Heart Failure
Associations between atrial fibrillation and early outcomes of patients with heart failure and reduced or preserved ejection fraction Zubin J. Eapen, MD, MHS, a,b Melissa A. Greiner, MS, a Gregg C. Fonarow, MD, c Zhong Yuan, MD, PhD, d Roger M. Mills, MD, d Adrian F. Hernandez, MD, MHS, a,b and Lesley H. Curtis, PhD a,b Durham, NC; Los Angeles, CA; and Raritan, NJ
Background The relative impact of atrial fibrillation on early outcomes of patients with heart failure with reduced or preserved ejection fraction (EF) is unknown. Methods We conducted a retrospective cohort study of clinical registry data linked to Medicare claims for patients with heart failure with reduced or preserved EF stratified by presence of atrial fibrillation at admission. Outcomes of interest were allcause mortality and readmission at 30 days. We used Kaplan-Meier methods to estimate mortality and calculated cumulative incidence estimates of readmission. We used Cox proportional hazards models to examine associations between atrial fibrillation and 30-day outcomes. Results Among 66,357 patients admitted to 283 hospitals between January 2001 and March 2006, 46% had atrial fibrillation (44% of patients with reduced EF and 48% of patients with preserved EF). After adjustment for other patient characteristics, atrial fibrillation was associated with a modestly higher risk of 30-day mortality (HR, 1.08; 95% CI, 1.03-1.14) and readmission (HR, 1.06; 95% CI, 1.02-1.11). In subgroup analyses, atrial fibrillation was associated with a higher risk of 30-day mortality (HR, 1.16; 95% CI, 1.08-1.25) among patients with preserved EF but not among patients with reduced EF. The association of atrial fibrillation with readmission did not differ by heart failure type (P = .37 for the interaction). Conclusions Atrial fibrillation was associated with higher 30-day mortality among patients with heart failure with preserved EF but not reduced EF. The association of atrial fibrillation with 30-day readmission was modest and did not differ by heart failure type. (Am Heart J 2014;167:369-375.e2.) As a chronic disease punctuated by acute exacerbations, heart failure is a significant burden to patients, providers, and payers. It is the leading cause of hospital readmission among older patients in the United States. 1 In 2010, more than $39 billion was spent on the care of patients with heart failure in the United States alone, mostly on inpatient care. 2 To increase hospitals' accountability in reducing preventable readmissions, the US Centers for Medicare & Medicaid Services (CMS) publicly reports hospital-level rates of all-cause mortality within 30 days after heart failure admission and all-cause readmission within 30 days after discharge. Hospitals are now financially accountable, as From the aDuke Clinical Research Institute, Duke University School of Medicine, Durham, NC, bDepartment of Medicine, Duke University School of Medicine, Durham, NC, c Ahmanson-UCLA Cardiomyopathy Center, University of California, Los Angeles, Los Angeles, CA, and dJanssen Research & Development, LLC, Raritan, NJ. Jalal K. Ghali, MD, served as guest editor for this article. Submitted September 13, 2013; accepted December 4, 2013. Reprint requests: Zubin J. Eapen, MD, MHS, Duke Clinical Research Institute, PO Box 17969, Durham, NC 27715. E-mail: [email protected] 0002-8703/$ - see front matter © 2014, Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.ahj.2013.12.001
well; in October 2012, the federal Hospital Readmissions Reduction Program began penalizing hospitals with higher-than-predicted readmission rates by up to 1% of their total Medicare reimbursement. The heart failure population for which hospitals are accountable includes patients with reduced ejection fraction (EF) and patients with preserved EF. Heart failure with preserved EF is increasing rapidly in prevalence and accounts for a growing share of heart failure hospitalizations. 3 Although evidence-based therapies for heart failure with preserved EF are lacking, guidelines emphasize the management of comorbid conditions such as atrial fibrillation. 4,5 Atrial fibrillation has been an independent predictor of adverse outcomes in the overall heart failure population in some but not all studies. 6-9 Understanding the relative impact of atrial fibrillation on early postdischarge outcomes by heart failure type could identify high-risk patients who affect outcomes at the population level. Using the Acute Decompensated Heart Failure National Registry (ADHERE) linked with Medicare inpatient claims, we examined associations between atrial fibrillation and 30-day mortality and readmission among patients with reduced or preserved EF.
American Heart Journal March 2014
370 Eapen et al
Methods Data sources We obtained hospitalization data from the ADHERE-Core registry, which was established to study the characteristics, treatments, and inpatient outcomes of patients hospitalized with acute decompensated heart failure. 10 More than 300 community and academic centers in the United States participated in the registry and enrolled more than 185,000 patients between January 2001 and March 2006. To analyze long-term follow-up data, we obtained Medicare standard analytic claim files from CMS. These files contain encrypted identifiers that allow for longitudinal follow-up of fee-for-service Medicare beneficiaries. We used indirect identifiers to link registry records to Medicare files using methods that have been described previously. 11 Almost 80% of ADHERE records have been linked with Medicare claims, and older patients enrolled in ADHERE constitute a nationally representative sample of fee-for-service Medicare beneficiaries. 11,12 The institutional review board of the Duke University Health System approved the study.
Study population We included patients 65 years or older living in the United States who had a registry hospitalization linked to Medicare claims. If multiple registry hospitalizations were identified for a single patient, we used the earliest as the index hospitalization. We restricted the population to patients enrolled in fee-forservice Medicare at the time of the index hospitalization. To identify heart failure type, we required that EF was recorded in the registry. For all readmission analyses, we restricted the population to patients who were discharged alive to home.
Outcomes Consistent with metrics reported publicly by CMS, the primary outcomes of interest were all-cause mortality within 30 days after admission and all-cause readmission within 30 days after discharge. We identified all-cause mortality on the basis of death dates recorded in the Medicare denominator files. We defined readmission as any subsequent inpatient claim except those for transfers to or from another hospital and admissions for rehabilitation (diagnosis-related group 462 or International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis code V57.xx). We classified readmissions based on the primary diagnosis on the inpatient claim. Secondary outcomes of interest were readmission for ischemic stroke (ICD9-CM code 433.x1 or 434.x1) 13 and readmission for heart failure (ICD-9-CM code 428.x, 402.x1, 404.x1, or 404.x3).
Covariates Variables from the registry included demographic characteristics, medical history, initial evaluation results, initial vital signs, laboratory test results, hospitalization dates, and discharge medications. We used quantitative and qualitative EF values to define heart failure with preserved EF (ie, EF of 40% or greater or normal/mildly impaired) and heart failure with reduced EF (ie, EF less than 40% or moderately/severely impaired). Variables had low rates of missingness (ie, less than 5% of records). Consistent with the approach used in the development of 30-day risk models for patients with heart failure, we imputed continuous variables to the overall median value and dichotomous variables to “no.” 14-16
We defined atrial fibrillation on the basis of a medical history of atrial fibrillation documented in the registry or any diagnosis of atrial fibrillation (ICD-9-CM code 427.31) or atrial flutter (ICD-9CM code 427.32) on the index hospitalization claim.
Subgroups We combined atrial fibrillation diagnosis and heart failure type to define 4 subgroups: preserved EF with or without atrial fibrillation, and reduced EF with or without atrial fibrillation.
Statistical analysis We present baseline characteristics of the study population by subgroup. We present categorical variables as frequencies and continuous variables as means with SDs. We tested for differences in baseline variables between groups using χ 2 tests for categorical variables and Kruskal-Wallis tests for continuous variables. We describe unadjusted 30-day event rates by subgroup. We used Kaplan-Meier methods to estimate mortality and log-rank tests to assess differences in mortality between groups. For allcause readmission, we calculated cumulative incidence estimates to account for the competing risk of mortality. We tested for differences in all-cause readmission between groups using Gray tests. For each subgroup, we identified the 5 most common causes of readmission within 30 days by grouping primary diagnosis codes into diagnostic categories including heart failure (as previously defined), acute myocardial infarction (ICD-9-CM code 410.x1), and additional categories based on clinical judgment. To examine adjusted associations between heart failure type, atrial fibrillation, and other patient characteristics with 30-day mortality and readmission, we used Cox proportional hazards models with robust standard errors to account for clustering of similar patients within hospitals. In multivariable models, we included variables for atrial fibrillation, reduced EF, age, sex, race, anemia, chronic obstructive pulmonary disease, chronic renal insufficiency, coronary artery disease, diabetes mellitus, hyperlipidemia, hypertension, peripheral vascular disease, myocardial infarction, stroke or transient ischemic attack, dyspnea, fatigue, rales, pulmonary edema, heart rate, systolic blood pressure, serum creatinine, serum sodium, hemoglobin, and year of index hospitalization. We selected covariates on the basis of clinical judgment, published literature, and availability in the registry. Consistent with prior studies, we included a variable for index length of stay greater than 7 days in the readmission models only. 17 We tested for an interaction between reduced EF and atrial fibrillation to examine whether the effect of atrial fibrillation differed by heart failure type and repeated the analysis within each heart failure type. We censored data for patients if they enrolled in Medicare managed care and at the time of death. In all analyses, we tested the proportionality assumption for the atrial fibrillation variable of interest. We considered a 2-sided P b .05 to be statistically significant for all tests. We used SAS software version 9.3 (SAS Institute Inc, Cary, North Carolina) for all analyses. This study was supported by a research agreement between Duke University and Johnson & Johnson.
