Warfarin Use and Stroke Risk Among Patients With Nonvalvular Atrial Fibrillation in a Large Managed Care Population

Background

Stroke prevention is a goal of atrial fibrillation (AF) management, but discontinuation of warfarin anticoagulation therapy is common.

Objective

To investigate the association between warfarin discontinuation and hospitalization for stroke among nonvalvular AF (NVAF) patients enrolled in managed care.

Methods

Patients with NVAF who initiated warfarin therapy from January 2005 through June 2009 were included. Warfarin discontinuation was defined as a supply gap >60 days without evidence of International Normalized Ratio measurements. Follow-up, which was a variable time period from warfarin initiation until the earlier of death, disenrollment from the health plan, or June 30, 2010, was divided into periods of warfarin treatment and discontinuation. Stroke events were identified based on claims for inpatient stays with a primary diagnosis of stroke or transient ischemic attack. Cox proportional hazards models were constructed to assess the relationship between warfarin discontinuation and incident stroke while adjusting for baseline demographics, stroke and bleeding risk, and comorbidities, as well as time-dependent antiplatelet use, stroke, and bleeding events in the previous warfarin treatment period.

Results

Among warfarin initiators with NVAF (N = 16,253), 51.4% discontinued warfarin therapy at least once during a mean follow-up of 668 days. Stroke risk was significantly greater during warfarin discontinuation periods compared with therapy periods (hazard ratio = 1.60; 95% CI, 1.35–1.90; P < 0.001).

Conclusions

More than half of patients on warfarin had treatment gaps or discontinued therapy. Therapy gaps were associated with increased stroke risk.     

Efficacy and safety of traditional Chinese medicine on thromboembolic events in patients with atrial fibrillation: A systematic review and meta-analysis

ObjectivesTraditional Chinese medicine (TCM) in combination with Western medicine (WM) has been widely used worldwide. This systematic review aimed to evaluate the efficacy and safety of TCM in prevention of thromboembolic events in patients with atrial fibrillation (AF).MethodsPotential studies were searched through the Cochrane Library, PubMed, EMBASE, CBM, VIP, CNKI, and Wanfang databases up to February 2016. Randomized controlled trials (RCTs) investigating the thromboembolic events and/or safety outcome of TCM in patients with AF were included.ResultsA total of 905 AF patients from 9 RCTs were identified. Meta-analysis showed that TCM in combination with warfarin was better than warfarin alone for preventing total thromboembolic events with a 68% reduction of risk (risk ratio [RR] 0.32; 95% confidence interval [CI] 0.13–0.78) without increasing the risk of total bleeding (RR 0.71; 95% CI 0.29–1.72). Compared with warfarin, TCM therapy was associated with lower risk of total bleeding (RR 0.13; 95% CI 0.04–0.47), but increased the risk of total thromboembolic events (RR 1.84; 95% CI 1.03–3.27).ConclusionsThis meta-analysis suggests that TCM combined with warfarin is superior to warfarin alone for the prevention of total thromboembolic events in patients with AF, with equal risk of bleeding as warfarin alone.     

Management patterns and outcomes of patients with venous thromboembolism in the usual community practice setting

OBJECTIVE: The objectives of this study were to observe a commercially insured sample diagnosed with a venous thromboembolism (VTE) event and treated postevent with warfarin and to detail the thromboembolic and bleeding outcomes in the time periods during warfarin therapy and after discontinuation of such therapy. METHODS: This retrospective, observational cohort study used medical, pharmacy, and eligibility data from 2 US health plans. Study inclusion required an inpatient diagnosis of deep venous thrombosis (DVT) or pulmonary embolism (PE) between January 1, 1998, and December 31, 2000; warfarin, heparin, or low-molecular-weight heparin within 30 days after diagnosis; no VTE diagnosis; and no anticoagulant use for 3 months preceding diagnosis. A random sample of medical charts was abstracted to validate VTE events and collect prothrombin time/international normalized ratio (INR) result data. Recurrent VTE events, bleeding events, and proportion of time within INR range were captured in the postindex VTE event time period. Univariate and multivariate statistical techniques were used to assess outcomes. RESULTS: A total of 2,090 patients were identified with a newly diagnosed VTE event (DVT only, 1450; PE with or without DVT, 640). Mean (SD) age was 61.7 (16) years; mean (SD) follow-up time after the index diagnosis was 21.3 (10) months. Overall mean (SD) length of warfarin therapy was 6.6 (6) months. During the follow-up period, 224 patients (10.7%) experienced a recurrent VTE event and 122 patients (5.8%) experienced a bleeding event requiring hospitalization. The cumulative incidence of recurrent VTE events over 3 and 6 months was 9.0% and 10.9%, respectively. Using the chart abstraction subset, patients were within the appropriate INR range 37.7% of the time while receiving warfarin. CONCLUSIONS: Negative outcomes associated with warfarin therapy-recurrent VTE events and bleeding requiring hospitalization-were experienced by 10.7% and 5.8% of patients, respectively. These data suggest that negative outcomes may be more prevalent in usual community medical practice compared with rates observed in the controlled environment of the clinical trial or specialized anticoagulation clinic.     

