Author + information
- Received January 10, 2015
- Revision received March 31, 2015
- Accepted April 14, 2015
- Published online June 30, 2015.
- Larry A. Allen, MD, MHS∗∗ (, )
- Gregg C. Fonarow, MD†,
- DaJuanicia N. Simon, MS‡,
- Laine E. Thomas, PhD‡,
- Lucas N. Marzec, MD∗,
- Sean D. Pokorney, MD, MBA‡,
- Bernard J. Gersh, MB, ChB, DPhil§,
- Alan S. Go, MD‖,
- Elaine M. Hylek, MD, MPH¶,
- Peter R. Kowey, MD#,
- Kenneth W. Mahaffey, MD∗∗,
- Paul Chang, MD††,
- Eric D. Peterson, MD, MPH‡,
- Jonathan P. Piccini, MD, MHS‡,
- ORBIT-AF Investigators
- ∗University of Colorado School of Medicine, Aurora, Colorado
- †University of California, Los Angeles, California
- ‡Duke Clinical Research Institute, Durham, North Carolina
- §Mayo Clinic College of Medicine, Rochester, Minnesota
- ‖Kaiser Permanente Northern California, Oakland, California
- ¶Boston University School of Medicine, Boston, Massachusetts
- #Lankenau Heart Institute and the Jefferson Medical College, Philadelphia, Pennsylvania
- ∗∗Stanford University School of Medicine, Palo Alto, California
- ††Janssen Scientific Affairs, Raritan, New Jersey
- ↵∗Reprint requests and correspondence:
Dr. Larry A. Allen, Division of Cardiology, University of Colorado School of Medicine, Academic Office 1, Room 7109, 12631 East 17th Avenue, Mail Stop B130, Aurora, Colorado 80045.
Background Although digoxin has long been used to treat atrial fibrillation (AF) and heart failure (HF), its safety remains controversial.
Objectives This study sought to describe digoxin use over time in patients with AF who were stratified by the presence or absence of HF, to characterize the predictors of digoxin use and initiation, and to correlate digoxin use with outcomes.
Methods Longitudinal patterns of digoxin use and its association with a variety of outcomes were assessed in a prospective outpatient registry conducted at 174 U.S. sites with enrollment from June 2010 to August 2011.
Results Among 9,619 patients with AF and serial follow-up every 6 months for up to 3 years, 2,267 (23.6%) received digoxin at study enrollment, 681 (7.1%) were initiated on digoxin during follow-up, and 6,671 (69.4%) were never prescribed digoxin. After adjusting for other medications, heart rate was 72.9 beats/min among digoxin users and 71.5 beats/min among nonusers (p < 0.0001). Prevalent digoxin use at registry enrollment was not associated with subsequent onset of symptoms, hospitalization, or mortality (in patients with HF, adjusted hazard ratio [HR] for death: 1.04; without HF, HR: 1.22). Incident digoxin use during follow-up was not associated with subsequent death in patients with HF (propensity adjusted HR: 1.05), but was associated with subsequent death in those without HF (propensity adjusted HR: 1.99).
Conclusions After adjustment for detailed clinical factors, digoxin use in registry patients with AF had a neutral association with outcomes under most circumstances. Because of the multiple conflicting observational reports about digoxin’s safety and possible concerns in specific clinical situations, a large pragmatic trial of digoxin therapy in AF is needed.
Cardiac glycosides, such as digoxin, have been used for decades to treat patients with atrial fibrillation (AF) and those with heart failure (HF) to slow atrioventricular nodal conduction and increase cardiac inotropy (1). With the development of alternative treatments for AF (2) and HF (3), as well as concerns about digoxin’s potential proarrhythmic properties and long-term effects on cardiac remodeling (4), prescribing digoxin has decreased and is no longer recommended as first-line therapy for either disease (3,5). However, there remain unmet needs for the treatment of many subgroups of AF patients, including those with HF, which has prompted calls for renewed use of digoxin in certain clinical situations (6).
Effectiveness and safety data for digoxin are relatively limited. The only large randomized trial of digoxin, the DIG (Digitalis Investigation Group) trial, showed no effect on mortality, but digoxin did reduce hospitalization among patients with heart failure and a reduced ejection fraction (HFrEF) (7). This trial only enrolled patients in sinus rhythm, was conducted between 1991 and 1993, and raised safety concerns at higher serum concentrations and in certain subgroups, including women (8–10). A more recent observational analysis of patients with incident HFrEF under routine care found that digoxin use was independently associated with higher mortality (11). There are currently no large randomized trials of digoxin in patients with AF. Two post-hoc nonrandomized analyses of data from the large AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) trial came to conflicting conclusions (12,13). Post-hoc analysis of other AF trials have shown higher mortality associated with digoxin use (14), as have real-world data from a large incident AF cohort from the Veterans Administration (15) and 2 large health maintenance organizations (16).
