Author + information
- Received April 28, 2014
- Revision received August 19, 2014
- Accepted September 16, 2014
- Published online December 16, 2014.
- Anders Nissen Bonde, MB∗,
- Gregory Y.H. Lip, MD†,
- Anne-Lise Kamper, MD, DMSc‡,
- Peter Riis Hansen, MD, PhD, DMSc∗,
- Morten Lamberts, MD, PhD∗,
- Kristine Hommel, MD, PhD‡,
- Morten Lock Hansen, MD, PhD∗,
- Gunnar Hilmar Gislason, MD, PhD∗,§,
- Christian Torp-Pedersen, MD, DMSc‖ and
- Jonas Bjerring Olesen, MD, PhD∗,¶∗ ()
- ∗Department of Cardiology, Copenhagen University Hospital Gentofte, Gentofte, Denmark
- †University of Birmingham Centre for Cardiovascular Sciences, City Hospital, Birmingham, United Kingdom
- ‡Department of Nephrology, Copenhagen University Hospital Rigshospitalet, Rigshospitalet, Denmark
- §The National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark
- ‖Department of Health, Science and Technology, Aalborg University, Aalborg, Denmark
- ¶Department of Cardiology, Copenhagen University Hospital Hvidovre, Hvidovre, Denmark
- ↵∗Reprint requests and correspondence:
Dr. Anders Nissen Bonde, Copenhagen University Hospital, Department of Cardiology, Gentofte, Niels Andersens Vej 65, Post 635, 2900 Hellerup, Denmark.
Background The balance between stroke reduction and increased bleeding associated with antithrombotic therapy among patients with atrial fibrillation (AF) and chronic kidney disease (CKD) is controversial.
Objectives This study assessed the risk associated with CKD in individual CHA2DS2-VASc (Congestive heart failure; Hypertension; Age ≥75 years; Diabetes mellitus; previous Stroke, transient ischemic attack, or thromboembolism; Vascular disease; Age 65 to 74 years; Sex category) strata and the net clinical benefit of warfarin in patients with AF and CKD in a nationwide cohort.
Methods By individual-level linkage of nationwide Danish registries, we identified all patients discharged with nonvalvular AF from 1997 to 2011. The stroke risk associated with non-end-stage CKD and end-stage CKD (e.g., patients on renal replacement therapy [RRT]) was estimated using Cox regression analyses. The net clinical benefit of warfarin was assessed using 4 endpoints: a composite endpoint of death/hospitalization from stroke/bleeding; a composite endpoint of fatal stroke/fatal bleeding; cardiovascular death; and all-cause death.
Results From nonvalvular AF patients (n = 154,259), we identified 11,128 patients (7.2%) with non-end-stage CKD and 1,728 (1.1%) receiving RRT. In all CHA2DS2-VASc risk groups, RRT was independently associated with a higher risk of stroke/thromboembolism, from a 5.5-fold higher risk in patients with CHA2DS2-VASc score = 0 to a 1.6-fold higher risk in patients with CHA2DS2-VASc score ≥2. In patients receiving RRT with CHA2DS2-VASc score ≥2, warfarin was associated with lower risk of all-cause death (hazard ratio [HR]: 0.85, 95% confidence interval [CI]: 0.72 to 0.99). In non-end-stage CKD patients with CHA2DS2-VASc score ≥2, warfarin was associated with a lower risk of a composite outcome of fatal stroke/fatal bleeding (HR: 0.71, 95% CI: 0.57 to 0.88), a lower risk of cardiovascular death (HR: 0.80, 95% CI: 0.74 to 0.88), and a lower risk of all-cause death (HR: 0.64, 95% CI: 0.60 to 0.69).
Conclusions CKD is associated with a higher risk of stroke/thromboembolism across stroke risk strata in AF patients. High-risk CKD patients (CHA2DS2-VASc ≥2) with AF benefit from warfarin treatment for stroke prevention.
The optimal management of thromboprophylaxis in patients with atrial fibrillation (AF) and chronic kidney disease (CKD) is complex. Whereas CKD patients are at high risk of stroke and thromboembolism (TE), these patients are also at high risk of death and major bleeding. This is particularly true of patients with end-stage CKD treated with renal replacement therapy (RRT), whether as dialysis or renal transplantation (1–4). Thromboprophylaxis in this high-risk group is therefore complex, and clinical decision making requires interpretation of the balance between the risks of these important endpoints in individual patients.