利伐沙班对合并心功能不全的高龄非瓣膜性心房颤动疗效及安全性

目的 分析利伐沙班对合并心功能不全的高龄非瓣膜性心房颤动（non valvular atrial fibrillation,NVAF）患者临床疗效及安全性。方法 随机连续纳入2016年1月至2020年1月间就诊于我院合并心功能不全的高龄NVAF患者182例,随机分为两组,其中91例患者应用利伐沙班抗凝归为观察组,余91例应用华法林抗凝归为对照组。随访记录两组抗凝期间血栓事件、出血等不良反应以及凝血功能和心功能情况。结果 观察组总出血事件显著低于对照组;观察组治疗后N末端B型利钠肽前体（NT proBNP）较对照组显著下降;COX风险模型发现,年龄及华法林是不良事件的危险因素。结论 利伐沙班对于合并心功能不全的高龄NVAF患者疗效不劣于华法林,但安全性显著高于华法林。     

Risks related with withholding and resuming anticoagulation in patients with non-variceal upper gastrointestinal bleeding while on warfarin therapy

Background: The use of warfarin is growing for the prevention or treatment of cardiovascular or cerebrovascular diseases. The risk of haemorrhagic side effects is increased in patients taking warfarin. Aims: To evaluate risks related with withholding and resuming anticoagulation in patients with upper gastrointestinal bleeding (UGIB) while on warfarin therapy and the role of the second‐look endoscopic examination (SEE). Methods: Records of 58 patients with native valvular heart diseases who presented with non‐variceal UGIB during chronic anticoagulation with warfarin were retrospectively reviewed. Age‐ and gender‐matched patients with non‐variceal UGIB during aspirin therapy because of ischaemic heart disease were recruited as the control group. Results: Development of both recurrent bleeding and thromboembolic events were more frequent in warfarin group than in control group (7.0% vs. 0% with p = 0.03 and 16.7% vs. 2.4% with p < 0.01, respectively). One of four cases of recurrent bleeding in warfarin group was found by SEE performed in an asymptomatic patient. There were six thromboembolic events which occurred on the 21st, 27th, 28th, 31st, 58th and 75th day from the presentation out of 36 patients who ceased anticoagulation. In contrast, only one from 41 in whom aspirin was discontinued experienced myocardial infarction. There was no difference in the failure of endoscopic haemostasis necessitating angiographic embolisation or surgery, hospital stay, the need of transfusion and overall mortality. Conclusions: Anticoagulation is recommended to be resumed before the 20th day from the cessation to prevent thromboembolic events. A routine SEE before resuming anticoagulation might be helpful to detect asymptomatic recurrent bleeding.     

Bleeding as an Outcome Among Patients With Nonvalvular Atrial Fibrillation in a Large Managed Care Population

Background

Patients with nonvalvular atrial fibrillation (NVAF) are at increased risk for stroke and bleeding events, but bleeding as an outcome has not been extensively studied in this patient population.

Objectives

The goal of this study was to estimate the incidence of bleeding events among patients with NVAF enrolled in managed care, investigate the relationships between bleeding incidence and bleeding and stroke risks, and estimate health care costs for patients who had a major bleeding event.

Methods

Adults with commercial insurance or Medicare Advantage coverage and health care claims related to AF between January 2005 and June 2009 but with no evidence of valvular disease were included in this retrospective claims data analysis. Baseline stroke risk (CHADS₂ [Congestive Heart Failure, Hypertension, Age >75 Years, Diabetes Mellitus, and Prior Stroke or Transient Ischemic Attack]) and bleeding risk (HAS-BLED [Hypertension, Abnormal Renal/Liver Function, Stroke, Bleeding History or Predisposition, Labile International Normalized Ratios, Elderly, Drugs/Alcohol]) were estimated. Bleeding events were identified during the variable follow-up period, which lasted from the date of the first qualifying AF visit until the earlier of death, disenrollment from the health plan, or June 30, 2010. Bleeding events were classified as major, serious nonmajor, or minor. Health care costs for patients with major bleeding events were calculated.