Due to limited and conflicting data, we set out to describe digoxin use over time among a large contemporary cohort of patients with AF who were stratified by the presence or absence of HF, to characterize predictors of digoxin use and initiation, and to clarify the association of digoxin use with heart rate, health-related quality of life (HRQOL) measures, hospitalization, and survival.
We used data from the ORBIT-AF (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation) study to assess the use of digoxin and its association with outcomes. Details of the ORBIT-AF study design have been published previously (17). Briefly, ORBIT-AF was a U.S.-based, prospective outpatient registry of AF conducted at 176 sites nationwide. The Duke Clinical Research Institute was responsible for ORBIT-AF site selection and study management. Eligible patients were 18 years of age and older with electrocardiographically confirmed AF. Enrolling providers included cardiologists, electrophysiologists, and primary care providers. Site personnel entered information on demographic characteristics, medical history, cardiovascular risk factors, AF management strategy, cardiac imaging, and provider characteristics into a standardized, web-based collection form. The presence or absence of HF and New York Heart Association (NYHA) functional class were determined at baseline by medical record review. Following initial enrollment, longitudinal information was collected during clinic visits at approximately 6-month intervals for up to 36 months, and included information on medication regimens, procedures, hospitalizations, quality of life, and vital status. We excluded patients who were missing information as to whether they were taking digoxin. Written informed consent was obtained from all study participants. The Duke Institutional Review Board approved the ORBIT-AF Registry; all participating sites obtained approval from local institutional review boards before entering patient data.
Medication use was collected prospectively at each study visit, including a field specific for digoxin. Dose and blood levels were not collected. The follow-up visit date at which digoxin was first reported was defined as the time period of initiation.
The primary outcome of interest was all-cause death. Additional outcomes of interest included heart rate, symptoms, HRQOL, all-cause hospitalization, and the composite of all-cause hospitalization and death. Symptoms were measured using the European Heart Rhythm Association (EHRA) score of AF-related symptoms (18). HRQOL was assessed by the Atrial Fibrillation Effect on Quality-of-Life questionnaire in a subset of patients at baseline, and at 12 and 24 months (19).
Characteristics between patients were described as frequency and percent for categorical variables and median (interquartile ranges) for continuous variables. The characteristics were compared using the chi-square test for categorical variables and the Wilcoxon rank sum test for continuous variables. The cohort was divided into those taking digoxin at study enrollment (prevalent use), those initiated on digoxin during follow-up (incident use), and those not on digoxin at any time during the study. Characteristics among the groups were compared using Pearson chi-square tests for categorical variables and Wilcoxon rank sum tests for continuous variables.
To examine factors associated with prevalent digoxin use, a multivariable hierarchical logistic regression model was constructed using backward selection for the binary outcome of digoxin use at baseline (see the covariates in Online Table 1 [54 pre-specified clinical and demographic characteristics and a random effect for the enrolling site] followed by the inclusion criterion of p < 0.05 final model covariates in Online Table 2a). Prevalent digoxin users at baseline were excluded from this model. Because digoxin use was measured at 6-month visit intervals, a second multivariable, discrete time Cox frailty model was constructed for the time to the first report of digoxin initiation (final model covariates in Online Table 2b). Patients were censored from the risk set when they were lost to follow-up (mainly due to staggered entry into the cohort). A third discrete time Cox frailty model was constructed for digoxin discontinuation among prevalent digoxin users (final model covariates in Online Table 2c). Results were presented as odds ratios (ORs) and/or hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) and p values.
Baseline heart rate was compared for baseline digoxin use using a linear regression model that accounted for other rate control medications (i.e., beta-blockers, verapamil, diltiazem, sotalol, and amiodarone). The adjusted mean heart rate was estimated by the model predicted heart rate, with and without digoxin, with the adjustment variables set equal to their population average.