American Heart Journal Volume 167, Number 3
Eapen et al 371
Table I. Baseline characteristics of the study population by type of heart failure and presence of atrial fibrillation Heart failure with reduced ejection fraction
Characteristic Age, mean (SD), y Male, No. (%) Race, No. (%) White Black Other/unknown Medical history, No. (%) Anemia Chronic obstructive pulmonary disease Chronic renal insufficiency Coronary artery disease Diabetes mellitus Hyperlipidemia Hypertension Myocardial infarction Peripheral vascular disease Stroke or transient ischemic attack CHADS2 score, mean (SD) Initial test results, No. (%) Hemoglobin b9 g/dL Hemoglobin 9–11 g/dL Serum creatinine N2.0 mg/dL Systolic blood pressure N150 mm Hg Discharge medications, No. (%) ACE inhibitor or ARB Aspirin β-Blocker Diuretic Clopidogrel Lipid-lowering agent Warfarin Index hospitalization length of stay Length of stay, mean (SD), d Length of stay N7 d, No. (%)
Without atrial fibrillation (n = 16,628)
With atrial fibrillation (n = 13,152)
77.6 (7.6) 9098 (54.7)
78.8 (7.3) 8006 (60.9)
12,825 (77.1) 2532 (15.2) 1271 (7.6)
11,360 (86.4) 983 (7.5) 809 (6.2)
Heart failure with preserved ejection fraction
P b.001 b.001 b.001
Without atrial fibrillation (n = 19,172)
With atrial fibrillation (n = 17,405)
78.9 (8.0) 6313 (32.9)
80.8 (7.4) 6415 (36.9)
14,910 (77.8) 2745 (14.3) 1517 (7.9)
15,161 (87.1) 1108 (6.4) 1136 (6.5)
P b.001 b.001 b.001
8372 4597 4778 11,399 7150 6842 11,742 6778 3350 2835 3.1
(50.3) (27.6) (28.7) (68.6) (43.0) (41.1) (70.6) (40.8) (20.1) (17.0) (1.2)
6551 3760 4154 9158 4755 5139 8872 5110 2617 2566 3.1
(49.8) (28.6) (31.6) (69.6) (36.2) (39.1) (67.5) (38.9) (19.9) (19.5) (1.2)
.36 .07 b.001 .046 b.001 b.001 b.001 b.001 .59 b .001 .15
11,183 6003 5254 10,492 9030 7556 15,497 4871 3626 3494 3.3
(58.3) (31.3) (27.4) (54.7) (47.1) (39.4) (80.8) (25.4) (18.9) (18.2) (1.1)
9505 5439 4210 9291 6170 5988 13,198 4030 3196 3484 3.3
(54.6) (31.2) (24.2) (53.4) (35.4) (34.4) (75.8) (23.2) (18.4) (20.0) (1.1)
b .001 .90 b.001 .01 b.001 b.001 b.001 b.001 .18 b .001 .03
616 3410 3232 5554
(3.7) (20.5) (19.4) (33.4)
394 2636 2620 3081
(3.0) (20.0) (19.9) (23.4)
b.001 .32 .30 b.001
1379 5473 3534 9906
(7.2) (28.5) (18.4) (51.7)
945 4543 2484 6769
(5.4) (26.1) (14.3) (38.9)
b.001 b.001 b.001 b.001
11,223 9388 10,740 13,677 2869 6578 2050
(67.5) (56.5) (64.6) (82.3) (17.3) (39.6) (12.3)
8416 5648 8127 10,857 1424 4507 6459
(64.0) (42.9) (61.8) (82.6) (10.8) (34.3) (49.1)
b.001 b.001 b.001 .50 b.001 b.001 b.001
11,184 9810 10,785 14,982 3129 7214 1369
(58.3) (51.2) (56.3) (78.1) (16.3) (37.6) (7.1)
9397 6595 9498 14,020 1722 5132 8254
(54.0) (37.9) (54.6) (80.6) (9.9) (29.5) (47.4)
b.001 b.001 .001 b.001 b.001 b.001 b.001
6.6 (6.0) 3791 (28.8)
b.001 b.001
6.3 (5.5) 4570 (26.3)
b.001 b.001
5.5 (4.7) 3526 (21.2)
5.4 (4.6) 3779 (19.7)
ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker.
Results We identified 80,416 patients with a linked registry record who were 65 years or older living in the United States with fee-for-service Medicare coverage at the time of hospital admission. We excluded 14,059 patients for whom EF was not recorded in the registry. Patients who did not have EF recorded had similar baseline characteristics and prevalence of comorbid atrial fibrillation compared with patients who did have EF recorded (the online Appendix Supplementary Table I). The final study population included 66,357 patients admitted to 283 hospitals between January 2001 and March 2006. Of these, 44.9% had reduced EF and 55.1% had preserved EF. Atrial fibrillation was present in 46.0% of all patients, including 44.2% of patients with reduced EF and 47.6% of patients with preserved EF.
Table I shows the baseline characteristics of the study population. The subgroup of patients with atrial fibrillation was slightly older and included more men and more white patients. Comorbid conditions were common. A history of stroke or transient ischemic attack was more frequent among patients with atrial fibrillation. Index hospitalization length of stay was greater among patients with atrial fibrillation. There were 47,447 patients discharged alive to home who were included for all readmission analyses. Comparisons of baseline characteristics by subgroup were similar to the full study population (the online Appendix Supplementary Table II). Table II and the Figure show that patients with atrial fibrillation had significantly worse unadjusted rates of mortality and readmission. Rates of ischemic stroke at 30 days were low among patients with and without atrial
American Heart Journal March 2014
372 Eapen et al
Table II. Cumulative incidence of mortality within 30 days after admission and cumulative incidence of readmission within 30 days after discharge⁎ Heart failure with reduced ejection fraction
Outcome Mortality All-cause readmission Heart failure readmission Ischemic stroke readmission
Without atrial fibrillation
With atrial fibrillation
1579/16,628 (9.5) 2766/12,415 (22.3) 1271/12,415 (10.2) 30/12,415 (0.2)
1502/13,152 (11.4) 2228/9275 (24.0) 1033/9275 (11.1) 22/9275 (0.2)
P b.001 .003 .04 .95
Heart failure with preserved ejection fraction Without atrial fibrillation
With atrial fibrillation
1218/19,172 (6.4) 2800/13,876 (20.2) 867/13,876 (6.3) 26/13,876 (0.2)
1606/14,405 (9.2) 2547/11,881 (21.4) 890/11,881 (7.5) 35/11,881 (0.3)
P b.001 .01 b.001 .08
⁎ Mortality values are expressed as number of events (Kaplan-Meier percentage). Readmission values are expressed as number of events (cumulative incidence percentage). Readmission was calculated among patients discharged alive only.
Figure
A
Cumulative Incidence, %
20%
HF−PEF+No AF HF−PEF+AF HF−REF+No AF HF−REF+AF
10%
0% 0
5
10
15
20
25
30
Days from Index Hospitalization Admission
B
Cumulative Incidence, %
30%
HF−PEF+No AF HF−PEF+AF HF−REF+No AF HF−REF+AF
20%
fibrillation. Among all patients readmitted within 30 days, heart failure was the most frequent reason for the first readmission, regardless of the type of heart failure or the presence of atrial fibrillation (Table III). Among patients with preserved EF and atrial fibrillation, atrial fibrillation was the third most frequent reason for readmission. Among patients with reduced EF and atrial fibrillation, atrial fibrillation was not among the 5 most frequent reasons for 30-day readmission. After adjustment for other patient factors, atrial fibrillation was associated with a modestly higher hazard of 30-day mortality (HR, 1.08; 95% CI, 1.03-1.14) and 30-day readmission (HR, 1.06; 95% CI, 1.02-1.11) in the overall study population (Table IV). The interaction between heart failure with reduced EF and atrial fibrillation in the mortality model indicated a differential effect of atrial fibrillation by type of heart failure (P b .001 for the interaction). After stratifying by heart failure type, the association of atrial fibrillation with higher risk of 30-day mortality was significant among patients with preserved EF (HR, 1.16; 95% CI, 1.08-1.25) but not patients with reduced EF (HR, 1.00; 95% CI, 0.941.08). The association of atrial fibrillation with 30-day readmission did not differ by heart failure type (P = .37 for the interaction). The proportionality assumption was met for the atrial fibrillation variable in all models.
Discussion 10%
0% 0
5
10
15
20
25
30
Days from Index Hospitalization Discharge Cumulative incidence of (A) mortality within 30 days of admission and (B) readmission within 30 days of discharge.
Through the Hospital Readmissions Reduction Program, CMS expects to recover $280 million in 2013 from financial penalties imposed on 2217 hospitals. 18 As the leading cause of hospitalization among older patients, heart failure presents a particular challenge. The impact of comorbid conditions like atrial fibrillation on postdischarge outcomes may differ according to heart failure type. In this study of a large nationwide registry of heart failure hospitalizations linked with Medicare data, we found that atrial fibrillation was modestly associated with 30-day mortality and readmission. After stratification by heart failure type, atrial fibrillation was associated with
American Heart Journal Volume 167, Number 3
Eapen et al 373
Table III. Five most frequent reasons for the first readmission among patients readmitted within 30 days after discharge⁎ Heart failure with reduced ejection fraction
Heart failure with preserved ejection fraction
Primary diagnosis frequency
Without atrial fibrillation (n = 2766)
With atrial fibrillation (n = 2228)
Without atrial fibrillation (n = 2800)
With atrial fibrillation (n = 2547)
Acute myocardial infarction Acute renal failure Atrial fibrillation Coronary atherosclerosis Heart failure Pneumonia Volume depletion
106 (3.8) 90 (3.3) — 135 (4.9) 1207 (43.6) 101 (3.7) —
53 (2.4) 96 (4.3) — 58 (2.6) 981 (44.0) 86 (3.9) —
87 (3.1) 112 (4.0) — 162 (5.8) 802 (28.6) 121 (4.3) —
— — 93 (3.7) 67 (2.6) 830 (32.6) 141 (5.5) 63 (2.5)
⁎ Values are expressed as number (percentage).