Results

Among 48,260 patients with NVAF (mean age, 67 years), 34% had an incident bleeding event during a mean (SD) follow-up period of 802 (540) days. Incidence rates for bleeding events of any severity and major events were 29.6 and 10.4 per 100 patient-years, respectively. Bleeding incidence rates increased with greater CHADS₂ and HAS-BLED risk scores. All-cause health care costs for patients during a major bleeding event averaged $16,830. Average costs per patient with a major event increased from approximately $52 per day in the prebleeding period to approximately $63 per day in the postbleeding period. Costs for patients who did not experience a major bleeding event averaged approximately $38 per day.

Conclusions

Bleeding incidence among patients with NVAF in a real-world setting was high and increased with greater stroke and bleeding risk scores. Health care costs for patients with major bleeding events were elevated. All rights reserved.     

Bleeding and thromboembolic risks of internal cardioversion for persistent atrial fibrillation

Although internal cardioversion is an effective method for converting AF the thromboembolic risk and bleeding complication of this procedure remains unclear. Retrospective analysis of the thromboembolic events and bleeding complications was performed in 114 consecutive patients (mean age 63 +/- 10 years, 91 men) who underwent internal cardioversion for persistent AF (mean AF duration 31 +/- 42 months). All patients received therapeutic warfarin for 3 weeks prior to the procedure, which was then replaced by periprocedural heparin therapy during internal cardioversion. The mean INR value before the internal cardioversion was 1.3 +/- 0.2. A total of 992 R wave synchronized shocks were delivered (mean 8.7 shocks/patients) without ventricular proarrhythmia. Successful internal cardioversion of AF was achieved in 100 (88%) patients. Two (2%) patients developed major bleeding complications with pericardial effusion after the procedure. Both of them subsided with conservative treatment. Minor bleeding with wound hematoma occurred in 8 (7%) of 114 after the procedure. Overall, there was no significant difference in the risk of procedure related complication between the use of heparin infusion and subcutaneous LMWH injection (6/78 [13%] vs 3/36 [8%], P = 0.9). No thromboembolic event was observed after 4 weeks of follow-up. In conclusion, the risk of thromboembolism after discontinuation of anticoagulation before the procedure is low in patients with persistent AF underwent internal cardioversion. However, the use of periprocedural heparin therapy was associated with an increase risk of procedure related bleeding complications.     

老年非瓣膜心房颤动及合并脑梗死住院患者抗栓治疗调查

目的探讨老年非瓣膜性心房颤动(NVAF)及合并脑梗死住院患者的临床特征、栓塞和出血风险及抗栓治疗现状。方法回顾分析583例非瓣膜老年心房颤动患者(≥60岁)的临床资料,分别应用CHADS2评分及CHA2DS2-VASc评分进行卒中风险分层,比较两种评分系统对卒中风险评估的差异,分析探讨各分层抗栓药物应用情况,并应用HAS-BLED评分进行出血风险评估。结果所有NVAF患者基础病中以高血压最常见占61.3%,其次冠心病占56.2%。583例NVAF心房颤动患者中,CHADS2评分≥2分351例,华法林用药率为2.6%,抗血小板用药率为82.3%,CHA2DS2-VASc评分≥2分522例,华法林用药率为2.1%,抗血小板用药率为84.7%。结论高血压和冠心病是大多数老年心房颤动患者的基础疾病,老年心房颤动及合并脑梗死患者应用华法林进行规范化抗凝治疗的比例低。     

Detecting nonvalvular atrial fibrillation and anticoagulant therapy in cardioembolic ischemic stroke

Nonvalvular Atrial fibrillation (NVAF) is the most common cardiac arrhythmia associated with an increase in risk of stroke and systemic thromboembolism. Strokes related to AF are associated with higher mortality, greater disability, longer hospital stays, and lower chance of being discharged home. The present review will focus on the current status of detecting NVAF and stroke prevention when there is AF. The CHA2DS2-VASc risk stratification scheme is discussed for the identification of patients who are at risk for thromboembolic stroke related to NVAF. Patient with a CHA2DS2-VASc score of 2 or greater are candidates for warfarin or a novel oral anticoagulant, irrespective of whether the strategy is for rate or rhythm control. Finally, guidelines and landmark clinical trials in NVAF patients with primary or secondary stroke prevention are discussed.     