Associations between prevalent digoxin use at baseline and subsequent all-cause death, all-cause hospitalization, cardiovascular hospitalization, and onset of symptoms were assessed in unadjusted and adjusted analysis. According to the pre-specified analysis plan, primary analyses were stratified a priori by the presence or absence of HF. Time to first reported symptoms was measured at 6-month visit intervals, and a discrete time Cox model was used for this outcome only; otherwise, exact event dates were used. The potential for clustering of patient outcomes within a site was handled by adding a random effect for site (multivariable Cox frailty model). Models were adjusted for all covariates listed in Online Table 3, which were determined to have: 1) particular clinical relevance, determined a priori; or 2) a statistically significant association with any of the outcomes under evaluation, as previously identified by backward selection with stay criteria of 0.05. The same set of covariates was used for adjustment of all outcomes. Adjusted associations for outcomes were displayed as HRs (95% CI).
Associations between incident digoxin use in follow-up and subsequent all-cause death, all-cause hospitalization, and cardiovascular hospitalization were assessed through propensity score matching between patients initiated on digoxin at follow-up. It was unusual to have subsequent follow-up in subjects who initiated digoxin at the final follow-up period (30 to 36 months); therefore, digoxin initiation was restricted to occur between 6 and 24 months. Analyses were conducted separately for patients with and without HF. Each case (incident digoxin use) was matched to 3 control subjects (noninitiators) using sequential stratification matching (20), which identified matches from the same point in follow-up at which digoxin was initiated and used all available covariate information up to that point (including HF status). The criteria for matching was a single propensity score, obtained from a logistic regression model for digoxin initiation. Matching was conducted sequentially, starting at 6 months and moving forward through follow-up. At each visit, patients who initiated digoxin were matched to other subjects who were still under follow-up at the same time, but who had not yet started digoxin. The criteria for identifying a match was “closeness” on a single propensity score value calculated at each visit. To be considered a match, patients had to have a difference in propensities no greater than a caliper of 20% of a SD. Standardized differences were used to evaluate the success of propensity matching at achieving balance. Outcomes assessment began immediately after the time period of initiation, and the model was fit using stratified Cox regression, stratified on the matched pair (21).
Pre-defined secondary analyses were performed in patient subgroups with renal function (estimated glomerular filtration rate [eGFR] <60 and ≥60 ml/min/1.73 m2) and left ventricular ejection fraction (LVEF) (<40% and ≥40%).
All candidate variables had <2% missing factors, except for level of education (4%), serum creatinine (7%), hematocrit (10%), LVEF (11%), and left atrial diameter (14%). Missing data were handled with single imputation. Imputed values were obtained by the Markov chain Monte Carlo method or by regression methods (22).
For all models, continuous variables were evaluated for nonlinearity with the outcome and when nonlinear fit with linear splines. All analyses were performed using SAS software version 9.3 (SAS Institute, Cary, North Carolina).
Patient characteristics and patterns of digoxin use
Between June 2010 and August 2011, 10,132 patients were enrolled in ORBIT-AF from 176 sites; 490 patients (4.8%) were then excluded due to lack of follow-up data, and 23 patients (0.2%) were excluded due to a missing response for digoxin use at baseline or follow-up. This resulted in a final cohort of 9,619 patients from 174 sites. Mean follow-up was 22 months (interquartile range: 17 to 25). Digoxin use was reported in 2,267 patients (23.6%) at the time of study enrollment, and an additional 681 patients (7.1%) were initiated on digoxin during follow-up, leaving 6,671 patients (69.4%) who were never on digoxin. Of those on digoxin at baseline, 794 patients (35.0%) discontinued digoxin during follow-up, and, of these, 217 (27.3%) subsequently resumed digoxin.
Baseline characteristics of the overall study population stratified by digoxin use are shown in Table 1. HF was present in 3,161 patients (32.9%) of the cohort, among whom prevalent digoxin use at baseline was present in 1,091 patients (34.5%). Incident use in follow-up was observed in another 268 patients (8.5%). Beta-blockers were prescribed in 69.6% of patients on digoxin compared with 62.4% of those never on digoxin; nondihydropyridine calcium channel blockers were prescribed in 18.6% of patients on digoxin compared with 16.0% of patients who never received digoxin. Antiarrhythmic medications were prescribed in 17.6% of patients on digoxin compared with 32.6% of patients who never received digoxin.