The published evidence for oral anticoagulation in patients with AF shows that the use of adjusted dose warfarin, compared with placebo, results in a 64% reduction in stroke and, compared with placebo or aspirin, a 26% reduction in all-cause mortality (5). However, patients with a creatinine clearance of <30 ml/min have been excluded from clinical trials (6). Thus, antithrombotic treatment in patients with CKD stages 4 to 5 and AF has hitherto been on the basis of a number of small observational studies and a few registry-based cohorts (7–12).
We have previously shown that in AF patients with CKD, warfarin treatment was associated with a decreased risk of stroke/TE, as well as an increased risk of bleeding (11). The present study investigated the net clinical benefit of antithrombotic therapy in patients with both AF and CKD (e.g., balancing efficacy and safety of the treatment). During recent years, there has been growing attention to the question whether CKD should be included in stroke risk stratification models (13–16), and validation of stroke stratification models should ideally be performed in non-anti-coagulated real-world cohorts. For this reason, the risk of stroke/TE associated with CKD was determined in patients not receiving warfarin, and the net clinical benefit of warfarin was determined in individual stroke risk strata, as determined by the CHA2DS2-VASc (Congestive heart failure; Hypertension; Age ≥75 years; Diabetes mellitus; previous Stroke, transient ischemic attack, or thromboembolism; Vascular disease; Age 65 to 74 years; Sex category) score. First, we hypothesized that CKD would be independently associated with a higher risk of stroke/TE in all stroke risk strata of non-anti-coagulated patients with AF. Second, we tested the hypothesis that the benefits of warfarin would outweigh its risks in AF patients with CKD and a high risk of stroke/TE.
In Denmark, linkage of data from nationwide registries on an individual level is possible using the unique, permanent, and personal registration number provided to all Danish citizens. All Danish hospital admissions since 1978 are registered in the National Patient Registry with diagnoses coded according to the International Classification of Diseases (ICD) system (17,18). Since 1996, invasive therapeutic procedures have been coded according to the Nordic Medical Statistics Committees Classification of Surgical Procedures. All prescriptions dispensed from Danish pharmacies are registered in the Danish National Prescription registry with information on drug strength, quantity, and anatomical therapeutic chemical code (19,20). Accurate data on all Danish patients actively treated for end-stage CKD with RRT are registered in the Danish Nephrology Registry (21). The civil registration system holds information on vital status for every citizen (22), and information on primary and contributing causes of death are registered in the National Cause of Death Registry (23). Diagnoses, surgical procedures, and pharmacotherapy used for defining the population, comorbidity, and outcomes are presented in Online Table 1. Ethical approval is not required for retrospective register-based studies in Denmark. The Danish Data Protection Agency approved the study.
We identified all patients discharged from a Danish hospital with nonvalvular AF (ICD-10 code I48 or ICD-8 code 4279) as either primary or secondary diagnosis in the 15-year study period from January 1, 1997 to December 31, 2011. Pharmacotherapy may have been changed or intensified in relation to the index AF hospitalization; therefore, follow-up began 7 days after discharge. Patients that experienced stroke/TE, a major bleeding, or died in this 7-day “quarantine” period were excluded (Figure 1).
Chronic kidney disease and renal replacement therapy
Patients with non-end-stage CKD were identified from the National Patient Registry using the diagnosis codes presented in Online Table 1. This definition of non-end-stage CKD has been used earlier (11,24). To evaluate the general level of renal function in patients diagnosed with non-end-stage CKD, a sample of 110 non-end-stage CKD patients (approximately 1% of the total non-end-stage CKD population) was studied; estimated glomerular filtration rate was calculated according to the Modification of Diet in Renal Disease equation, using the first plasma creatinine value during hospitalization, sex, and age (25). Of these, 9 patients (8.2%) were stage-1, 12 (10.9%) were stage-2, 22 (20.0%) were stage-3, 40 (36.4%) were stage-4, and 27 (24.5%) were stage-5 CKD. Patients on RRT (i.e., hemodialysis, peritoneal dialysis, or living with a renal transplant) were identified from the well-validated Danish Nephrology Registry (11,24). CKD status was determined at baseline and time-dependently throughout follow-up: 1) non-CKD patients could change status to non-end-stage CKD or RRT during follow-up; 2) patients with non-end-stage CKD could change status to RRT; and 3) RRT patients could not change renal status (Figure 1). We had data from the Danish Nephrology Register from 1997 to 2010.