Atrial Fibrillation and Anticoagulation in Patients with Permanent Pacemakers: Implications for Stroke Prevention

Several large prospective randomized trials have demonstrated that anticoagulation with warfarin reduces the risk of thromboembolic stroke in high risk patients with chronic AF by approximately 70%. Large numbers of patients with permanent pacemakers have AF, and anticoagulation rates in this population have not been described. In a prospective analysis of 110 consecutive patients attending the pacemaker clinic of a large university hospital, we assessed the number of patients with AF and the proportion of these patients who were receiving anticoagulation to prevent thromboembolic stroke. Where necessary, temporary pacemaker reprogramming to low ventricular rates was utilized to facilitate the diagnosis of AF. Fifty-three of the 110 patients (48%) were diagnosed with AF, all of whom (100%) had accepted high risk factors for thromboembolic stroke. Only eight of the 53 (15%) had been anticoagulated with warfarin. Thirty-six of the 53 patients (68%) diagnosed with AF had no prior documented diagnosis of chronic AF, and the majority had no symptoms suggesting AF. A single lead II ECG was insufficient in 67 of the 110 patients (61%) to diagnose the underlying atrial rhythm; the remainder required 12-lead ECGs or temporary pacemaker reprogramming to low ventricular rates to diagnose the underlying atrial rhythm. AF is common in patients with permanent pacemakers. It is commonly asymptomatic, and anticoagulation is markedly underutilized in reducing stroke risk in these patients. Attention to the possibility of AF in paced patients should allow prompt diagnosis and allow both the initiation of anticoagulation in order to reduce thromboembolic stroke risk and consideration for cardioversion of AF to sinus rhythm.     

Results of an open-label, prospective study of anticoagulant therapy for atrial fibrillation in an outpatient anticoagulation clinic

OBJECTIVE: The goal of this study was to investigate the complications and control of warfarin treatment in patients with nonvalvular atrial fibrillation (NVAF) newly referred to an outpatient anticoagulation clinic. METHODS: This study included new patients with NVAF who were referred to an anticoagulation clinic for warfarin therapy over a recruitment period of 21 months. To reflect real-world clinical practice, patient selection for anticoagulation and patient management were left to the referring physicians, who were blinded to their patients' participation in the study. Patients were interviewed in person at the first clinic visit and then by telephone every 4 to 6 weeks. They were questioned about any bleeding or thromboembolic events. RESULTS: A total of 402 patients were included (100% of all new referrals over 21 months). The mean (SD) age was 72.3 (10.3) years, and 224 (56%) patients were men. The mean (SD) international normalized ratio (INR) was 2.4 (0.31). Patients were followed up for a mean (SD) of 19 (8.1) months (range, 1.0-31.0 months). They spent a mean (SD) 66% (18.3) of time in the target range of INR (ie, 2.0-3.0). Annual event rates were 1.7% (95% CI, 0.4%-3.0%) for major bleeding, 16.6% (95% CI, 13.0%-20.2%) for minor bleeding, 1.2% (95% CI, 0.1%-2.3%) for ischemic stroke, and 0.3% (95% CI, 0.2%-0.8%) for transient ischemic attacks. There were no cases of hemorrhagic stroke or fatal bleeding. Variability of INR and number of medications were identified as risk factors for bleeding (P = 0.03 and P = 0.001, respectively). There was no significant association between age and bleeding. CONCLUSIONS: Based on this analysis, the risks of long-term oral anticoagulation therapy in an outpatient anticoagulation clinic appear to reflect the results of clinical trials. Rates of ischemic stroke, major bleeding, and anticoagulation control were comparable. There was no age-related risk of complications.     

Cardiac Resynchronization Therapy Device Implantation in Patients with Therapeutic International Normalized Ratios