Factors independently associated with digoxin use at baseline included the following: rate control strategy and absence of previous ablation; permanent AF; worse HF functional class and LVEF; sinus node dysfunction; larger left atria; lower diastolic blood pressure; better renal function; faster heart rate; history of diabetes, hyperthyroidism, or chronic obstructive pulmonary disease; and female sex (the full model is presented in Online Table 2a). Multivariable predictors of the initiation of digoxin in follow-up were relatively similar to those for baseline use (the full model is shown in Online Table 2b). Multivariable baseline predictors of discontinuation during follow-up were the following: new-onset, paroxysmal, or persistent AF; bradycardia or tachycardia; no previous myocardial infarction; previous use of an antiarrhythmic drug; lower eGFR; higher EHRA score; and higher systolic blood pressure (the full model is shown in Online Table 2c).
The overall mean heart rate at baseline was 71.9 ± 13 beats/min, with a higher unadjusted heart rate (73.1 ± 12.7 beats/min) among prevalent digoxin users compared with those not on digoxin at baseline (71.5 ± 13.1 beats/min; p for comparison < 0.001). After adjustment for other rate control medications, the adjusted heart rate remained slightly higher among digoxin users (72.9 beats/min) than nonusers at baseline (71.5 beats/min; p < 0.0001).
Quality of life, symptoms, and outcomes
EHRA symptom scores were not significantly different among prevalent or incident digoxin patients versus patients not on digoxin (p = 0.09 and p = 0.58, respectively) (Table 1). In the quality-of-life substudy, unadjusted HRQOL was slightly lower in patients who received digoxin than those who were never on digoxin, both at baseline (Atrial Fibrillation Effect on Quality-of-Life questionnaire overall score median [interquartile range]: prevalent 78.7 [62.0 to 92.6]; incident 79.2 [64.8 to 88.9]; never 83.3 [68.5 to 93.5]; p = 0.0002) and at 1 year (prevalent 80.6 [66.7 to 91.7]; incident 79.6 [64.8 to 95.4]; never 86.6 [73.1 to 95.4]; p = 0.0001). The majority of prevalent digoxin patients who had follow-up were without symptoms, with an EHRA score of 0 in 51% to 53% of patients at 6, 12, 18, and 24 months. In multivariable analysis, digoxin use had a neutral association with the frequency of worsening AF symptoms in both patients with HF (Central Illustration, panel A) and patients without HF (Central Illustration, panel B).
Hospitalization for any cause occurred in 4,326 patients (45.0%), cardiovascular hospitalization in 2,485 patients (26.0%), and death in 865 patients (9.0%). Prevalent use of digoxin at registry enrollment among patients with and without HF was not associated with all-cause hospitalization, cardiovascular hospitalization, symptoms, or death (Central Illustration, panels A and B). Among patients with eGFR <60 ml/min/1.73 m2, digoxin use at enrollment had a borderline association with subsequent death (adjusted HR: 1.23; 95% CI: 1.00 to 1.51) and first all-cause hospitalization (adjusted HR: 1.14; 95% CI: 1.01 to 1.28).
Incident use of digoxin during follow-up among patients with and without HF after propensity matching (Online Tables 4a and 4b) was not associated with all-cause hospitalization, cardiovascular hospitalization, or symptoms (Central Illustration, panels C and D). Incident digoxin use was not associated with subsequent death in those with HF (adjusted HR: 1.05; 95% CI: 0.66 to 1.65) (Central Illustration, panel C), but was associated with subsequent death in those without HF (adjusted HR: 1.99; 95% CI: 1.12 to 3.56) (Central Illustration, panel D). Similarly, the association of incident digoxin use with death was confined to those with a LVEF >40% (adjusted HR: 2.21; 95% CI: 1.32 to 3.71). Among patients with eGFR <60 ml/min/1.73 m2, incident digoxin use was not associated with death (adjusted HR: 0.96; 95% CI: 0.51 to 1.82).
Within the context of multiple recent descriptions of prescribing digoxin in a variety of patient populations, the ORBIT-AF registry provides a broadly representative and clinically detailed look at the patterns of digoxin use in patients with existing AF. Despite the growing availability of alternative treatments for AF patients with HF, the prevalent use of digoxin in ORBIT-AF was 24% overall and even higher among patients with HF, lower blood pressure, higher heart rates, and female sex. Other contemporary data showed similar rates of digoxin use: 23% use among patients with incident AF (15) and 18% use among patients with incident HFrEF (11). In follow-up out to 3 years, 15% of ORBIT-AF patients either initiated or discontinued the drug, suggesting that use was dynamic. Digoxin was given with beta-blockers and other rate and rhythm control agents in the majority of patients.