Pharmacological treatment for all drugs other than warfarin and aspirin at baseline was determined by claimed prescriptions from 180 days before discharge to 7 days thereafter. Patients receiving antiplatelet drugs other than aspirin (i.e., clopidogrel, prasugrel, or dipyridamole) were excluded from the study population (Figure 1). Periods in which each patient was treated with warfarin, aspirin, or both drugs were determined throughout follow-up by dividing the number of tablets dispensed with the estimated daily dosage (26,27). Dabigatran was only recently approved in Denmark (October 2010) and is unlikely to be used in most CKD patients due to its renal excretion. Therefore, we did not include dabigatran data in our study.
The CHA2DS2-VASc and HAS-BLED scores
The risk of stroke/TE for all patients was assessed according to the CHA2DS2-VASc score (28,29). The CHA2DS2-VASc score is calculated by adding 1 point each for the presence of heart failure, hypertension, diabetes, vascular disease, age 65 to 74 years, and female sex, and 2 points each for the presence of previous stroke/TE and age ≥75 years (30). A score of 0 was considered “low risk,” 1 was “intermediate risk,” and ≥2 was “high risk” of stroke/TE. The risk of bleeding for all patients was assessed using the HAS-BLED (Hypertension, Abnormal renal function, abnormal liver function, Stroke, Bleeding, Labile international normalized ratio, Elderly, Drug therapy, alcohol intake) score (11,31). The HAS-BLED score is calculated by adding 1 point each for the presence of hypertension, abnormal renal or liver function, previous stroke/TE, a history of bleeding, labile international normalized ratio, age ≥65 years, concomitant therapy with nonsteroidal anti-inflammatory drugs, and excessive alcohol intake. As CKD was the object of the study and because international normalized ratio data were not available, these risk factors were not included in the score.
Study outcomes under investigation were the following: 1) hospitalization/death from stroke/TE (32) (i.e., peripheral arterial embolism, ischemic stroke, and transient ischemic attack); 2) a composite outcome of death/hospitalization from stroke/TE/bleeding (i.e., gastrointestinal, intracranial, urinary tract, and airway bleeding); 3) a composite of fatal stroke/fatal bleeding; 4) cardiovascular death; or 5) death from any cause (23). Outcome 1 was used to determine the risk of stroke/TE associated with CKD. Outcomes 2 to 5 were used to determine the net clinical benefit of antithrombotic therapy in patients with AF and CKD.
Event rates of stroke/TE according to CKD, warfarin treatment, and CHA2DS2-VASc score were estimated time-dependently. The risk of stroke/TE associated with non-end-stage CKD, and RRT in patients that were not treated with warfarin was estimated in Cox proportional-hazards survival models. Separate analyses were conducted for patients at low, intermediate, and high risk of stroke/TE. First, the Cox regression models were adjusted for risk factors included in the CHA2DS2-VASc score and aspirin treatment. Second, the Cox regression models were adjusted for all baseline characteristics, with age included in the model as a continuous covariate. We used the Wald test to test for differences between the non-end-stage CKD group and the RRT group in our model. The net clinical benefit of warfarin and aspirin treatment was estimated in time-dependent Cox proportional hazards survival models outcomes 2 to 5. Analyses were conducted separately for patients with non-end-stage CKD and RRT, respectively, and according to CHA2DS2-VASc score, with adjustment for changes in renal status and antithrombotic treatment during follow-up. The models were adjusted for risk factors in the CHA2DS2-VASc score and the HAS-BLED score, with age included in the model as a continuous covariate.
Treatment is often discontinued in terminal patients, and they may therefore erroneously appear to die during an untreated period. This could lead to false estimates that would favor treatment over no treatment. For this reason, net clinical benefit analyses of “fatal only” outcomes were made, with patients analyzed as being in treatment until 30 days after treatment discontinuation. Sensitivity analyses were later conducted for all outcomes with treatment periods of 0, 30, and 90 days, respectively.