Table IV. Associations between atrial fibrillation and 30-day mortality and readmission All-cause mortality⁎ Heart failure type Any heart failure Heart failure with reduced ejection fraction Heart failure with preserved ejection fraction
Adjusted HR (95% CI) 1.08 (1.03-1.14) 1.00 (0.94-1.08) 1.16 (1.08-1.25)
All-cause readmission† P .003 .92 b.001
Adjusted HR (95% CI) 1.06 (1.02-1.11) 1.05 (0.99-1.11) 1.07 (1.01-1.14)
P .004 .12 .02
⁎ Adjusted for atrial fibrillation, reduced ejection fraction, age, sex, race, anemia, chronic obstructive pulmonary disease, chronic renal insufficiency, coronary artery disease, diabetes mellitus, hyperlipidemia, hypertension, peripheral vascular disease, myocardial infarction, stroke or transient ischemic attack, dyspnea, fatigue, rales, pulmonary edema, heart rate, systolic blood pressure, serum creatinine, serum sodium, hemoglobin, and year of index hospitalization. † Adjusted for the above variables and a variable for index length of stay greater than 7 days.
30-day mortality among patients with preserved EF but not patients with reduced EF. In addition, we found that heart failure was the most common cause of readmission among patients with either type of heart failure, regardless of the presence of comorbid atrial fibrillation. These findings suggest that while management of heart failure should be the primary focus of efforts to improve 30-day outcomes, targeting improved management of atrial fibrillation may be of benefit, particularly in patients with preserved EF. Atrial fibrillation may serve as a marker of illness severity in patients with heart failure, a result of increased left atrial pressure. The lack of atrial kick and rapid ventricular response associated with atrial fibrillation may also promote development of decompensated heart failure. Atrial fibrillation is a common comorbid condition among patients with heart failure; up to 40% of patients with advanced heart failure have atrial fibrillation. 7,19-21 Prevalence of both conditions is higher in older patients like Medicare beneficiaries. 22,23 Studies of the impact of atrial fibrillation on patients with either type of heart failure have yielded conflicting results. Some studies have found atrial fibrillation to be an independent predictor of mortality in patients with reduced EF, whereas others have not. 24-28 Results are similarly mixed in analyses of patients with preserved EF. 29-33 Our study is among the first to evaluate the impact of atrial fibrillation on 30-day
mortality and readmission among patients with either type of heart failure. Our study builds upon previous findings that atrial fibrillation is associated with inhospital mortality among patients with either reduced or preserved EF. 6 Our findings reveal a modest association between atrial fibrillation and 30-day mortality and readmission overall and, after stratification by heart failure type, among patients with preserved EF. Although clinicians should optimize the management of atrial fibrillation, our findings suggest that this is unlikely to considerably influence publicly reported 30-day outcomes in patients with reduced EF. Although administratively coded the same as reduced EF, preserved EF represents a different type of heart failure, which can require different management strategies and present different cost burdens. Identifying ways to improve postdischarge outcomes for patients with preserved EF is of particular importance. In-hospital mortality has improved for patients with preserved EF, but temporal trends in hospitalization for heart failure suggest that preserved EF may surpass reduced EF as the most prevalent form of heart failure. 3 Consequently, preserved EF may increasingly influence aggregate readmission rates. Management of comorbid conditions will remain important for patients with preserved EF, but therapies that directly address preserved EF to improve postdischarge outcomes are needed.
American Heart Journal March 2014
374 Eapen et al
Our study has some limitations. Administrative claims are subject to incomplete or incorrect coding that may affect outcome measurement. Rather than complete case analysis, we imputed missing values so that the final study population was not overly restrictive and because variables had low rates of missingness (ie, less than 5% of records). Hospitals that participate in ADHERE may be more likely to have better adherence to performance measures compared with other hospitals. Data were restricted to patients 65 years and older in the ADHERE registry, which limited our ability to generalize the findings to the broader population of patients with atrial fibrillation and heart failure. However, the study population is relevant to hospitals accountable for the care of Medicare beneficiaries through the Hospital Readmissions Reduction Program. Several clinical variables that are likely to be associated with clinical outcomes were not available, including New York Heart Association functional classification, symptom severity, frailty, and sinus rhythm during or after hospitalization. We could not account for socioeconomic status, education level, and health literacy. Residual or unmeasured confounding may have contributed to these findings. In conclusion, atrial fibrillation had a modest association with all-cause mortality within 30 days after admission and with all-cause readmission within 30 days after discharge among patients with heart failure. After stratification by heart failure type, atrial fibrillation was associated with 30-day mortality in patients with preserved EF but not in patients with reduced EF. The association of atrial fibrillation with 30-day readmission did not differ by heart failure type. Whether improved treatment of patients with atrial fibrillation reduces 30day readmission and mortality rates, particularly in patients with preserved EF, requires further study.
Acknowledgements Additional Contributions: Laura G. Qualls, MS, and Damon M. Seils, MA, Duke University, provided editorial assistance and prepared the manuscript.
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
References 1. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009; 360:1418-28. 2. Lloyd-Jones D, Adams RJ, Brown TM, et al. Heart disease and stroke statistics–2010 update: a report from the American Heart Association. Circulation 2010;121:e46-215. 3. Steinberg BA, Zhao X, Heidenreich PA, et al. Trends in patients hospitalized with heart failure and preserved left ventricular ejection fraction: prevalence, therapies, and outcomes. Circulation 2012;126:65-75. 4. Hunt SA, Abraham WT, Chin MH, et al. 2009 focused update incorporated into the ACC/AHA 2005 Guidelines for the Diagnosis and Management of Heart Failure in Adults: a report of the American
16.
17.
18. 19. 20.
College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. Circulation 2009;119:e391-479. Shah SJ, Gheorghiade M. Heart failure with preserved ejection fraction: treat now by treating comorbidities. JAMA 2008;300:431-3. Mountantonakis SE, Grau-Sepulveda MV, Bhatt DL, et al. Presence of atrial fibrillation is independently associated with adverse outcomes in patients hospitalized with heart failure: an analysis of get with the guidelines-heart failure. Circ Heart Fail 2012;5:191-201. Wang TJ, Larson MG, Levy D, et al. Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality: the Framingham Heart Study. Circulation 2003;107: 2920-5. Dries DL, Exner DV, Gersh BJ, et al. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: a retrospective analysis of the SOLVD trials. Studies of Left Ventricular Dysfunction. J Am Coll Cardiol 1998;32: 695-703. Kober L, Swedberg K, McMurray JJ, et al. Previously known and newly diagnosed atrial fibrillation: a major risk indicator after a myocardial infarction complicated by heart failure or left ventricular dysfunction. Eur J Heart Fail 2006;8:591-8. Adams Jr KF, Fonarow GC, Emerman CL, et al. Characteristics and outcomes of patients hospitalized for heart failure in the United States: rationale, design, and preliminary observations from the first 100,000 cases in the Acute Decompensated Heart Failure National Registry (ADHERE). Am Heart J 2005;149:209-16. Hammill BG, Hernandez AF, Peterson ED, et al. Linking inpatient clinical registry data to Medicare claims data using indirect identifiers. Am Heart J 2009;157:995-1000. Kociol RD, Hammill BG, Fonarow GC, et al. Generalizability and longitudinal outcomes of a national heart failure clinical registry: comparison of Acute Decompensated Heart Failure National Registry (ADHERE) and non-ADHERE Medicare beneficiaries. Am Heart J 2010;160:885-92. Wahl PM, Rodgers K, Schneeweiss S, et al. Validation of claimsbased diagnostic and procedure codes for cardiovascular and gastrointestinal serious adverse events in a commercially-insured population. Pharmacoepidemiol Drug Saf 2010;19:596-603. Keenan PS, Normand SL, Lin Z, et al. An administrative claims measure suitable for profiling hospital performance on the basis of 30-day all-cause readmission rates among patients with heart failure. Circ Cardiovasc Qual Outcomes 2008;1:29-37. Krumholz HM, Wang Y, Mattera JA, et al. An administrative claims model suitable for profiling hospital performance based on 30-day mortality rates among patients with heart failure. Circulation 2006; 113:1693-701. Harrell FE. Multivariable modeling strategies. Regression modeling strategies: with applicaions to linear models, logistic regression, and survival analysis. New York: Springer-Verlag; 2001. p. 53-85. Krumholz HM, Parent EM, Tu N, et al. Readmission after hospitalization for congestive heart failure among Medicare beneficiaries. Arch Intern Med 1997;157:99-104. Joynt KE, Jha AK. A path forward on Medicare readmissions. N Engl J Med 2013;368:1175-7. Stevenson WG, Stevenson LW. Atrial fibrillation and heart failure— five more years. N Engl J Med 2004;351:2437-40. Koelling TM, Chen RS, Lubwama RN, et al. The expanding national burden of heart failure in the United States: the influence of heart failure in women. Am Heart J 2004;147:74-8.