Digoxin effectiveness and safety
These data add to the suboptimal body of evidence regarding the effectiveness and safety of digoxin in AF patients with and without HF. All positive cardiac inotropes that act through calcium handling and sensitization—with the notable exception of digoxin—have been shown in controlled trials to promote left ventricular remodeling and adverse events (23–26). The exception, the DIG trial, demonstrated equal survival in HF patients randomized to digoxin versus placebo (7). However, its conduct in the early 1990s, which predates most modern therapies for HFrEF, and its exclusion of patients with AF, leave open many questions. Since DIG, no high-quality, large, randomized trials of digoxin have been performed. Meanwhile, a variety of post hoc analyses of trial data have found digoxin to be neutral to harmful in certain subgroups. In SPORTIF III and V (Stroke Prevention Using an Oral Thrombin Inhibitor in Atrial Fibrillation) studies, digitalis use was 53%, and users had a higher mortality than nonusers (14). The AFFIRM study analysis, which incorporated time-dependent assessment of digoxin use, found an association between incident digoxin use and mortality (12,13).
Observational data from real-world practice have generally come to similar conclusions. In a large study of 2,891 Kaiser patients with newly diagnosed HFrEF, of whom 22.9% had AF, incident digoxin use was associated with higher mortality (HR: 1.72; 95% CI: 1.25 to 2.36), but there was no significant difference in the risk of HF hospitalization (HR: 1.05; 95% CI: 0.82 to 1.34) (11). The ATRIA-CVRN (Anticoagulation and Risk factors in Atrial Fibrillation-Cardiovascular Research Network) study, which was a propensity score matching analysis of 14,787 Kaiser patients with incident AF and without HF, found that incident digoxin use was independently associated with a higher risk of death (HR: 1.71; 95% CI: 1.52 to 1.93) and a higher risk of hospitalization (HR: 1.63; 95% CI: 1.56 to 1.71) (16). Recently, the TREAT-AF (Retrospective Evaluation and Assessment of Therapies in Atrial Fibrillation) study in 122,465 veterans with new-onset AF showed higher mortality in those treated with digoxin (multivariable HR: 1.26; 95% CI: 1.23 to 1.29; and propensity matching HR: 1.21; 95% CI: 1.17 to 1.25) (15).
Study advantages and limitations
The ORBIT-AF registry has advantages over these other studies. Unlike trial cohorts with narrow eligibility criteria and mandated systematic follow-up, ORBIT-AF captured a wide range of patients in routine care. Unlike the Kaiser and TREAT-AF studies, which relied on administrative coding for diagnoses and did not study LVEF and functional status measures, the prospective and rigorous clinical data captured in ORBIT-AF were more likely to limit misclassification and accurately measure potential confounders. This might be why the adjustment process in ORBIT-AF showed greater reduction in the unadjusted to adjusted HRs for digoxin in most of the analyses performed.
Obviously, unaccounted for treatment selection biases are likely to affect observational associations. The contradictory findings from different statistical analyses of the same AFFIRM database highlight this potential (12,13). Although extensive covariates were collected and modeled in this detailed prospective ORBIT-AF registry, including LVEF, NYHA functional class, vital signs, laboratory values, and concomitant medications, unmeasured reasons for starting digoxin might have at least partially confounded the association observed between digoxin and outcomes. Digoxin use is often dictated by hypotension, intolerance to more typical agents (e.g., beta-blockers), and worsening left ventricular dysfunction, all of which indicate worse disease. Digoxin is prescribed for AF largely for its ability to slow atrioventricular nodal conduction; yet, even after adjustment, digoxin was associated with higher heart rates, suggesting residual unmeasured differences or issues with medication adherence that were not measured.
Propensity matching, multivariable adjustment, and stratified analysis all help to reduce such confounding, but these methods do not completely address disease severity or many other facets of cardiac health. Ultimately, higher quality data from randomized trial designs that can remove treatment selection biases are necessary to definitively assess the effectiveness and safety of digoxin. Although another large randomized trial of digoxin like the DIG study is unlikely to be funded, evolving pragmatic clinical trial designs offer opportunities in the near future to test such a question through randomization of real-world practices (27). Because there are 33 million individuals with AF across the globe, determining whether or not digoxin is safe and effective should be a key priority in future clinical investigations.
Other limitations should be considered. ORBIT-AF participating sites were selected to be representative of the national AF population, but were not a true cross section, such that results might not be generalizable to all patients with AF; they also purposely did not represent patients outside the United States. Dose, serum digoxin concentration, and exact timing of and reasons for digoxin initiation or discontinuation were not collected.
