 were prescribed rivaroxaban (37.8%), 596 were prescribed apixaban (39.0%), and 106 were prescribed edoxaban (6.9%).
Fig. 2Overall proportions of OACs and each type of NOACs. OAC, oral anticoagulant; NOAC, non-vitamin K oral anticoagulant.
SDR was observed in 19.6% of the patients and NSDR was observed in 14.4%. The frequency of SDR and NSDR in each NOAC group is presented in Fig. 3. The percentage of NSDR was the lowest in patients who were administered dabigatran, followed by those who were administered rivaroxaban, edoxaban, and apixaban, (4.4%, 12.8%, 17.0%, and 19.6%, respectively). In the subgroup analyses, according to the risk of embolism (as determined with the CHA2DS2-VASc score), patients with a high risk of stroke (i.e. CHA2DS2-VASc score ≥2) were more likely to be treated with NSDR than were low- and intermediate-risk patients (17.7% for high-risk, 3.1% for low-risk, and 8.7% for intermediate-risk patients, Fig. 4).
Fig. 3Percentages of full doses, SDR, and NSDR among patients on NOACs. NOAC, non-vitamin K oral anticoagulant; SDR, standardized dose reduction; NSDR, non-standardized dose reduction.
Fig. 4Distributions of NOAC doses according to the CHA2DS2-VASc score. NOAC, non-vitamin K oral anticoagulant; SDR, standardized dose reduction; NSDR, non-standardized dose reduction.
Fig. 5 shows the trend in the use of OACs and NOAC dose. During the study period, the use of NOACs increased (from 51.2% in 2012–13 to 74.4% in 2016–17, p < 0.001); conversely, the use of warfarin decreased (from 25.6% in 2012–13 to 11.5% in 2016–17, p < 0.001). Among patients who were prescribed NOACs during the study period, the proportion of patients who were prescribed a full dose gradually decreased (from 69.0% in 2012–13 to 62.7% in 2016–17) and NSDR increased (from 8.6% in 2012–13 to 16.0% in 2016–17). Yet, these differences were not statistically significant (p > 0.01 for all).
Fig. 5Overall trend in the use of OACs and NOACs. OAC, oral anticoagulant; NOAC, non-vitamin K oral anticoagulant; SDR, standardized dose reduction; NSDR, non-standardized dose reduction.
Baseline characteristics of the patients who were prescribed NOACs
The baseline characteristics of the patients who were administered NOACs, stratified by dose, are shown in Table 2. The median age was the highest in the SDR group, followed by the NSDR, and full-dose group (79 years in the SDR group vs. 75 years in the SDR group vs. 66 years in the full-dose group, p < 0.001). Conversely the proportion of men was the lowest in the SDR group, followed by the NSDR, and full-dose group (46.0% in the SDR group vs. 67.3% in the NSDR group vs. 77.3% in the full-dose group, p < 0.001).
Table 2Baseline characteristics of patients who were prescribed non-vitamin K oral anticoagulants.
Furthermore, patients in the SDR group were more likely to have a history of hypertension, heart failure, stroke, and chronic kidney disease (eGFR <60 ml/min), than those in the SDR or full-dose group (p < 0.01 for all). Importantly, reflecting more complex backgrounds, the median CHA2DS2-VASc score was the highest in the SDR group, followed by the NSDR, and full-dose group (p < 0.001).
Factors associated with NSDR among patients with NOACs
The results of the multivariable analysis are shown in Fig. 6. In the NOAC group, older age (≥65 years), concomitant antiplatelet therapy, chronic kidney disease (eGFR <60 ml/min), and prior heart failure or left ventricular dysfunction were independently associated with the risk of NSDR. There were no significant associations between NSDR and a history of major bleeding or lower body mass index (≤18.5).
Fig. 6Independent predictors that were associated with reduction of the non-standardized dose among patients who were prescribed non-vitamin K oral anticoagulants. Adjusted ORs (point estimate) and 95% CIs (error bars) indicate the likelihood of non-standardized dose reduction in a logistic regression analysis. ORs <1 indicate decreased odds of non-standardized dose reduction. Covariates: age ≥65 years, concomitant antiplatelet therapy, chronic kidney disease (eGFR <60 ml/min), prior heart failure or LVEF <40%, female sex (vs male), hypertension, prior history of stroke, prior major bleeding, diabetes mellitus, BMI ≤18.5, vascular disease. OR, odds ratio; CI, confidence interval; LVEF, left ventricular ejection fraction; eGFR, estimated glomerular filtration rate; NSDR, non-standardized dose reduction; BMI, body mass index. *: eGFR <60 ml/min.