American Heart Journal Volume 167, Number 3
21. Maisel WH, Stevenson LW. Atrial fibrillation in heart failure: epidemiology, pathophysiology, and rationale for therapy. Am J Cardiol 2003;91:2D-8D. 22. Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study. JAMA 2001;285:2370-5. 23. Naccarelli GV, Varker H, Lin J, et al. Increasing prevalence of atrial fibrillation and flutter in the United States. Am J Cardiol 2009;104:1534-9. 24. Dries D, Exner D, Gersh B, et al. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: a retrospective analysis of the SOLVD trials. J Am Coll Cardiol 1998;32:695-703. 25. Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance of atrial fibrillation in advanced heart failure. A study of 390 patients. Circulation 1991;84:40-8. 26. Swedberg K, Olsson LG, Charlesworth A, et al. Prognostic relevance of atrial fibrillation in patients with chronic heart failure on longterm treatment with beta-blockers: results from COMET. Eur Heart J 2005;26:1303-8. 27. Carson PE, Johnson GR, Dunkman WB, et al. The influence of atrial fibrillation on prognosis in mild to moderate heart failure. The V-HeFT Studies. The V-HeFT VA Cooperative Studies Group. Circulation 1993;87:VI102-10.
Eapen et al 375
28. Crijns HJ, Tjeerdsma G, de Kam PJ, et al. Prognostic value of the presence and development of atrial fibrillation in patients with advanced chronic heart failure. Eur Heart J 2000;21: 1238-45. 29. Owan TE, Hodge DO, Herges RM, et al. Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med 2006;355:251-9. 30. Lenzen MJ, Scholte op Reimer WJ, Boersma E, et al. Differences between patients with a preserved and a depressed left ventricular function: a report from the EuroHeart Failure Survey. Eur Heart J 2004;25:1214-20. 31. Grigorian Shamagian L, Roman AV, Seara JG, et al. Atrial fibrillation in patients hospitalized for congestive heart failure: the same prognostic influence independently of left ventricular systolic function? Int J Cardiol 2006;110:366-72. 32. Olsson LG, Swedberg K, Ducharme A, et al. Atrial fibrillation and risk of clinical events in chronic heart failure with and without left ventricular systolic dysfunction: results from the Candesartan in Heart failure-Assessment of Reduction in Mortality and morbidity (CHARM) program. J Am Coll Cardiol 2006;47: 1997-2004. 33. Rusinaru D, Leborgne L, Peltier M, et al. Effect of atrial fibrillation on long-term survival in patients hospitalised for heart failure with preserved ejection fraction. Eur J Heart Fail 2008;10: 566-72.
American Heart Journal Volume 167, Number 3
Eapen et al 375.e1
Appendix Supplementary Table I. Characteristics of the Study Population by Ejection Fraction Status Characteristic Age, mean (SD), y Male, No. (%) Race, No. (%) Black White Other/unknown Medical history, No. (%) Atrial fibrillation Anemia Chronic obstructive pulmonary disease Chronic renal insufficiency Coronary artery disease Diabetes mellitus Hyperlipidemia/dyslipidemia Hypertension Peripheral vascular disease Myocardial infarction Stroke or transient ischemic attack CHADS2 score, mean (SD) Initial test results, No. (%) Hemoglobin b9 g/dL Hemoglobin 9–11 g/dL Serum creatinine N2.0 mg/dL Systolic blood pressure N150 mm Hg Discharge medications, No. (%) ACE inhibitor or ARB Aspirin β-Blocker Diuretic Clopidogrel Lipid-lowering Warfarin Index hospitalization length of stay Length of stay, mean (SD), d Length of stay N7 days, No. (%)
Missing Ejection Fraction (n = 14,059)
Not Missing Ejection Fraction (n = 66,357)
P Value b.001 b.001 b.001
80.4 (8.2) 5501 (39.1)
79.0 (7.7) 29,832 (45.0)
1796 (12.8) 11,187 (79.6) 1076 (7.7)
7368 (11.1) 54,256 (81.8) 4733 (7.1)
6064 7802 4294 3912 7957 5665 3989 10,132 2417 3575 2677 3.2
(43.1) (55.5) (30.5) (27.8) (56.6) (40.3) (28.4) (72.1) (17.2) (25.4) (19.0) (1.1)
30,557 35,611 19,799 18,396 40,340 27,105 25,525 49,309 12,789 20,789 12,379 3.2
(46.0) (53.7) (29.8) (27.7) (60.8) (40.8) (38.5) (74.3) (19.3) (31.3) (18.7) (1.1)
b.001 b.001 .10 .80 b.001 .23 b.001 b.001 b.001 b.001 .29 .18
928 3609 2848 5481
(6.6) (25.7) (20.3) (39.0)
3334 16,062 11,870 25,310
(5.0) (24.2) (17.9) (38.1)
b.001 b.001 b.001 .06
6879 5151 6536 10,443 1658 3747 3293
(48.9) (36.6) (46.5) (74.3) (11.8) (26.7) (23.4)
40,220 31,441 39,150 53,536 9144 23,431 18,132
(60.6) (47.4) (59.0) (80.7) (13.8) (35.3) (27.3)
b.001 b.001 b.001 b.001 b.001 b.001 b.001
5.9 (5.2) 15666 (23.6)
b.001 b.001
4.7 (4.1) 2212 (15.7)
Abbreviations: ACE, Angiotensin-converting enzyme; ARB, Angiotensin receptor blocker.
American Heart Journal March 2014
375.e2 Eapen et al
Supplementery Table II. Baseline Characteristics of Patients Discharged Alive by Type of Heart Failure and Presence of Atrial Fibrillation Heart Failure With Reduced Ejection Fraction
Characteristic Age, mean (SD), y Male, No. (%) Race, No. (%) Black White Other/unknown Medical history, No. (%) Anemia Chronic obstructive pulmonary disease Chronic renal insufficiency Coronary artery disease Diabetes mellitus Hyperlipidemia Hypertension Myocardial infarction Peripheral vascular disease Stroke or transient ischemic attack CHADS2 score, mean (SD) Initial test results, No. (%) Hemoglobin b9 g/dL Hemoglobin 9–11 g/dL Serum creatinine N2.0 mg/dL Systolic blood pressure N150 mm Hg Discharge medications, No. (%) ACE inhibitor or ARB Aspirin β-Blocker Diuretic Clopidogrel Lipid-lowering agent Warfarin Index hospitalization length of stay Length of stay, mean (SD), d Length of stay N7 d, No. (%)
Without Atrial Fibrillation (n = 12,415)
With Atrial Fibrillation (n = 9275)
Heart Failure With Preserved Ejection Fraction
P Value b.001 b.001 b.001
Without Atrial Fibrillation (n = 13,876)
With Atrial Fibrillation (n = 11,881) 79.8 (7.2) 4664 (39.3)
2166 (15.6) 10,533 (75.9) 1177 (8.5)
781 (6.6) 10,281 (86.5) 819 (6.9)
b.001 b.001 b.001
76.7 (7.3) 7064 (56.9)
77.8 (7.0) 5951 (64.2)
2013 (16.2) 9437 (76.0) 965 (7.8)
716 (7.7) 7985 (86.1) 574 (6.2)
5962 3382 3330 8594 5311 5441 8837 5110 2396 1896 3.0
(48.0) (27.2) (26.8) (69.2) (42.8) (43.8) (71.2) (41.2) (19.3) (15.3) (1.1)
4397 (47.4) 2618 (28.2) 2702 (29.1) 6580 (70.9) 3347 (36.1) 3875 (41.8) 6271 (67.6) 3678 (39.7) 1767 (19.1) 1661 (17.9) 3.1 (1.1)
.37 .11 b.001 .006 b.001 .003 b.001 .03 .65 b.001 .21
7964 4378 3669 7757 6687 5904 11,390 3586 2565 2311 3.3
(57.4) (31.6) (26.4) (55.9) (48.2) (42.5) (82.1) (25.8) (18.5) (16.7) (1.1)
6188 3595 2643 6516 4172 4522 9062 2818 2116 2195 3.2
(52.1) (30.3) (22.2) (54.8) (35.1) (38.1) (76.3) (23.7) (17.8) (18.5) (1.1)
b.001 .03 b.001 .09 b.001 b.001 b.001 b.001 .16 b.001 .008
400 2320 2185 4397
(3.2) (18.7) (17.6) (35.4)
239 (2.6) 1656 (17.9) 1606 (17.3) 2298 (24.8)
.005 .12 .59 b.001
949 3798 2472 7486
(6.8) (27.4) (17.8) (53.9)
574 2857 1544 4942
(4.8) (24.0) (13.0) (41.6)
b.001 b.001 b.001 b.001
9200 7649 8713 10,983 2286 5638 1695
(74.1) (61.6) (70.2) (88.5) (18.4) (45.4) (13.7)
6625 (71.4) 4276 (46.1) 6331 (68.3) 8,377 (90.3) 1066 (11.5) 3704 (39.9) 5253 (56.6)
b.001 b.001 .002 b.001 b.001 b.001 b.001
8707 7579 8290 11,435 2405 5961 1022
(62.7) (54.6) (59.7) (82.4) (17.3) (43.0) (7.4)
7070 4743 7037 10,218 1212 4151 6506
(59.5) (39.9) (59.2) (86.0) (10.2) (34.9) (54.8)
b.001 b.001 .40 b.001 b.001 b.001 b.001
4.8 (3.6) 1937 (15.6)
5.6 (4.4) 2041 (22.0)
b.001 b.001
5.3 (4.0) 2217 (18.7)
b.001 b.001
Abbreviations: ACE, Angiotensin-converting-enzyme; ARB, Angiotensin receptor blocker.
