Download Presence of Atrial Fibrillation Is Independently Associated With

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
Stavros E. Mountantonakis, MD; Maria V. Grau-Sepulveda, MD; Deepak L. Bhatt, MD, MPH;
Adrian F. Hernandez, MD, MHS; Eric D. Peterson, MD, MPH; Gregg C. Fonarow, MD
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
Background—It is unclear if the presence of atrial fibrillation (AF) on admission is associated with worse in-hospital
outcomes in patients hospitalized with heart failure (HF). This study evaluated the clinical characteristics, management,
length of stay, and mortality of HF patients with and without AF.
Methods and Results—We studied 99 810 patients from 255 sites admitted with HF enrolled in Get With The
Guidelines–Heart Failure between January 1, 2005, and December 31, 2010. Patients with AF on admission were
compared with patients in sinus rhythm. A total of 31 355 (31.4%) HF patients presented with AF, of which 6701
(21.3%) were newly diagnosed. Patients in AF were older (77⫾12 versus 70⫾15, P⬍0.001) and were more likely to
have history of stroke and valvular heart disease. AF patients had higher B-type natriuretic peptide levels and ejection
fraction (42⫾17% versus 39⫾17%, P⬍0.001). AF patients were more likely to be hospitalized ⬎4 days (48.8% versus
41.5%, P⬍0.001), discharged to a facility other than home (28.5% versus 19.7%, P⬍0.001), and had higher hospital
mortality rate (4.0% versus 2.6%, P⬍0.001). AF, particularly newly diagnosed, was independently associated with
adverse outcomes (adjusted odds ratios and 95% confidence intervals for mortality 1.17, 1.05–1.29, P⫽0.0029, and
1.29, 1.10 –1.52, P⫽0.0023 for AF and newly diagnosed AF, respectively).
Conclusions—In patients hospitalized with HF, AF is present in one-third and is independently associated with adverse
hospital outcomes and longer length of stay. Whether prompt restoration of sinus rhythm would improve outcomes in
patients hospitalized with HF and new-onset or paroxysmal AF is unclear and requires further study. (Circ Heart Fail.
2012;5:191-201.)
Key Words: atrial fibrillation 䡲 heart failure 䡲 hospitalization 䡲 mortality 䡲 outcomes
A
HF has been conflicting when analyzed within randomized
HF trial populations and other cohorts.6 –15 In the largest
series, AF was found to be an independent predictor of
long-term mortality.6,9 –11,15 In contrast, less is known about
the effects of AF, particularly newly diagnosed AF, on acute
outcomes in patients hospitalized with HF.16
trial fibrillation (AF) and heart failure (HF) are both
major causes of cardiovascular morbidity and mortality.1,2 The prevalence of both AF and HF rises with advancing
age. Thus, as the population of the United States ages, each
year our country must treat more and more patients with these
disorders, alone or in combination. The two also can precipitate one another. Worsening HF can bring on AF through
increased left atrial pressure, whereas AF itself can worsen
HF symptoms due to its loss of left atrial kick and irregular
and rapid heart rates. Overall, it is estimated that the annual
incidence of AF in the general HF population is 5%, whereas
as many as 40% of patients with advanced HF have AF.2– 6
The long-term prognostic significance of AF in patients with
Clinical Perspective on p 201
The purpose of this study is to examine the prevalence of
AF among patients hospitalized with HF and evaluate the
associations with short-term clinical outcomes. We also
seek to identify any differences in the application of
current HF guidelines in the management of patients with
Received September 22, 2011; accepted February 10, 2012.
From the North Shore University Hospital and Hofstra School of Medicine, Manhasset, NY (S.E.M.); Duke Clinical Research Institute, Durham, NC
(M.V.G.-S., A.F.H., E.D.P.); Veterans Affairs Boston Healthcare System, Brigham and Women’s Hospital, and Harvard Medical School, Boston, MA
(D.L.B.); and Ahmanson-University of California Los Angeles Cardiomyopathy Center, Los Angeles, CA (G.C.F.).
Guest Editor for this article was Gregory Y.H. Lip, MD, FRCP, FESC, FACC.