Digoxin use remains common in the contemporary treatment of AF. Overall, after statistical adjustment for detailed clinical factors, digoxin had a neutral association with a wide range of outcomes. Because of ongoing questions about the safety of this commonly used medication, high-quality data derived from a pragmatic clinical trial of real-world contemporary digoxin use is greatly needed.
COMPETENCY IN MEDICAL KNOWLEDGE: Previous studies have suggested that digoxin use might be associated with adverse events in patients with AF; however, after adjustment for patient characteristics in a contemporary cohort, there were no significant interactions across a range of outcomes.
TRANSLATIONAL OUTLOOK: Mixed results regarding the safety of digoxin across multiple observational analyses call for higher quality evidence derived from pragmatic clinical trials of digoxin in patients with AF.
The authors would like to thank the staff and participants of the ORBIT-AF Registry for their important contributions to this work.
ORBIT-AF is sponsored by Janssen Scientific Affairs, LLC, Raritan, New Jersey. This project was supported in part by funding from the Agency of Healthcare Research and Quality through cooperative agreement number 1U19 HS021092. Dr. Allen is supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health (NIH) under Award Number K23HL105896. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Dr. Allen has served as a consultant for Amgen, Inc., Janssen Pharmaceuticals, and Novartis. Dr. Fonarow has received research grants or other research support from the NIH and the Agency for Healthcare Research and Quality; has consulted for Novartis, Amgen, Bayer, Gambro, Medtronic, and Janssen Pharmaceuticals; and has received support from the Ahmanson Foundation and the Corday Foundation. Dr. Pokorney has received educational grant support from AstraZeneca; research support from Gilead and Boston Scientific; and modest support from the advisory board of Janssen Pharmaceuticals. Dr. Kowey has been a paid consultant for Johnson & Johnson on multiple projects, including the development of rivaroxaban; has served as a consultant for the American College of Cardiology, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Cubist, Eli Lilly, Elsevier, Forest, GlaxoSmithKline, Medtronic, Omthera, Protola Pharma, and Spring Publishing; and has received research grants from Medtronic and St. Jude Medical. Dr. Peterson has received research funding from Janssen Pharmaceuticals, Sanofi, Genentech, Daiichi-Sankyo, Eli Lilly, and AstraZeneca; and has served as a consultant for Boehringer Ingelheim, AstraZeneca, and Janssen Pharmaceuticals. Dr. Piccini has received research grants from ARCA biopharma, Boston Scientific, Gilead, Janssen Pharmaceuticals, ResMed, St. Jude Medical; and has served as a consultant for Bayer, ChanRx, Johnson & Johnson, Medtronic, and Spectranetics. 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
- estimated glomerular filtration rate
- European Heart Rhythm Association
- heart failure
- heart failure with reduced ejection fraction
- hazard ratio
- health-related quality of life
- left ventricular ejection fraction
- odds ratio
- Received January 10, 2015.
- Revision received March 31, 2015.
- Accepted April 14, 2015.
- American College of Cardiology Foundation
- Gheorghiade M.,
- van Veldhuisen D.J.,
- Colucci W.S.
- Anderson J.L.,
- Halperin J.L.,
- Albert N.M.,
- et al.
- Yancy C.W.,
- Jessup M.,
- Bozkurt B.,
- et al.
- January C.T.,
- Wann L.S.,
- Alpert J.S.,
- et al.
- Ambrosy A.P.,
- Butler J.,
- Ahmed A.,
- et al.
- Ahmed A.,
- Rich M.W.,
- Love T.E.,
- et al.
- Freeman J.V.,
- Yang J.,
- Sung S.H.,
- Hlatky M.A.,
- Go A.S.
- Whitbeck M.G.,
- Charnigo R.J.,
- Khairy P.,
- et al.
- Gheorghiade M.,
- Fonarow G.C.,
- van Veldhuisen D.J.,
- et al.
- Gjesdal K.,
- Feyzi J.,
- Olsson S.B.
- Turakhia M.P.,
- Santangeli P.,
- Winkelmayer W.C.,
- et al.
- Freeman J.V.,
- Reynolds K.,
- Fang M.,
- et al.
- Brignole M.,
- Auricchio A.,
- Baron-Esquivias G.,
- et al.
- Spertus J.,
- Dorian P.,
- Bubien R.,
- et al.
- Austin P.C.
- Enders C.K.
- Packer M.,
- Colucci W.,
- Fisher L.,
- et al.