77.8 (7.7) 4873 (35.1)
P Value
4.7 (3.4) 1973 (14.2)
Associations between atrial fibrillation and early outcomes of patients with heart failure and reduced or preserved ejection fraction Zubin J. Eapen, MD, MHS, a,b Melissa A. Greiner, MS, a Gregg C. Fonarow, MD, c Zhong Yuan, MD, PhD, d Roger M. Mills, MD, d Adrian F. Hernandez, MD, MHS, a,b and Lesley H. Curtis, PhD a,b Durham, NC; Los Angeles, CA; and Raritan, NJ
Background The relative impact of atrial fibrillation on early outcomes of patients with heart failure with reduced or preserved ejection fraction (EF) is unknown. Methods We conducted a retrospective cohort study of clinical registry data linked to Medicare claims for patients with heart failure with reduced or preserved EF stratified by presence of atrial fibrillation at admission. Outcomes of interest were allcause mortality and readmission at 30 days. We used Kaplan-Meier methods to estimate mortality and calculated cumulative incidence estimates of readmission. We used Cox proportional hazards models to examine associations between atrial fibrillation and 30-day outcomes. Results Among 66,357 patients admitted to 283 hospitals between January 2001 and March 2006, 46% had atrial fibrillation (44% of patients with reduced EF and 48% of patients with preserved EF). After adjustment for other patient characteristics, atrial fibrillation was associated with a modestly higher risk of 30-day mortality (HR, 1.08; 95% CI, 1.03-1.14) and readmission (HR, 1.06; 95% CI, 1.02-1.11). In subgroup analyses, atrial fibrillation was associated with a higher risk of 30-day mortality (HR, 1.16; 95% CI, 1.08-1.25) among patients with preserved EF but not among patients with reduced EF. The association of atrial fibrillation with readmission did not differ by heart failure type (P = .37 for the interaction). Conclusions Atrial fibrillation was associated with higher 30-day mortality among patients with heart failure with preserved EF but not reduced EF. The association of atrial fibrillation with 30-day readmission was modest and did not differ by heart failure type. (Am Heart J 2014;167:369-375.e2.) As a chronic disease punctuated by acute exacerbations, heart failure is a significant burden to patients, providers, and payers. It is the leading cause of hospital readmission among older patients in the United States. 1 In 2010, more than $39 billion was spent on the care of patients with heart failure in the United States alone, mostly on inpatient care. 2 To increase hospitals' accountability in reducing preventable readmissions, the US Centers for Medicare & Medicaid Services (CMS) publicly reports hospital-level rates of all-cause mortality within 30 days after heart failure admission and all-cause readmission within 30 days after discharge. Hospitals are now financially accountable, as From the aDuke Clinical Research Institute, Duke University School of Medicine, Durham, NC, bDepartment of Medicine, Duke University School of Medicine, Durham, NC, c Ahmanson-UCLA Cardiomyopathy Center, University of California, Los Angeles, Los Angeles, CA, and dJanssen Research & Development, LLC, Raritan, NJ. Jalal K. Ghali, MD, served as guest editor for this article. Submitted September 13, 2013; accepted December 4, 2013. Reprint requests: Zubin J. Eapen, MD, MHS, Duke Clinical Research Institute, PO Box 17969, Durham, NC 27715. E-mail: [email protected] 0002-8703/$ - see front matter © 2014, Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.ahj.2013.12.001
well; in October 2012, the federal Hospital Readmissions Reduction Program began penalizing hospitals with higher-than-predicted readmission rates by up to 1% of their total Medicare reimbursement. The heart failure population for which hospitals are accountable includes patients with reduced ejection fraction (EF) and patients with preserved EF. Heart failure with preserved EF is increasing rapidly in prevalence and accounts for a growing share of heart failure hospitalizations. 3 Although evidence-based therapies for heart failure with preserved EF are lacking, guidelines emphasize the management of comorbid conditions such as atrial fibrillation. 4,5 Atrial fibrillation has been an independent predictor of adverse outcomes in the overall heart failure population in some but not all studies. 6-9 Understanding the relative impact of atrial fibrillation on early postdischarge outcomes by heart failure type could identify high-risk patients who affect outcomes at the population level. Using the Acute Decompensated Heart Failure National Registry (ADHERE) linked with Medicare inpatient claims, we examined associations between atrial fibrillation and 30-day mortality and readmission among patients with reduced or preserved EF.
American Heart Journal March 2014
370 Eapen et al
Methods Data sources We obtained hospitalization data from the ADHERE-Core registry, which was established to study the characteristics, treatments, and inpatient outcomes of patients hospitalized with acute decompensated heart failure. 10 More than 300 community and academic centers in the United States participated in the registry and enrolled more than 185,000 patients between January 2001 and March 2006. To analyze long-term follow-up data, we obtained Medicare standard analytic claim files from CMS. These files contain encrypted identifiers that allow for longitudinal follow-up of fee-for-service Medicare beneficiaries. We used indirect identifiers to link registry records to Medicare files using methods that have been described previously. 11 Almost 80% of ADHERE records have been linked with Medicare claims, and older patients enrolled in ADHERE constitute a nationally representative sample of fee-for-service Medicare beneficiaries. 11,12 The institutional review board of the Duke University Health System approved the study.
Study population We included patients 65 years or older living in the United States who had a registry hospitalization linked to Medicare claims. If multiple registry hospitalizations were identified for a single patient, we used the earliest as the index hospitalization. We restricted the population to patients enrolled in fee-forservice Medicare at the time of the index hospitalization. To identify heart failure type, we required that EF was recorded in the registry. For all readmission analyses, we restricted the population to patients who were discharged alive to home.
Outcomes Consistent with metrics reported publicly by CMS, the primary outcomes of interest were all-cause mortality within 30 days after admission and all-cause readmission within 30 days after discharge. We identified all-cause mortality on the basis of death dates recorded in the Medicare denominator files. We defined readmission as any subsequent inpatient claim except those for transfers to or from another hospital and admissions for rehabilitation (diagnosis-related group 462 or International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis code V57.xx). We classified readmissions based on the primary diagnosis on the inpatient claim. Secondary outcomes of interest were readmission for ischemic stroke (ICD9-CM code 433.x1 or 434.x1) 13 and readmission for heart failure (ICD-9-CM code 428.x, 402.x1, 404.x1, or 404.x3).
Covariates Variables from the registry included demographic characteristics, medical history, initial evaluation results, initial vital signs, laboratory test results, hospitalization dates, and discharge medications. We used quantitative and qualitative EF values to define heart failure with preserved EF (ie, EF of 40% or greater or normal/mildly impaired) and heart failure with reduced EF (ie, EF less than 40% or moderately/severely impaired). Variables had low rates of missingness (ie, less than 5% of records). Consistent with the approach used in the development of 30-day risk models for patients with heart failure, we imputed continuous variables to the overall median value and dichotomous variables to “no.” 14-16
We defined atrial fibrillation on the basis of a medical history of atrial fibrillation documented in the registry or any diagnosis of atrial fibrillation (ICD-9-CM code 427.31) or atrial flutter (ICD-9CM code 427.32) on the index hospitalization claim.
Subgroups We combined atrial fibrillation diagnosis and heart failure type to define 4 subgroups: preserved EF with or without atrial fibrillation, and reduced EF with or without atrial fibrillation.
Statistical analysis We present baseline characteristics of the study population by subgroup. We present categorical variables as frequencies and continuous variables as means with SDs. We tested for differences in baseline variables between groups using χ 2 tests for categorical variables and Kruskal-Wallis tests for continuous variables. We describe unadjusted 30-day event rates by subgroup. We used Kaplan-Meier methods to estimate mortality and log-rank tests to assess differences in mortality between groups. For allcause readmission, we calculated cumulative incidence estimates to account for the competing risk of mortality. We tested for differences in all-cause readmission between groups using Gray tests. For each subgroup, we identified the 5 most common causes of readmission within 30 days by grouping primary diagnosis codes into diagnostic categories including heart failure (as previously defined), acute myocardial infarction (ICD-9-CM code 410.x1), and additional categories based on clinical judgment. To examine adjusted associations between heart failure type, atrial fibrillation, and other patient characteristics with 30-day mortality and readmission, we used Cox proportional hazards models with robust standard errors to account for clustering of similar patients within hospitals. In multivariable models, we included variables for atrial fibrillation, reduced EF, age, sex, race, anemia, chronic obstructive pulmonary disease, chronic renal insufficiency, coronary artery disease, diabetes mellitus, hyperlipidemia, hypertension, peripheral vascular disease, myocardial infarction, stroke or transient ischemic attack, dyspnea, fatigue, rales, pulmonary edema, heart rate, systolic blood pressure, serum creatinine, serum sodium, hemoglobin, and year of index hospitalization. We selected covariates on the basis of clinical judgment, published literature, and availability in the registry. Consistent with prior studies, we included a variable for index length of stay greater than 7 days in the readmission models only. 17 We tested for an interaction between reduced EF and atrial fibrillation to examine whether the effect of atrial fibrillation differed by heart failure type and repeated the analysis within each heart failure type. We censored data for patients if they enrolled in Medicare managed care and at the time of death. In all analyses, we tested the proportionality assumption for the atrial fibrillation variable of interest. We considered a 2-sided P b .05 to be statistically significant for all tests. We used SAS software version 9.3 (SAS Institute Inc, Cary, North Carolina) for all analyses. This study was supported by a research agreement between Duke University and Johnson & Johnson.