We used a contemporary outpatient dataset to investigate the patterns of NOAC dosing and evaluate the risk of reducing the non-standardized dose and found that there is a high rate of NOAC use in Japan, well beyond the global average [
]. Approximately 15% of patients received an NSDR, representing nearly half of the dose-reduced NOAC prescriptions. Our time-series analysis showed that the use of NOACs further increased after an initial plateau. More importantly, the subgroup analysis, which was based on the patient's stroke risk, showed that the incidence of NSDR increased in patients with a high risk of stroke, suggesting that this is a potentially serious clinical issue. NSDR was also a concern in older patients (≥65 years), those with impaired renal function, concomitant antiplatelet therapy, and prior heart failure or left ventricular dysfunction who required close monitoring owing to their increased susceptibility to both systemic embolism and bleeding events.
The rate of NOAC use was 5% in the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) study in the USA [
]; however, the rate was 48% in the Global Registry on Long-Term Oral Anti-thrombotic Treatment in Patients with Atrial Fibrillation study, which included China and the Middle East [
]. Compared with these other regions, Japan shows a high use of NOACs. This may be attributable to the following: (1) prescriptions are covered by the National Health Insurance system in Japan [
]; (2) doctors may prescribe medications using their discretion, especially in the case of NOACs, as there are no set restrictions; and (3) a growing body of data suggests that NOACs may be preferred for East Asians [
], 36% to 59% of patients experienced NSDR. However, these figures were compiled when warfarin was used more frequently. As NOAC use has become more prevalent, reducing the dose has become a global problem. In a registry study in Denmark [
Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study.
], the incidence of cerebral infarction and hemorrhage that was associated with a reduced dose was equal to that with the use of warfarin; thus, the advantages of using a NOAC are minimal. The recent ORBIT-AF analysis showed that the rate of NSDR was only 9.4%; thus, the dose had no effect on the incidence of cerebral infarction or hemorrhage [
Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study.
]. Overall, NSDR is a relatively widespread problem that requires a global solution.
Regarding the rationale for reducing the dose, our results were in line with those of previous studies, despite the differences in the baseline characteristics between people in Western and non-Western countries. Steinberg et al. [
] examined the possibility of assessing renal dysfunction based on the calculation of the CrCl level and stated that the concurrent use of other drugs may prompt physicians to prescribe reduced doses. Furthermore, Nielsen et al. [
Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study.
] suggested that the rationale for reduced NOAC doses is determined not only by an understanding of the pharmacological effects of the drug but also by the physician's and patient's perceptions about embolism and hemorrhage. Although an increased risk of hemorrhage is unavoidable when adequate anticoagulation is used, inadequate anticoagulation is also clearly unsafe.
In this study, the NSDR group included younger men with normal renal function and a higher body mass index than those in the SDR group. This finding indicates that physicians in Japan should avoid NSDR when prescribing NOACs to patients with normal renal function, but NSDR may be appropriate in individuals with a normal habitus. Although the package inserts for NOACs in Japan stipulate that reducing the dose should be based on the CrCl level, the estimated CrCl level has a wide margin of error. The eGFR is frequently used to evaluate renal function; the results of this study support previous findings that a low eGFR is a risk factor for underdosing [
]; in the present study, NSDR was prevalent, reflecting Japanese preferences. However, it has also been reported that physicians are insufficiently aware that patients are more concerned about cerebral infarction than they are about hemorrhage [
]; this suggests the importance of having a detailed discussion with the patient when considering anticoagulation therapy. Additionally, because NSDR may have a significant effect on the patient's prognosis, it must be used with great caution.
Limitations
The present study had several limitations. First, the study had the inherent limitations of its nonrandomized, observational study design, and unmeasured confounders among the associated variables may have contributed to the results, even after rigorous measurement and recording of key clinical variables. Non-quantifiable parameters (such as the presence of dementia or physical frailty) or the use of substances known to interact with NOACs (e.g. cytochrome P450 3A4 inhibitors, P-glycoprotein inhibitors, or antifungal drugs) were not recorded in the registry. Thus, there was the possibility of an overestimation of the number of NSDR patients. Second, the number of patients was relatively small, particularly for the analysis of clinical outcomes (including stroke or bleeding events); therefore, it is still unknown whether NSDR in patients with NOACs actually contributed to worse outcomes. Despite this limitation, we compared outcomes within the NOAC sub-groups using data up to the 1-year follow-up examination that were available in July 2017 [n = 997; (1) full dose: n = 671 (67.3%); (2) SDR: n = 190 (19.0%); and (3) NSDR: n = 136 (13.6%)]. Stroke is defined as a new, sudden, focal neurologic deficit that persists beyond 24 h and is not due to a readily identifiable, nonvascular cause (e.g. seizure). Major bleeding was defined by the International Society of Thrombosis and Hemostasis criteria as a bleeding event meeting at least one of the following criteria: (1) fatal bleeding, and/or; (2) symptomatic bleeding in a critical area or organ, such as intracranial, intraspinal, intraocular, retroperitoneal, intraarticular or pericardial, or intramuscular with compartment syndrome, and/or; (3) bleeding causing a fall in the hemoglobin level of 2 g/dL or more, or leading to transfusion of two or more units of whole blood or red cells. The number of patients who suffered from stroke was one in the full-dose group, two in the SDR group, and zero in the NSDR group; there were no significant differences. In addition, the number of patients who suffered from major bleeding was eight in the full-dose group, six in the SDR group, and four in the NSDR group; there were no significant differences (Supplemental Table 1). Third, we excluded patients who were prescribed standard-dose NOACs with a renal indication for a reduced dose (such as potential overdosing or patients with chronic kidney disease on hemodialysis). Finally, our classification of the NOAC dose was based on the CrCl level at one time point. However, the CrCl can fluctuate over time, and we may have over- or underestimated SDR or NSDR.