Background: Many patients who need cardiac resynchronization therapy (CRT) require chronic anticoagulation. Current guidelines recommend discontinuation of warfarin and the initiation of anticoagulant “bridging” therapy during these procedures. We evaluated the safety of CRT‐device (CRT‐D) implantation without interruption of warfarin therapy. Methods: A total of 123 consecutive patients requiring CRT‐D therapy were enrolled, 49 identified as high risk for thromboembolic events who received either intravenous heparin, low molecular weight heparin, or warfarin therapy. The control group comprised 74 patients with low risk of thromboembolic events who required only cessation of warfarin perioperatively. Patients were evaluated at discharge and 15 and 30 days postoperatively for pocket hematomas, thromboembolic events, and bleeding. Patients’ length of stay was also catalogued. Results: Patients in the bridging arm had a significant increase in the rate of pocket hematomas (4.1%[control] vs 5.0%[warfarin] vs 20.7%[bridging], P = 0.03) and subsequent longer length of stay (1.6 ± 1.6 [control] vs 2.9 ± 2.7 [warfarin] vs 3.7 ± 3.2 [bridging], P < 0.001). Hematoma formation postoperatively was not different among patients undergoing an upgrade procedure versus those without preexisting cardiac rhythm devices (12% vs 6.2%, P = NS). Patients with a prosthetic mechanical mitral valve had a higher incidence of pocket hematoma formation (1.8% vs 20%, P = 0.03). Conclusions: Our findings suggest that implantation of CRT‐Ds without interruption of warfarin therapy in patients at high risk of thromboembolic events is a safe alternative to routine bridging therapy. This strategy is associated with reduced risk of pocket hematomas and shorter length of hospital stay. (PACE 2010; 400–406)     

高危老年非瓣膜性心房颤动患者抗凝治疗调查

目的 调查高危老年非瓣膜性心房颤动(NVAF)患者的抗凝治疗情况.方法 回顾性调查2008年1月至2010年1月住院治疗且CHADS2评分2分以上的198例NVAF患者,分析华法林的使用率和效果、未使用华法林原因,并探讨相应的对策.结果 (1)按指南须口服华法林者中,华法林实际应用仅51例(25.8%).服用阿司匹林107例,未抗栓40例.与无抗栓组、阿司匹林组对比,华法林组缺血性脑卒中发生率明显减少(P<0.05);与无抗栓组对比,阿司匹林组缺血性脑卒中发生率也减少(P<0.05).各组间出血、全因死亡情况均无统计学差异.(2)高龄、并存疾病多、用药复杂、肝肾功能障碍、认识不足及过分担心出血并发症是未使用华法林的主要原因.结论 目前必须加强高危老年NVAF患者长期抗凝治疗的管理.     

Asymptomatic embolisation in non-valvular atrial fibrillation and its relationship to anticoagulation therapy.

OBJECTIVE: A prospective study to determine the prevalence of embolic signals (ES) detected with transcranial Doppler ultrasound (TCD) in subjects with non-valvular atrial fibrillation (NVAF) who were and were not taking anticoagulants. METHODS: Sixty-eight subjects with NVAF taking warfarin and sixty-three NVAF subjects not taking warfarin were recruited from the cardiology department. No patients had suffered a previous stroke or clinical systemic embolisation. Thirty subjects in sinus rhythm were recruited as controls. TCD recordings were performed bilaterally from the middle cerebral arteries for 1 h. In 126 subjects (96%) the recording was repeated 1 week later. All Doppler recordings were reviewed blindly by an observer for ES. RESULTS: After one recording ES were detected in one warfarin NVAF subject (1.5%), four non-warfarin NVAF subjects (6.3%) and no sinus rhythm controls. Following both recordings ES were detected in one warfarin NVAF subject (1.5%) and seven non-warfarin NVAF subjects (11.9%) P = 0.017. CONCLUSIONS: Asymptomatic ES can be detected in a minority of subjects with NVAF. These signals were significantly less common in NVAF subjects taking warfarin, and this is consistent with asymptomatic ES having clinical significance in this disease. TCD detection of ES may have application in patients with NVAF for risk stratification, and assessment of the efficacy of new anti-platelet and anti-thrombotic regimens in the condition. Further large prospective studies are required to determine whether ES predict stroke risk in this patient group.     

心房颤动抗凝治疗进展

心房颤动是临床上最常见的心律失常,增加卒中风险。华法林抗凝效果虽已受到广泛的肯定,但同时存在出血风险、治疗窗狭窄、需要长期监测国际标准化比率以调整药量等缺点。新型口服抗凝药的应用如达比加群、利伐沙班、阿哌沙班可有效预防卒中及血栓栓塞。经皮左心耳封堵术亦可成为预防心房颤动血栓事件的有效替代治疗方式。     

Evaluation of antithrombotic usage for atrial fibrillation in aged care facilities