American Heart Journal Volume 167, Number 3
Eapen et al 371
Table I. Baseline characteristics of the study population by type of heart failure and presence of atrial fibrillation Heart failure with reduced ejection fraction
Characteristic Age, mean (SD), y Male, No. (%) Race, No. (%) White Black Other/unknown Medical history, No. (%) Anemia Chronic obstructive pulmonary disease Chronic renal insufficiency Coronary artery disease Diabetes mellitus Hyperlipidemia Hypertension Myocardial infarction Peripheral vascular disease Stroke or transient ischemic attack CHADS2 score, mean (SD) Initial test results, No. (%) Hemoglobin b9 g/dL Hemoglobin 9–11 g/dL Serum creatinine N2.0 mg/dL Systolic blood pressure N150 mm Hg Discharge medications, No. (%) ACE inhibitor or ARB Aspirin β-Blocker Diuretic Clopidogrel Lipid-lowering agent Warfarin Index hospitalization length of stay Length of stay, mean (SD), d Length of stay N7 d, No. (%)
Without atrial fibrillation (n = 16,628)
With atrial fibrillation (n = 13,152)
77.6 (7.6) 9098 (54.7)
78.8 (7.3) 8006 (60.9)
12,825 (77.1) 2532 (15.2) 1271 (7.6)
11,360 (86.4) 983 (7.5) 809 (6.2)
Heart failure with preserved ejection fraction
P b.001 b.001 b.001
Without atrial fibrillation (n = 19,172)
With atrial fibrillation (n = 17,405)
78.9 (8.0) 6313 (32.9)
80.8 (7.4) 6415 (36.9)
14,910 (77.8) 2745 (14.3) 1517 (7.9)
15,161 (87.1) 1108 (6.4) 1136 (6.5)
P b.001 b.001 b.001
8372 4597 4778 11,399 7150 6842 11,742 6778 3350 2835 3.1
(50.3) (27.6) (28.7) (68.6) (43.0) (41.1) (70.6) (40.8) (20.1) (17.0) (1.2)
6551 3760 4154 9158 4755 5139 8872 5110 2617 2566 3.1
(49.8) (28.6) (31.6) (69.6) (36.2) (39.1) (67.5) (38.9) (19.9) (19.5) (1.2)
.36 .07 b.001 .046 b.001 b.001 b.001 b.001 .59 b .001 .15
11,183 6003 5254 10,492 9030 7556 15,497 4871 3626 3494 3.3
(58.3) (31.3) (27.4) (54.7) (47.1) (39.4) (80.8) (25.4) (18.9) (18.2) (1.1)
9505 5439 4210 9291 6170 5988 13,198 4030 3196 3484 3.3
(54.6) (31.2) (24.2) (53.4) (35.4) (34.4) (75.8) (23.2) (18.4) (20.0) (1.1)
b .001 .90 b.001 .01 b.001 b.001 b.001 b.001 .18 b .001 .03
616 3410 3232 5554
(3.7) (20.5) (19.4) (33.4)
394 2636 2620 3081
(3.0) (20.0) (19.9) (23.4)
b.001 .32 .30 b.001
1379 5473 3534 9906
(7.2) (28.5) (18.4) (51.7)
945 4543 2484 6769
(5.4) (26.1) (14.3) (38.9)
b.001 b.001 b.001 b.001
11,223 9388 10,740 13,677 2869 6578 2050
(67.5) (56.5) (64.6) (82.3) (17.3) (39.6) (12.3)
8416 5648 8127 10,857 1424 4507 6459
(64.0) (42.9) (61.8) (82.6) (10.8) (34.3) (49.1)
b.001 b.001 b.001 .50 b.001 b.001 b.001
11,184 9810 10,785 14,982 3129 7214 1369
(58.3) (51.2) (56.3) (78.1) (16.3) (37.6) (7.1)
9397 6595 9498 14,020 1722 5132 8254
(54.0) (37.9) (54.6) (80.6) (9.9) (29.5) (47.4)
b.001 b.001 .001 b.001 b.001 b.001 b.001
6.6 (6.0) 3791 (28.8)
b.001 b.001
6.3 (5.5) 4570 (26.3)
b.001 b.001
5.5 (4.7) 3526 (21.2)
5.4 (4.6) 3779 (19.7)
ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker.
Results We identified 80,416 patients with a linked registry record who were 65 years or older living in the United States with fee-for-service Medicare coverage at the time of hospital admission. We excluded 14,059 patients for whom EF was not recorded in the registry. Patients who did not have EF recorded had similar baseline characteristics and prevalence of comorbid atrial fibrillation compared with patients who did have EF recorded (the online Appendix Supplementary Table I). The final study population included 66,357 patients admitted to 283 hospitals between January 2001 and March 2006. Of these, 44.9% had reduced EF and 55.1% had preserved EF. Atrial fibrillation was present in 46.0% of all patients, including 44.2% of patients with reduced EF and 47.6% of patients with preserved EF.
Table I shows the baseline characteristics of the study population. The subgroup of patients with atrial fibrillation was slightly older and included more men and more white patients. Comorbid conditions were common. A history of stroke or transient ischemic attack was more frequent among patients with atrial fibrillation. Index hospitalization length of stay was greater among patients with atrial fibrillation. There were 47,447 patients discharged alive to home who were included for all readmission analyses. Comparisons of baseline characteristics by subgroup were similar to the full study population (the online Appendix Supplementary Table II). Table II and the Figure show that patients with atrial fibrillation had significantly worse unadjusted rates of mortality and readmission. Rates of ischemic stroke at 30 days were low among patients with and without atrial
American Heart Journal March 2014
372 Eapen et al
Table II. Cumulative incidence of mortality within 30 days after admission and cumulative incidence of readmission within 30 days after discharge⁎ Heart failure with reduced ejection fraction
Outcome Mortality All-cause readmission Heart failure readmission Ischemic stroke readmission
Without atrial fibrillation
With atrial fibrillation
1579/16,628 (9.5) 2766/12,415 (22.3) 1271/12,415 (10.2) 30/12,415 (0.2)
1502/13,152 (11.4) 2228/9275 (24.0) 1033/9275 (11.1) 22/9275 (0.2)
P b.001 .003 .04 .95
Heart failure with preserved ejection fraction Without atrial fibrillation
With atrial fibrillation
1218/19,172 (6.4) 2800/13,876 (20.2) 867/13,876 (6.3) 26/13,876 (0.2)
1606/14,405 (9.2) 2547/11,881 (21.4) 890/11,881 (7.5) 35/11,881 (0.3)
P b.001 .01 b.001 .08
⁎ Mortality values are expressed as number of events (Kaplan-Meier percentage). Readmission values are expressed as number of events (cumulative incidence percentage). Readmission was calculated among patients discharged alive only.
Figure
A
Cumulative Incidence, %
20%
HF−PEF+No AF HF−PEF+AF HF−REF+No AF HF−REF+AF
10%
0% 0
5
10
15
20
25
30
Days from Index Hospitalization Admission
B
Cumulative Incidence, %
30%
HF−PEF+No AF HF−PEF+AF HF−REF+No AF HF−REF+AF
20%
fibrillation. Among all patients readmitted within 30 days, heart failure was the most frequent reason for the first readmission, regardless of the type of heart failure or the presence of atrial fibrillation (Table III). Among patients with preserved EF and atrial fibrillation, atrial fibrillation was the third most frequent reason for readmission. Among patients with reduced EF and atrial fibrillation, atrial fibrillation was not among the 5 most frequent reasons for 30-day readmission. After adjustment for other patient factors, atrial fibrillation was associated with a modestly higher hazard of 30-day mortality (HR, 1.08; 95% CI, 1.03-1.14) and 30-day readmission (HR, 1.06; 95% CI, 1.02-1.11) in the overall study population (Table IV). The interaction between heart failure with reduced EF and atrial fibrillation in the mortality model indicated a differential effect of atrial fibrillation by type of heart failure (P b .001 for the interaction). After stratifying by heart failure type, the association of atrial fibrillation with higher risk of 30-day mortality was significant among patients with preserved EF (HR, 1.16; 95% CI, 1.08-1.25) but not patients with reduced EF (HR, 1.00; 95% CI, 0.941.08). The association of atrial fibrillation with 30-day readmission did not differ by heart failure type (P = .37 for the interaction). The proportionality assumption was met for the atrial fibrillation variable in all models.
Discussion 10%
0% 0
5
10
15
20
25
30
Days from Index Hospitalization Discharge Cumulative incidence of (A) mortality within 30 days of admission and (B) readmission within 30 days of discharge.
Through the Hospital Readmissions Reduction Program, CMS expects to recover $280 million in 2013 from financial penalties imposed on 2217 hospitals. 18 As the leading cause of hospitalization among older patients, heart failure presents a particular challenge. The impact of comorbid conditions like atrial fibrillation on postdischarge outcomes may differ according to heart failure type. In this study of a large nationwide registry of heart failure hospitalizations linked with Medicare data, we found that atrial fibrillation was modestly associated with 30-day mortality and readmission. After stratification by heart failure type, atrial fibrillation was associated with
American Heart Journal Volume 167, Number 3
Eapen et al 373
Table III. Five most frequent reasons for the first readmission among patients readmitted within 30 days after discharge⁎ Heart failure with reduced ejection fraction
Heart failure with preserved ejection fraction
Primary diagnosis frequency
Without atrial fibrillation (n = 2766)
With atrial fibrillation (n = 2228)
Without atrial fibrillation (n = 2800)
With atrial fibrillation (n = 2547)
Acute myocardial infarction Acute renal failure Atrial fibrillation Coronary atherosclerosis Heart failure Pneumonia Volume depletion
106 (3.8) 90 (3.3) — 135 (4.9) 1207 (43.6) 101 (3.7) —
53 (2.4) 96 (4.3) — 58 (2.6) 981 (44.0) 86 (3.9) —
87 (3.1) 112 (4.0) — 162 (5.8) 802 (28.6) 121 (4.3) —
— — 93 (3.7) 67 (2.6) 830 (32.6) 141 (5.5) 63 (2.5)
⁎ Values are expressed as number (percentage).