Conclusions
In our contemporary Japanese registry, nearly half of the NOAC dose reductions in patients with NVAF were non-standardized. This suggests that the physician's apprehension regarding excessive bleeding must be balanced with a consideration for an increased risk of embolic events, particularly in patients with NVAF with a significant risk profile.
Sources of funding
This study was funded by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (Grant No. 16H05215) and by an unrestricted research grant from Bayer Yakuhin Ltd.
Disclosure of interest
Dr Kohsaka received an unrestricted research grant from Bayer Yakuhin and Daiichi Sankyo Co., Ltd., on behalf of the Department of Cardiology, Keio University School of Medicine. The authors have no other relationships that could be construed as a conflict of interest.
Acknowledgments
We are indebted to all of the study coordinators, investigators, and patients who participated in the Keio Inter-hospital Cardiovascular Studies for Atrial Fibrillation registry.
Appendix A.
Site investigators:
Yukihiko Momiyama, Munehisa Sakamoto, Jun Fuse, Kojiro Tanimoto, Yoko Tanimoto, Yukinori Ikegami, Kohei Inagawa (National Hospital Organization Tokyo Medical Center); Iwao Nakamura, Jyunji Suzuki, Tomohiro Matsuhashi, Hiroshi Shiga (Hino Municipal Hospital); Seiji Takatsuki, Yoshiyasu Aizawa, Takahiko Nishiyama, Yoshinori Katsumata, Akira Kunitomi, Kazuaki Nakajima, Taishi Fujisawa (Keio University School of Medicine); Masahiro Suzuki, Keisuke Matsumura, Tomohiko Ono, Hanako Tokuda, Ryutaro Yamaguchi, Hiroaki Tanaka (National Hospital Organization Saitama National Hospital); Shigetaka Noma, Takashi Yagi, Kenichiro Shimoji, Koji Ueno, Satoshi Mogi (Saiseikai Utsunomiya Hospital); Takashi Koyama, Shiro Ishikawa, Hideaki Kanki, Takashi Akima, Masahito Munakata, Kazutaka Miyamoto (Saitama City Hospital); Hideo Mitamura, Kazunori Moritani, Masaru Shibata, Toshimi Kageyama (Tachikawa Hospital); Takahiro Oki, Akiyasu Baba, Yoshinori Mano, Hiroaki Sukegawa (Tokyo Dental College Ichikawa General Hospital); Kouji Negishi, Takahiro Koura, Daisuke Shinmura, Kotaro Fukumoto, Hiroyuki Yamakawa, Shin Kashimura (Yokohama Municipal Citizen's Hospital). Keiichi Nagami, Kazuhiro Oyamada, Kotaro Naitou, Keijiro Chiba (Keiyu Hospital); Megumi Shimada (Tokai University Oiso Hospital); Makoto Akaishi (Tokai University Tokyo Hospital).
Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials.
Non-vitamin K antagonist oral anticoagulants have better efficacy and equivalent safety compared to warfarin in elderly patients with atrial fibrillation: a systematic review and meta-analysis.
Predictive factors for bleeding during treatment with rivaroxaban and warfarin in Japanese patients with atrial fibrillation – subgroup analysis of J-ROCKET AF.
Bleeding risk of apixaban, dabigatran, and low-dose rivaroxaban compared with warfarin in Japanese patients with non-valvular atrial fibrillation: a propensity matched analysis of administrative claims data.
Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study.
30269-7&id=gr1.jpg){kind=link}
30269-7&id=gr2.jpg){kind=link}
30269-7&id=gr3.jpg){kind=link}
30269-7&id=gr4.jpg){kind=link}
30269-7&id=gr5.jpg){kind=link}
30269-7&id=gr6.jpg){kind=link}