The online-only Data Supplement is available with this article at http://circheartfailure.ahajournals.org/lookup/suppl/doi:10.1161/
CIRCHEARTFAILURE.111.965681/-/DC1.
Correspondence to Stavros E. Mountantonakis, MD, North Shore University Hospital, Department of Cardiology, 300 Community Dr, Manhasset, NY
11030. E-mail [email protected]
© 2012 American Heart Association, Inc.
Circ Heart Fail is available at http://circheartfailure.ahajournals.org
191
DOI: 10.1161/CIRCHEARTFAILURE.111.965681
192
Circ Heart Fail
March 2012
Definitions
Several patient populations were classified according to the temporal
characteristics history of their AF and their respective rhythm
findings on admission ECG. Current AF was defined as the presence
of AF on admission irrespective of history of AF. Patients in
preexisting AF were in AF on admission and had a reported medical
history of AF. New onset was defined as presence of AF on
admission with negative medical history of AF. Patients in the
category “sinus rhythm” were in sinus rhythm on presentation
irrespective of history of AF.
Outcome Measures
The primary outcome measure of interest was in-hospital mortality.
Additional outcomes of interest included length of hospital stay and
discharge location (home versus other). Quality of care was assessed
using the GWTG-HF performance and quality measures as previously described.17,18
Statistical Analysis
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
Figure 1. Flow diagram showing selection of study population
and study subgroup based on atrial fibrillation status.
AF. We hypothesize that the detrimental hemodynamic
effects of AF, especially in patients with new-onset AF,
may lead to or worsen in-hospital outcomes for these
patients.
Methods
Data Collection
The Get With The Guidelines–Heart Failure (GWTG-HF) program is
among the largest quality improvement initiatives focusing on
patients hospitalized with HF, the design of which has been previously described.17,18 Hospitals participating in the registry use a
web-based patient management tool (PMT, Outcomes Sciences Inc,
Cambridge, MA) to collect data for consecutive patients admitted
with HF to receive recommendations for qualitative improvement in
medical management. Patients hospitalized with new or worsening
HF as primary diagnosis or patients that developed significant HF
symptoms such that HF was the primary discharge diagnosis were
included. Patients were enrolled into the program regardless of their
left ventricular function. Hospitals from all regions of the United
States are represented and a variety of institutions participate, from
community hospitals to large tertiary medical centers. Data collected
for each HF patient include demographics, medical/surgical history
including any history of AF, admission medications, physical examination, rhythm at time of admission, serum laboratory tests, pharmacological and nonpharmacologic interventions, in-hospital outcomes, and discharge information. Trained hospital personnel enter
the data by using standardized definitions. All participating hospitals
were required to submit the GWTG protocol to their institutional
review board for approval. Because data collected were used for
qualitative hospital improvement, sites were granted a waiver of
informed consent under the common rule. Outcome Sciences Inc
serves as the data collection and the Duke Clinical Research Institute
(Durham, NC) serves as the data analysis center.
Study Population
The study evaluated patients enrolled in the GWTG-HF registry
between January 2005 and January 2011. A total of 136 465 patients
from 282 participating hospitals were registered during that period.
From this group, patients were excluded due to missing discharge
status (n⫽4320) or missing initial heart rhythm and/or history of AF
(n⫽32 335). The final study population consisted of 99 810 patients
from 255 sites (Figure 1).
Categorical variables were summarized using percentages and compared using Cochran Mantel Haenzel general association statistic
test; continuous variables were presented using means and standard
deviation and compared using Wilcoxon rank sum test. Patients with
AF were compared with those in sinus rhythm. Subgroups of patients
with preexisting AF and new-onset AF were also compared. The
multivariable association between presence of AF and primary end
points was assessed using a logistic regression model with the
generalized estimating equations method to account for withinhospital clustering and obtaining unadjusted and adjusted odds ratios
and confidence intervals. For adjusted models, the following variables were included: age (per 10 years), race, sex, chronic obstructive pulmonary disease, stroke/transient ischemic attack, depression,
diabetes mellitus, renal failure, dyslipidemia, hypertension, peripheral artery disease, smoking, ischemic etiology, valvular heart
disease, heart rate, blood pressure, ejection fraction, sodium level,
hemoglobin, serum creatinine, serum blood urine nitrogen, hospital
region where patient is admitted, hospital academic status, and
hospital heart transplant capability. To evaluate the importance of
ejection fraction (EF ⱖ40% versus ⬍40%) on outcomes, we include
an interaction term in the overall model and compute the odds ratios
for AF based on the above 2 subgroups. We also provide the
probability value for the interaction or p for heterogeneity. All
analyses were performed using SAS software (version 9.1, SAS
Institute, Cary, NC). Probability values were 2-sided, with P⬍0.01
considered statistically significant to adjust for multiple comparisons
among our 3 groups. All authors had access to data and assume
responsibility of the accuracy of the analysis.