Table IV. Associations between atrial fibrillation and 30-day mortality and readmission All-cause mortality⁎ Heart failure type Any heart failure Heart failure with reduced ejection fraction Heart failure with preserved ejection fraction
Adjusted HR (95% CI) 1.08 (1.03-1.14) 1.00 (0.94-1.08) 1.16 (1.08-1.25)
All-cause readmission† P .003 .92 b.001
Adjusted HR (95% CI) 1.06 (1.02-1.11) 1.05 (0.99-1.11) 1.07 (1.01-1.14)
P .004 .12 .02
⁎ Adjusted for atrial fibrillation, reduced ejection fraction, age, sex, race, anemia, chronic obstructive pulmonary disease, chronic renal insufficiency, coronary artery disease, diabetes mellitus, hyperlipidemia, hypertension, peripheral vascular disease, myocardial infarction, stroke or transient ischemic attack, dyspnea, fatigue, rales, pulmonary edema, heart rate, systolic blood pressure, serum creatinine, serum sodium, hemoglobin, and year of index hospitalization. † Adjusted for the above variables and a variable for index length of stay greater than 7 days.
30-day mortality among patients with preserved EF but not patients with reduced EF. In addition, we found that heart failure was the most common cause of readmission among patients with either type of heart failure, regardless of the presence of comorbid atrial fibrillation. These findings suggest that while management of heart failure should be the primary focus of efforts to improve 30-day outcomes, targeting improved management of atrial fibrillation may be of benefit, particularly in patients with preserved EF. Atrial fibrillation may serve as a marker of illness severity in patients with heart failure, a result of increased left atrial pressure. The lack of atrial kick and rapid ventricular response associated with atrial fibrillation may also promote development of decompensated heart failure. Atrial fibrillation is a common comorbid condition among patients with heart failure; up to 40% of patients with advanced heart failure have atrial fibrillation. 7,19-21 Prevalence of both conditions is higher in older patients like Medicare beneficiaries. 22,23 Studies of the impact of atrial fibrillation on patients with either type of heart failure have yielded conflicting results. Some studies have found atrial fibrillation to be an independent predictor of mortality in patients with reduced EF, whereas others have not. 24-28 Results are similarly mixed in analyses of patients with preserved EF. 29-33 Our study is among the first to evaluate the impact of atrial fibrillation on 30-day
mortality and readmission among patients with either type of heart failure. Our study builds upon previous findings that atrial fibrillation is associated with inhospital mortality among patients with either reduced or preserved EF. 6 Our findings reveal a modest association between atrial fibrillation and 30-day mortality and readmission overall and, after stratification by heart failure type, among patients with preserved EF. Although clinicians should optimize the management of atrial fibrillation, our findings suggest that this is unlikely to considerably influence publicly reported 30-day outcomes in patients with reduced EF. Although administratively coded the same as reduced EF, preserved EF represents a different type of heart failure, which can require different management strategies and present different cost burdens. Identifying ways to improve postdischarge outcomes for patients with preserved EF is of particular importance. In-hospital mortality has improved for patients with preserved EF, but temporal trends in hospitalization for heart failure suggest that preserved EF may surpass reduced EF as the most prevalent form of heart failure. 3 Consequently, preserved EF may increasingly influence aggregate readmission rates. Management of comorbid conditions will remain important for patients with preserved EF, but therapies that directly address preserved EF to improve postdischarge outcomes are needed.
American Heart Journal March 2014
374 Eapen et al
Our study has some limitations. Administrative claims are subject to incomplete or incorrect coding that may affect outcome measurement. Rather than complete case analysis, we imputed missing values so that the final study population was not overly restrictive and because variables had low rates of missingness (ie, less than 5% of records). Hospitals that participate in ADHERE may be more likely to have better adherence to performance measures compared with other hospitals. Data were restricted to patients 65 years and older in the ADHERE registry, which limited our ability to generalize the findings to the broader population of patients with atrial fibrillation and heart failure. However, the study population is relevant to hospitals accountable for the care of Medicare beneficiaries through the Hospital Readmissions Reduction Program. Several clinical variables that are likely to be associated with clinical outcomes were not available, including New York Heart Association functional classification, symptom severity, frailty, and sinus rhythm during or after hospitalization. We could not account for socioeconomic status, education level, and health literacy. Residual or unmeasured confounding may have contributed to these findings. In conclusion, atrial fibrillation had a modest association with all-cause mortality within 30 days after admission and with all-cause readmission within 30 days after discharge among patients with heart failure. After stratification by heart failure type, atrial fibrillation was associated with 30-day mortality in patients with preserved EF but not in patients with reduced EF. The association of atrial fibrillation with 30-day readmission did not differ by heart failure type. Whether improved treatment of patients with atrial fibrillation reduces 30day readmission and mortality rates, particularly in patients with preserved EF, requires further study.
Acknowledgements Additional Contributions: Laura G. Qualls, MS, and Damon M. Seils, MA, Duke University, provided editorial assistance and prepared the manuscript.
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
References 1. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009; 360:1418-28. 2. Lloyd-Jones D, Adams RJ, Brown TM, et al. Heart disease and stroke statistics–2010 update: a report from the American Heart Association. Circulation 2010;121:e46-215. 3. Steinberg BA, Zhao X, Heidenreich PA, et al. Trends in patients hospitalized with heart failure and preserved left ventricular ejection fraction: prevalence, therapies, and outcomes. Circulation 2012;126:65-75. 4. Hunt SA, Abraham WT, Chin MH, et al. 2009 focused update incorporated into the ACC/AHA 2005 Guidelines for the Diagnosis and Management of Heart Failure in Adults: a report of the American
16.
17.
18. 19. 20.
College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. Circulation 2009;119:e391-479. Shah SJ, Gheorghiade M. Heart failure with preserved ejection fraction: treat now by treating comorbidities. JAMA 2008;300:431-3. Mountantonakis SE, Grau-Sepulveda MV, Bhatt DL, et al. Presence of atrial fibrillation is independently associated with adverse outcomes in patients hospitalized with heart failure: an analysis of get with the guidelines-heart failure. Circ Heart Fail 2012;5:191-201. Wang TJ, Larson MG, Levy D, et al. Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality: the Framingham Heart Study. Circulation 2003;107: 2920-5. Dries DL, Exner DV, Gersh BJ, et al. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: a retrospective analysis of the SOLVD trials. Studies of Left Ventricular Dysfunction. J Am Coll Cardiol 1998;32: 695-703. Kober L, Swedberg K, McMurray JJ, et al. Previously known and newly diagnosed atrial fibrillation: a major risk indicator after a myocardial infarction complicated by heart failure or left ventricular dysfunction. Eur J Heart Fail 2006;8:591-8. Adams Jr KF, Fonarow GC, Emerman CL, et al. Characteristics and outcomes of patients hospitalized for heart failure in the United States: rationale, design, and preliminary observations from the first 100,000 cases in the Acute Decompensated Heart Failure National Registry (ADHERE). Am Heart J 2005;149:209-16. Hammill BG, Hernandez AF, Peterson ED, et al. Linking inpatient clinical registry data to Medicare claims data using indirect identifiers. Am Heart J 2009;157:995-1000. Kociol RD, Hammill BG, Fonarow GC, et al. Generalizability and longitudinal outcomes of a national heart failure clinical registry: comparison of Acute Decompensated Heart Failure National Registry (ADHERE) and non-ADHERE Medicare beneficiaries. Am Heart J 2010;160:885-92. Wahl PM, Rodgers K, Schneeweiss S, et al. Validation of claimsbased diagnostic and procedure codes for cardiovascular and gastrointestinal serious adverse events in a commercially-insured population. Pharmacoepidemiol Drug Saf 2010;19:596-603. Keenan PS, Normand SL, Lin Z, et al. An administrative claims measure suitable for profiling hospital performance on the basis of 30-day all-cause readmission rates among patients with heart failure. Circ Cardiovasc Qual Outcomes 2008;1:29-37. Krumholz HM, Wang Y, Mattera JA, et al. An administrative claims model suitable for profiling hospital performance based on 30-day mortality rates among patients with heart failure. Circulation 2006; 113:1693-701. Harrell FE. Multivariable modeling strategies. Regression modeling strategies: with applicaions to linear models, logistic regression, and survival analysis. New York: Springer-Verlag; 2001. p. 53-85. Krumholz HM, Parent EM, Tu N, et al. Readmission after hospitalization for congestive heart failure among Medicare beneficiaries. Arch Intern Med 1997;157:99-104. Joynt KE, Jha AK. A path forward on Medicare readmissions. N Engl J Med 2013;368:1175-7. Stevenson WG, Stevenson LW. Atrial fibrillation and heart failure— five more years. N Engl J Med 2004;351:2437-40. Koelling TM, Chen RS, Lubwama RN, et al. The expanding national burden of heart failure in the United States: the influence of heart failure in women. Am Heart J 2004;147:74-8.