Further, sensitivity analyses on the net clinical benefit of warfarin and aspirin were conducted with separate columns for hemodialysis patients, peritoneal dialysis patients, and renal transplant recipients, respectively.
A 2-sided p value of <0.05 was considered statistically significant. All analyses were performed with SAS statistical software (version 9.2, SAS Institute Inc., Cary, North Carolina) and R (version 2.15.2, R Development Core Team).
We included 154,259 patients with nonvalvular AF during the 15-year study period and identified a total of 11,128 patients (7.2%) with non-end-stage CKD and 1,728 (1.2%) receiving RRT. Of these, 4,519 had non-end-stage CKD and 1,142 were receiving RRT at inclusion (Table 1). During follow-up, 6,609 patients developed non-end-stage CKD and 586 initiated RRT. Of the 1,728 patients receiving RRT, 1,245 (72.0%) were receiving hemodialysis, 435 (25.1%) were receiving peritoneal dialysis, and 48 (2.8%) were kidney transplant recipients. Overall, non-end-stage CKD patients were older, had more comorbidities, and received more pharmacological treatment, whereas RRT patients were more often men and hypertensive. All CKD patients were less likely to receive warfarin treatment. Among patients who had CKD and CHA2DS2-VASc score = 0 at baseline (n = 153), compared with intermediate and high-risk CKD patients, there were more patients with Wegener granulomatosis, chronic glomerulonephritis, or autosomal dominant polycystic kidney disease. Of the 19,045 patients with CHA2DS2-VASc score = 1, 5,642 patients (29.6%) had this score because of their female sex.
Non-CKD patients had a median follow-up of 1,179 (interquartile range [IQR]: 397 to 2,412) days and in this period 19,877 patients (13.4%) experienced stroke/TE. Non-end-stage CKD patients had a median follow-up of 312 (IQR: 48 to 952) days and 1,087 (9.9%) experienced stroke/TE. RRT patients had a median follow-up of 603 (IQR: 225 to 1,300) days and 227 (13.2%) had stroke/TE. A total of 322 patients emigrated during the study period and were censored on the date of emigration. Online Table 2 shows the exact numbers of all events in our study. Online Figure 1 shows cumulative incidence curves of cardiovascular death according to CHA2DS2-VASc score and CKD group.
Risk of stroke/TE
Table 2 displays rates of stroke/TE according to CKD, warfarin treatment, and CHA2DS2-VASc score. In patients with CHA2DS2-VASc score = 0 not treated with warfarin, the stroke/TE rate per 100 person-years increased from 0.8 (95% confidence interval [CI]: 0.7 to 0.8) in patients without CKD to 4.2 (95% CI: 1.5 to 6.9) in RRT patients. In patients with CHA2DS2-VASc score ≥2 not treated with warfarin, the stroke/TE rate was substantially increased in both non-end-stage CKD and RRT patients when compared with AF patients without CKD. Event rates of stroke/TE were generally lower in patients treated with warfarin than in those patients not treated with warfarin. When female patients with CHA2DS2-VASc score = 1 were reclassified as CHA2DS2-VASc score = 0, given that such patients are “low risk,” the following event rates were obtained in the non-anti-coagulated patients: non-end-stage CKD patients as low risk had a stroke/TE rate of 1.6 per 100 person-years, compared with a stroke/TE rate of 1.8 in non-end-stage CKD patients at intermediate risk, and RRT patients at low risk had a stroke/TE rate of 3.1, compared with a stroke/TE rate of 3.2 in RRT patients at intermediate risk, respectively.
Table 3 shows the results from Cox proportional hazards survival model, adjusted for CHA2DS2-VASc score and aspirin treatment, and CHA2DS2-VASc score and all baseline medication, respectively. Among AF patients with CHA2DS2-VASc score = 0, CKD was associated with a higher risk of stroke/TE among both non-end-stage CKD and RRT patients when compared with patients without CKD. This association was not found in non-end-stage CKD patients with CHA2DS2-VASc score = 1 (hazard ratio [HR]: 0.94, 95% CI: 0.60 to 1.48). CKD was independently associated with a higher risk of stroke/TE in high-risk patients (CHA2DS2-VASc score ≥2): from a 1.3-fold higher risk with non-end-stage CKD to a 1.6-fold higher risk with RRT. There was a significant difference between the non-end-stage CKD group and the RRT group in patients with CHA2DS2-VASc score ≥1 (p value for interaction <0.0001).