Results
Of the 99 810 patients from 255 hospital sites in the final
analysis, 31 355 (31.4%) patients presented in AF, of which
6701 (6.7%) were newly diagnosed (Figure 1). AF was more
prevalent in patients with history of HF compared with
patients with newly diagnosed HF (32.78% versus 28.82%,
P⬍0.001), whereas newly diagnosed AF was more prevalent
in patients with newly diagnosed HF (9.09% versus 5.27%,
P⬍0.001) (online-only Data Supplement Table I). Baseline
characteristics of the study population are shown in Table 1.
Patients in AF were older (77⫾12 versus 70⫾15 years,
P⬍0.001) and had similar gender distribution with patients in
sinus rhythm. Comorbidities that were more frequent in
patients with AF included history of stroke, valvular heart
disease, HF, and pulmonary disease. Diabetes mellitus was
highly prevalent in both groups, but moderately less frequent
in AF patients. On presentation, patients in AF had higher
heart rates and lower blood pressure. Laboratory tests on
Mountantonakis et al
Adverse Hospital Outcomes in HF Patients With AF
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
admission in patients with AF were significant for lower
serum creatinine and blood urea nitrogen, higher frequency of
abnormal troponin levels and higher serum B-type natriuretic
peptide (BNP) levels. Echocardiogram performed during
admission showed a higher EF in AF patients.
Subgroup analysis showed that patients with newly diagnosed AF had different clinical profiles from patients in sinus
rhythm or preexisting AF. Although older than patients in
sinus rhythm, patients with newly diagnosed AF were
younger than patients with preexisting AF. The prevalence of
comorbid diagnoses was significantly less than the other 2
groups except for diabetes and need for dialysis (Table 1).
Notably, patients with new-onset AF had less frequent HF
history compared with the 2 other study groups. At presentation, these patients had heart rates that were higher than
both patients in sinus rhythm and preexisting AF, whereas
systolic blood pressure was slightly higher than in patients
with preexisting AF. Last, EF was higher than patients with
sinus rhythm and clinically comparable with EF in patients
with preexisting AF (Table 1).
Differences in medical therapy before hospital admission
in patients in sinus rhythm and AF were also noted (onlineonly Data Supplement Table II). Patients in AF were less
likely to be receiving therapy with angiotensin-converting
enzyme inhibitors (ACEi), aspirin, hydralazine, statins, and
nitrates and were more likely to be on treatment with
antiarrhythmics, digoxin, calcium channel blockers, diuretics,
and warfarin. Medication treatment rates before hospitalization for patients with preexisting diagnosis of HF and reduced
EF are shown in online-only Data Supplement Table III. In
these patients, treatment rates with ACEi/ARB and
␤-blockers were above 60% and patients in SR were more
likely to be treated compared with AF patients. Significant
differences among the populations stratified by AF status
were also identified with respect to hospital management
(Table 2). Patients in AF underwent less invasive procedures
in general, including lower frequency of right cardiac catheterization, stress testing, and coronary angiography. Multivariable analysis showed that presence of AF was independently associated with less cardiac catheterization after
adjusting for age, medical history, laboratories, vitals, and
hospital characteristics (online-only Data Supplement Table
IV). As expected, cardioversion was significantly more prevalent in the AF group. Renal replacement therapy during
hospitalization was more common for the sinus rhythm
group.
In regard to conformity with HF performance and quality
measures during hospitalization and on discharge, small
differences were identified between the 2 patient groups.
Patients in AF were discharged with all-or-none composite
measure rate of 86.8%. Rates of conformity with each
performance and quality measure are shown in (Table 3).