American Heart Journal Volume 167, Number 3
21. Maisel WH, Stevenson LW. Atrial fibrillation in heart failure: epidemiology, pathophysiology, and rationale for therapy. Am J Cardiol 2003;91:2D-8D. 22. Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study. JAMA 2001;285:2370-5. 23. Naccarelli GV, Varker H, Lin J, et al. Increasing prevalence of atrial fibrillation and flutter in the United States. Am J Cardiol 2009;104:1534-9. 24. Dries D, Exner D, Gersh B, et al. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: a retrospective analysis of the SOLVD trials. J Am Coll Cardiol 1998;32:695-703. 25. Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance of atrial fibrillation in advanced heart failure. A study of 390 patients. Circulation 1991;84:40-8. 26. Swedberg K, Olsson LG, Charlesworth A, et al. Prognostic relevance of atrial fibrillation in patients with chronic heart failure on longterm treatment with beta-blockers: results from COMET. Eur Heart J 2005;26:1303-8. 27. Carson PE, Johnson GR, Dunkman WB, et al. The influence of atrial fibrillation on prognosis in mild to moderate heart failure. The V-HeFT Studies. The V-HeFT VA Cooperative Studies Group. Circulation 1993;87:VI102-10.
Eapen et al 375
28. Crijns HJ, Tjeerdsma G, de Kam PJ, et al. Prognostic value of the presence and development of atrial fibrillation in patients with advanced chronic heart failure. Eur Heart J 2000;21: 1238-45. 29. Owan TE, Hodge DO, Herges RM, et al. Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med 2006;355:251-9. 30. Lenzen MJ, Scholte op Reimer WJ, Boersma E, et al. Differences between patients with a preserved and a depressed left ventricular function: a report from the EuroHeart Failure Survey. Eur Heart J 2004;25:1214-20. 31. Grigorian Shamagian L, Roman AV, Seara JG, et al. Atrial fibrillation in patients hospitalized for congestive heart failure: the same prognostic influence independently of left ventricular systolic function? Int J Cardiol 2006;110:366-72. 32. Olsson LG, Swedberg K, Ducharme A, et al. Atrial fibrillation and risk of clinical events in chronic heart failure with and without left ventricular systolic dysfunction: results from the Candesartan in Heart failure-Assessment of Reduction in Mortality and morbidity (CHARM) program. J Am Coll Cardiol 2006;47: 1997-2004. 33. Rusinaru D, Leborgne L, Peltier M, et al. Effect of atrial fibrillation on long-term survival in patients hospitalised for heart failure with preserved ejection fraction. Eur J Heart Fail 2008;10: 566-72.
American Heart Journal Volume 167, Number 3
Eapen et al 375.e1
Appendix Supplementary Table I. Characteristics of the Study Population by Ejection Fraction Status Characteristic Age, mean (SD), y Male, No. (%) Race, No. (%) Black White Other/unknown Medical history, No. (%) Atrial fibrillation Anemia Chronic obstructive pulmonary disease Chronic renal insufficiency Coronary artery disease Diabetes mellitus Hyperlipidemia/dyslipidemia Hypertension Peripheral vascular disease Myocardial infarction Stroke or transient ischemic attack CHADS2 score, mean (SD) Initial test results, No. (%) Hemoglobin b9 g/dL Hemoglobin 9–11 g/dL Serum creatinine N2.0 mg/dL Systolic blood pressure N150 mm Hg Discharge medications, No. (%) ACE inhibitor or ARB Aspirin β-Blocker Diuretic Clopidogrel Lipid-lowering Warfarin Index hospitalization length of stay Length of stay, mean (SD), d Length of stay N7 days, No. (%)
Missing Ejection Fraction (n = 14,059)
Not Missing Ejection Fraction (n = 66,357)
P Value b.001 b.001 b.001
80.4 (8.2) 5501 (39.1)
79.0 (7.7) 29,832 (45.0)
1796 (12.8) 11,187 (79.6) 1076 (7.7)
7368 (11.1) 54,256 (81.8) 4733 (7.1)
6064 7802 4294 3912 7957 5665 3989 10,132 2417 3575 2677 3.2
(43.1) (55.5) (30.5) (27.8) (56.6) (40.3) (28.4) (72.1) (17.2) (25.4) (19.0) (1.1)
30,557 35,611 19,799 18,396 40,340 27,105 25,525 49,309 12,789 20,789 12,379 3.2
(46.0) (53.7) (29.8) (27.7) (60.8) (40.8) (38.5) (74.3) (19.3) (31.3) (18.7) (1.1)
b.001 b.001 .10 .80 b.001 .23 b.001 b.001 b.001 b.001 .29 .18
928 3609 2848 5481
(6.6) (25.7) (20.3) (39.0)
3334 16,062 11,870 25,310
(5.0) (24.2) (17.9) (38.1)
b.001 b.001 b.001 .06
6879 5151 6536 10,443 1658 3747 3293
(48.9) (36.6) (46.5) (74.3) (11.8) (26.7) (23.4)
40,220 31,441 39,150 53,536 9144 23,431 18,132
(60.6) (47.4) (59.0) (80.7) (13.8) (35.3) (27.3)
b.001 b.001 b.001 b.001 b.001 b.001 b.001
5.9 (5.2) 15666 (23.6)
b.001 b.001
4.7 (4.1) 2212 (15.7)
Abbreviations: ACE, Angiotensin-converting enzyme; ARB, Angiotensin receptor blocker.
American Heart Journal March 2014
375.e2 Eapen et al
Supplementery Table II. Baseline Characteristics of Patients Discharged Alive by Type of Heart Failure and Presence of Atrial Fibrillation Heart Failure With Reduced Ejection Fraction
Characteristic Age, mean (SD), y Male, No. (%) Race, No. (%) Black White Other/unknown Medical history, No. (%) Anemia Chronic obstructive pulmonary disease Chronic renal insufficiency Coronary artery disease Diabetes mellitus Hyperlipidemia Hypertension Myocardial infarction Peripheral vascular disease Stroke or transient ischemic attack CHADS2 score, mean (SD) Initial test results, No. (%) Hemoglobin b9 g/dL Hemoglobin 9–11 g/dL Serum creatinine N2.0 mg/dL Systolic blood pressure N150 mm Hg Discharge medications, No. (%) ACE inhibitor or ARB Aspirin β-Blocker Diuretic Clopidogrel Lipid-lowering agent Warfarin Index hospitalization length of stay Length of stay, mean (SD), d Length of stay N7 d, No. (%)
Without Atrial Fibrillation (n = 12,415)
With Atrial Fibrillation (n = 9275)
Heart Failure With Preserved Ejection Fraction
P Value b.001 b.001 b.001
Without Atrial Fibrillation (n = 13,876)
With Atrial Fibrillation (n = 11,881) 79.8 (7.2) 4664 (39.3)
2166 (15.6) 10,533 (75.9) 1177 (8.5)
781 (6.6) 10,281 (86.5) 819 (6.9)
b.001 b.001 b.001
76.7 (7.3) 7064 (56.9)
77.8 (7.0) 5951 (64.2)
2013 (16.2) 9437 (76.0) 965 (7.8)
716 (7.7) 7985 (86.1) 574 (6.2)
5962 3382 3330 8594 5311 5441 8837 5110 2396 1896 3.0
(48.0) (27.2) (26.8) (69.2) (42.8) (43.8) (71.2) (41.2) (19.3) (15.3) (1.1)
4397 (47.4) 2618 (28.2) 2702 (29.1) 6580 (70.9) 3347 (36.1) 3875 (41.8) 6271 (67.6) 3678 (39.7) 1767 (19.1) 1661 (17.9) 3.1 (1.1)
.37 .11 b.001 .006 b.001 .003 b.001 .03 .65 b.001 .21
7964 4378 3669 7757 6687 5904 11,390 3586 2565 2311 3.3
(57.4) (31.6) (26.4) (55.9) (48.2) (42.5) (82.1) (25.8) (18.5) (16.7) (1.1)
6188 3595 2643 6516 4172 4522 9062 2818 2116 2195 3.2
(52.1) (30.3) (22.2) (54.8) (35.1) (38.1) (76.3) (23.7) (17.8) (18.5) (1.1)
b.001 .03 b.001 .09 b.001 b.001 b.001 b.001 .16 b.001 .008
400 2320 2185 4397
(3.2) (18.7) (17.6) (35.4)
239 (2.6) 1656 (17.9) 1606 (17.3) 2298 (24.8)
.005 .12 .59 b.001
949 3798 2472 7486
(6.8) (27.4) (17.8) (53.9)
574 2857 1544 4942
(4.8) (24.0) (13.0) (41.6)
b.001 b.001 b.001 b.001
9200 7649 8713 10,983 2286 5638 1695
(74.1) (61.6) (70.2) (88.5) (18.4) (45.4) (13.7)
6625 (71.4) 4276 (46.1) 6331 (68.3) 8,377 (90.3) 1066 (11.5) 3704 (39.9) 5253 (56.6)
b.001 b.001 .002 b.001 b.001 b.001 b.001
8707 7579 8290 11,435 2405 5961 1022
(62.7) (54.6) (59.7) (82.4) (17.3) (43.0) (7.4)
7070 4743 7037 10,218 1212 4151 6506
(59.5) (39.9) (59.2) (86.0) (10.2) (34.9) (54.8)
b.001 b.001 .40 b.001 b.001 b.001 b.001
4.8 (3.6) 1937 (15.6)
5.6 (4.4) 2041 (22.0)
b.001 b.001
5.3 (4.0) 2217 (18.7)
b.001 b.001
Abbreviations: ACE, Angiotensin-converting-enzyme; ARB, Angiotensin receptor blocker.
77.8 (7.7) 4873 (35.1)
P Value
4.7 (3.4) 1973 (14.2)