Warfarin and aspirin treatment
Results from Cox regression analyses on net clinical benefit are shown in Figure 2. Among high-risk non-end-stage CKD patients, warfarin was associated with a lower risk of cardiovascular death (HR: 0.80, 95% CI: 0.74 to 0.88) and all-cause mortality (HR: 0.64, 95% CI: 0.60 to 0.69) than no antithrombotic treatment was. Among low-/intermediate-risk non-end-stage CKD patients, warfarin was associated with a significantly lower risk of all-cause mortality (HR: 0.62, 95% CI: 0.49 to 0.79) and a non-significant trend toward lower risk of cardiovascular death. Treatments with aspirin and aspirin plus warfarin were associated with a lower risk of all-cause mortality, but not with a lower risk of any other outcome.
Among high-risk RRT patients, warfarin was associated with a significantly lower risk of all-cause mortality (HR: 0.85, 95% CI: 0.72 to 0.99) and nonsignificant trends toward lower risks of cardiovascular death and of a composite of death/hospitalization from stroke/TE/bleeding, respectively. Treatment with aspirin or aspirin plus warfarin was not associated with a lower risk of any outcome among RRT patients.
Online Figure 2 shows results from time-dependent analyses on “fatal stroke/fatal bleeding.” Among high-risk non-end-stage CKD patients, warfarin was associated with a lower risk of a composite outcome of fatal stroke/fatal bleeding (HR: 0.71, 95% CI: 0.57 to 0.88).
Analyses of associations between of antithrombotic treatment and all outcomes investigated in this study were also performed for hemodialysis patients, peritoneal dialysis patients, and renal transplant recipients (Online Table 3). There was a significant interaction between the 2 dialysis modalities on warfarin in low-risk patients (p = 0.002) and on aspirin in high-risk patients for the outcome of stroke/TE/bleeding (p = 0.041). There was a trend toward fewer negative outcomes with warfarin among high-risk hemodialysis patients than among high-risk peritoneal dialysis patients. Excluding renal transplanted patients from the RRT group did not alter the direction of any of the results.
Online Table 4 shows net clinical benefit analyses of antithrombotic treatment in patients with AF and CKD when prolonging treatment periods in the analyses for 0, 30, and 90 days, respectively. Prolonging treatment periods resulted in more conservative risk estimates and wider confidence intervals for all study outcomes and all treatments investigated. Warfarin was associated with a lower risk of all-cause mortality in patients with non-end-stage CKD after prolonging treatment periods for both 30 and 90 days. After prolonging treatment periods for 30 days, warfarin was still significantly associated with a lower risk of all-cause mortality in patients with AF on RRT, but this association was not found after prolonging treatment periods for 90 days. The study design and main results are summarized in the Central Illustration.
This nationwide study investigated the risk of stroke/TE and the net clinical benefit of antithrombotic treatment in patients with AF and CKD. First, RRT patients in all stroke risk (CHA2DS2-VASc score) strata had a higher HR of stroke/TE. Second, in RRT patients with CHA2DS2-VASc score ≥2, warfarin was associated with a significant net clinical benefit, as warfarin was associated with a lower risk of all-cause mortality, a nonsignificant trend toward lower risk of cardiovascular death and a composite outcome of death/hospitalization from stroke/TE/bleeding. Third, non-end-stage CKD patients with CHA2DS2-VASc score ≥2 or CHA2DS2-VASc = 0 had higher HR of stroke/TE. Fourth, treatment with warfarin in high-risk non-end-stage CKD patients was associated with a significant net clinical benefit, as it was associated with a lower risk of the composite outcome of fatal stroke/fatal bleeding, a lower risk of cardiovascular death, and a lower risk of all-cause mortality, respectively.
Of note, CKD patients not treated with warfarin with CHA2DS2-VASc = 1 had lower rates of stroke/TE than did CKD patients with CHA2DS2-VASc = 0 in our study. When we grouped female patients with CHA2DS2-VASc score = 1 together with CHA2DS2-VASc score = 0 (e.g., classifying these patients as “low risk” in accordance with 2012 European Society of Cardiology guidelines ), other CKD patients with CHA2DS2-VASc score = 1 had higher rates of stroke/TE than did CKD patients with CHA2DS2-VASc score = 0.