What is known and Objective: Warfarin is an important drug for the prevention of thromboembolic events such as stroke in patients with atrial fibrillation (AF). However, it is commonly implicated in major adverse drug events, which may result in reluctance to prescribe warfarin, especially in the institutionalised elderly population. This study aimed to assess the current trends in the antithrombotic management of AF in aged care facilities (ACFs) in Tasmania, Australia, and to compare this with current recommendations. Methods: We performed a non‐experimental, retrospective cohort study designed to evaluate antithrombotic usage for AF in ACF residents in Tasmania. Residents with AF were identified on a patient‐by‐patient basis from residential case‐note summaries collected from 29 of the 64 ACFs in Tasmania. The CHADS‐2 score and the presence or absence of documented contraindications were used to determine the appropriateness of the current antithrombotic therapy prescribed. Results and Discussion: Fifteen per cent (262/1737) of the ACF residents were diagnosed with AF. Nine per cent of the residents with AF had a contraindication to antithrombotic therapy (either warfarin or antiplatelet therapy). Eighty‐one per cent of residents were eligible for treatment with warfarin according to the CHADS‐2 score and did not appear to have a contraindication to warfarin. Of these, only 38% were prescribed warfarin; 16% (40/255) did not receive any antithrombotic treatment, despite being eligible for treatment with warfarin or antiplatelet therapy. Residents who did not receive any antithrombotic treatment or who received antiplatelet treatment were significantly older than those treated with anticoagulants. What is new and Conclusion: Our results indicate that antithrombotic therapy, particularly warfarin, is underused in ACF‐dwelling elderly patients who are eligible for treatment.     

d-dimer predicts major bleeding,cardiovascular events and all-cause mortality during warfarin treatment

ObjectivesPrevious studies have shown that biomarkers in blood plasma can predict bleeding complications during anticoagulant treatment as well as thromboembolic events and may improve existing risk stratification schemes in patients on or considered for oral anticoagulant treatment. The aim of this study was to investigate if levels of d-dimer, tissue plasminogen activator (tPA) and its complex with plasminogen inhibitor type 1 (tPA/PAI-1 complex) can predict major bleedings, cardiovascular events and all-cause mortality in patients with warfarin treatment.Design and methodsIn a longitudinal cohort study, 719 patients on oral anticoagulant treatment were followed for a total of 3001 treatment years. Major bleeding, stroke, arterial emboli, myocardial infarction and death were recorded and classified. Blood samples collected at baseline were analyzed for d-dimer, tPA, and tPA/PAI-1 complex.ResultsIn multivariate Cox regression analysis, high levels of d-dimer were associated with major bleeding (HR 1.27 per SD; 95% CI: 1.01–1.60), cardiovascular events (HR 1.23 per SD; 95% CI: 1.05–1.45) and all-cause mortality (HR 1.25 per SD; 95% CI: 1.06–1.47). Neither tPA nor the tPA/PAI-1 complex was associated with major bleeding, cardiovascular events or all-cause mortality.ConclusionWe conclude that high levels of d-dimer predict major bleeding, cardiovascular events and all-cause mortality during warfarin treatment.     

A review of anticoagulation with warfarin in patients with nonvalvular atrial fibrillation

BACKGROUND: Warfarin therapy has proved safe and effective in a number of randomized controlled trials of stroke prophylaxis in patients with nonvalvular atrial fibrillation (NVAF), reducing the risk of stroke in these patients by two thirds. However, participants in the clinical trials were carefully selected and younger than patients in actual clinical practice. OBJECTIVE: This analysis sought to determine whether the results of clinical trials in patients with NV can be extrapolated to the general population seen in clinical practice. METHODS: A MEDLINE search from 1966 to the present was used to identify observational trials of anticoagulation in patients with NVAF that addressed warfarin use, anticoagulation control, efficacy, and complications. The search terms used were atrial fibrillation and anticoagulation. RESULTS: Although warfarin prophylaxis against stroke in patients with NVAF appeared to be as well tolerated and effective in clinical practice as in clinical trials, it was generally underused, particularly in the elderly. Anticoagulation control was not as good in clinical practice as in clinical trials, although the rates of stroke and major bleeding were comparable. CONCLUSIONS: Judicious use of warfarin, tailored to individual stroke risk, seems to be a reasonable policy. Warfarin therapy increases quality-adjusted survival in patients at high risk for stroke, and it is recommended for medium-risk patients unless their risk of bleeding is high or their quality of life while taking warfarin would be poor. Patients at a low risk for stroke will have equivalent health outcomes and incur lower costs if treated with aspirin. Despite the increased risk of hemorrhage in elderly patients, the net benefit of warfarin therapy is greater in this age group because of the higher risk of stroke. Active involvement of patients and their caregivers in an anticoagulation service setting may improve outcomes of anticoagulation therapy.