It has recently been debated whether CKD should be added as an extra risk point to the CHA2DS2-VASc score (13–16). We found that RRT was independently associated with a higher HR of stroke/TE in all stroke risk strata of the CHA2DS2-VASc score, from a 5.5-fold higher risk in low-risk patients to a 1.6-fold higher risk in high-risk patients. Non-end-stage CKD was independently associated with a higher HR of stroke/TE in patients with CHA2DS2-VASc = 0 or 2. Warfarin was associated with a lower risk of all-cause mortality in non-end-stage CKD patients with CHA2DS2-VASc ≤1. In these latter low-/intermediate-risk CKD patients, our results may therefore suggest a potential benefit of warfarin.
One previous study found warfarin to be beneficial in AF patients with non-end-stage CKD compared to no-treatment patients (33). Our study supports these findings, considering that warfarin was found to be associated with a net clinical benefit among high-risk non-end-stage CKD patients. Among patients with AF receiving RRT, the value of warfarin has previously been challenged, and routine anticoagulation for primary stroke prevention has not been recommended (34), given that these patients were excluded from previous clinical trials. Some observational studies have reported higher rates of stroke or bleeding with warfarin among RRT patients (7,9,12), whereas others did not find this association (8,11,35), and 1 study, like the present analysis, found a lower risk of all-cause mortality with warfarin (10). Compared with previous studies, our study perhaps has greater clinical relevance and application, given the large size and the fact that we accounted for different stroke risk strata in our analyses. We found a net clinical benefit of warfarin in RRT patients with CHA2DS2-VASc score ≥2. This association was still significant when prolonging treatment periods for 30 days but not when prolonging treatment periods for 90 days, indicating that some of the apparent positive effect associated with warfarin may be caused by treatment cessation in some of the terminal patients. When compared with no antithrombotic treatment, aspirin, which has previously been used for antithrombotic therapy among RRT patients who are considered unsuitable for warfarin treatment, was not associated with a net clinical benefit in any analysis. The relationship between reduced stroke/TE and increased bleeding with antithrombotic therapy might differ across the various RRT modalities. We observed a trend toward fewer negative outcomes with warfarin among high-risk hemodialysis patients than among high-risk peritoneal dialysis patients, but this difference was not seen on all-cause mortality.
Study strengths and limitations
This study is limited by its observational cohort design and therefore residual confounding may still be present. Data on renal function were not available for all patients, and we were therefore unable to stratify the non-end-stage CKD group into glomerular filtration rate categories or to account for changes in renal function during the study period. Patients emigrating during the study period were lost to follow-up, and we had no information on eventual health care contacts outside of Denmark. Because diabetes, heart failure, and hypertension were identified on the basis of filled prescriptions, the frequency of these risk factors might have been underestimated, and the number of patients with low risk of stroke/TE might have been overestimated, because an unknown number of patients may have been treated with diet and lifestyle interventions alone. We were not able to control for blood pressure per se due to the lack of data on blood pressure measurements. The inclusion of only hospitalized patients with AF increased the proportion of patients at high risk of stroke/TE and death. We were not able to distinguish between paroxysmal and persistent AF. Because aspirin can be bought over the counter in Denmark, we might have underestimated the number of aspirin users. Because some dialysis centers in Denmark have reported that they use warfarin for patients without prescription, we might have underestimated the number of RRT patients that used warfarin. We did not have access to brain imaging data, and some of the strokes reported as ischemic might have been hemorrhagic, although a previous study (32) on the subject did not find this potential bias to be of major importance. The strengths of our data are that the positive predictive value of the AF diagnosis in our study is very high (36), and the stroke diagnosis is well validated (32). Information on all-cause mortality is accurate (23). Our RRT information is also precise (21).
We have found that among AF patients, CKD was associated with a higher risk of stroke/TE across stroke risk (CHA2DS2-VASc) strata, most notably among RRT patients. In RRT patients with CHA2DS2-VASc score ≥2, warfarin was associated with a net clinical benefit, as it was associated with a lower risk of all-cause mortality. In non-end-stage CKD patients with CHA2DS2-VASc score ≥2, warfarin was associated with a net clinical benefit, as it was associated with a lower risk of all-cause mortality, a lower risk of cardiovascular death, and a lower risk of the composite outcome of fatal stroke/fatal bleeding, respectively. These observations provide guidance for the optimal use of antithrombotic therapy in AF patients with CKD.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with AF who have severe CKD have been excluded from most randomized trials involving anticoagulation for stroke prevention, but warfarin is associated with net clinical benefit.
COMPETENCY IN PATIENT CARE: Warfarin therapy is generally beneficial for patients with AF and CKD unless contraindicated for other reasons.
TRANSLATIONAL OUTLOOK: There is a need for further studies, including randomized trials comparing warfarin with dose-adjusted target-specific oral anticoagulants, in patients with AF and CKD to assess safety and efficacy for prevention of thromboembolism in addition to the impact of treatment on the progression of renal disease.
For supplemental figures and tables, please see the online version of this paper.
The Capital Region of Denmark, Foundation for Health Research has financed this study through a research grant for Dr. Jonas Bjerring Olesen. Dr. Lip has served as a consultant for Bayer Healthcare, Astellas, Merck & Co., Inc., Sanofi, BMS/Pfizer, Daiichi-Sankyo, Biotronik, Medtronic, Portola, AstraZeneca, and Boehringher Ingelheim; and has served on the speakers bureaus for Bayer Healthcare, BMS/Pfizer, Boehringher Ingelheim, Daiichi-Sankyo, Medtronic, and Sanofi. Dr. Gislason has received research grants and speaking fees from AstraZeneca and Bristol-Myers Squibb. Dr. Torp-Pedersen has developed antiarrhythmic drugs; and has received fees for lecturing, serving on advisory boards, and steering committees for Cardiome, Sanofi, and Merck & Co., Inc. Dr. Bjerring Olesen has received speaking fees from Bristol-Myers Squibb; and funding for research from the Lundbeck Foundation, Bristol-Myers Squibb, and The Capital Region of Denmark, Foundation for Health Research. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- atrial fibrillation
- confidence interval
- chronic kidney disease
- hazard ratio
- International Classification of Diseases
- interquartile range
- renal replacement therapy
- Received April 28, 2014.
- Revision received August 19, 2014.
- Accepted September 16, 2014.
- American College of Cardiology Foundation
- Go A.S.,
- Fang M.C.,
- Udaltsova N.,
- et al.,
- for the ATRIA Study Investigators
- Reinecke H.,
- Engelbertz C.,
- Schäbitz W.R.
- Di Angelantonio E.,
- Chowdhury R.,
- Sarwar N.,
- Aspelund T.,
- Danesh J.,
- Gudnason V.
- Marinigh R.,
- Lane D.A.,
- Lip G.Y.
- Chan K.E.,
- Lazarus J.M.,
- Thadhani R.,
- Hakim R.M.
- Winkelmayer W.C.,
- Liu J.,
- Setoguchi S.,
- Choudhry N.K.
- Shah M.,
- Avgil T.M.,
- Jackevicius C.A.,
- et al.
- Apostolakis S.,
- Guo Y.,
- Lane D.A.,
- Buller H.,
- Lip G.Y.
- Piccini J.P.,
- Stevens S.R.,
- Chang Y.,
- et al.,
- for the ROCKET AF Steering Committee and Investigators
- Banerjee A.,
- Fauchier L.,
- Vourc'h P.,
- et al.
- Lynge E.,
- Sandegaard J.L.,
- Rebolj M.
- Kildemoes H.W.,
- Sørensen H.T.,
- Hallas J.
- Hommel K.,
- Rasmussen S.,
- Madsen M.,
- Kamper A.L.
- Pedersen C.B.
- Helweg-Larsen K.
- Blicher T.M.,
- Hommel K.,
- Olesen J.B.,
- Torp-Pedersen C.,
- Madsen M.,
- Kamper A.L.
- Sørensen R.,
- Hansen M.L.,
- Abildstrom S.Z.,
- et al.
- Olesen J.B.,
- Lip G.Y.,
- Hansen M.L.,
- et al.
- Camm A.J.,
- Lip G.Y.,
- De Caterina R.,
- et al.,
- for the ESC Committee for Practice Guidelines
- Hart R.G.,
- Pearce L.A.,
- Asinger R.W.,
- Herzog C.A.