Patients in AF were less commonly discharged on ACEi or
ARB, evidenced-based ␤-blocker and lipid-lowering agent
and received less frequently prophylaxis for deep venous
thrombosis during hospitalization.
Hospital outcomes were significantly different among
the patient groups by the presence or absence of AF
(Table 4). Hospital length of stay was significantly longer
193
in patients in AF (mean, 5 versus 4 days; P⬍0.001, in
patients with AF and sinus rhythm, respectively), and
patients in AF were more likely to be discharged to a
facility other than home (28.5% versus 19.7%, P⬍0.001).
Hospital mortality rate was significantly higher in the AF
group (4.0% versus 2.6%, P⬍0.001). Among patients in
AF, those with newly diagnosed AF had even lengthier
hospitalization and higher hospital mortality than both
patients in sinus rhythm and preexisting AF (Table 4 and
Figure 2). In multivariable analysis, the presence of AF
was independently associated with adverse in-hospital
outcomes including length of stay more than 4 days,
discharge to a facility other than home, and hospital
mortality (Table 4). Notably, newly diagnosed AF had a
stronger independent association with adverse outcome
(Table 4).
The relationships between AF and outcomes were also
examined separately in patients with reduced and preserved
EF HF patients. The associations between AF and outcomes
in the adjusted analyses were similar for both EF subgroups
(EF ⱖ40% and ⬍40%), suggesting that the relationship
between AF and outcomes applies irrespective of type of HF
(preserved versus reduced EF HF) (Table 5).
Discussion
In this large, multicenter, national HF registry of patients
admitted with HF, 31.4% of patients presented in AF, and, of
those in AF, 21.3% were newly diagnosed. Patients in AF
were older and more likely to have a history of stroke and
valvular heart disease. AF patients underwent fewer invasive
procedures including right and left heart catheterization. AF
patients were more likely to be hospitalized ⬎4 days, discharged to a facility other than home, and had higher hospital
mortality rates. AF, particularly newly diagnosed, was independently associated with adverse in-hospital outcomes
among patients hospitalized with HF, including mortality.
The present study demonstrates the very high prevalence of
AF among patients hospitalized with HF patients and expands
on current knowledge regarding the potential contributions of
AF to adverse clinical outcomes among patients hospitalized
with HF.
The prevalence of AF in this patient population is remarkably high. This incidence is higher than most prior case series
with ambulatory patients with HF that have reported an
incidence of AF between 13–27%.7,8,11,19 –21 This difference
could be attributed to the fact that our series captures patients
in a state of decompensated HF requiring admission. Therefore, adverse hemodynamics of increased left atrial pressure,
enhanced cathecholaminergic state, hypoxemia, and
tachycardia may have promoted the development of AF. In
agreement with our finding in a much smaller cohort of 4596
patients admitted to the Mayo Clinic with HF, 34.9% were in
AF.19 A main difference in baseline characteristics of the 2
groups is that patients in AF were older. Development of AF
has been previously associated with age in the Framingham
study, which showed a 1 in 4 lifetime risk of developing AF
in adults over the age of 40.22 Conditions that are classically
associated with development or presence of AF, namely,
194
Circ Heart Fail
Table 1.
March 2012
Baseline Characteristics Stratified by AF Status on Admission
Variable
Sinus Rhythm
(n⫽68 455)
Current AF
(n⫽31 355)
P Value
Preexisting AF
(n⫽24 654)
New-Onset
AF (n⫽6701)
P Value†
P Value‡
Age, y, mean⫾SD
70.0⫾15.0
77.2⫾11.8
⬍0.0001§
77.5⫾11.6
76.1⫾12.6
⬍0.0001§
⬍0.0001§
0.2362
0.0922
⬍0.0001§
Male, n, %
33 861
49.5%
15 456
0.8229
49.3%
12 123
49.2%
3333
49.7%
Presenting vital signs
84⫾19
87⫾23
⬍0.0001§
86⫾22
92⫾26
⬍0.0001§
Systolic blood pressure, mm Hg
143⫾32
135⫾28
⬍0.0001§
135⫾27
136⫾28
⬍0.0001§
0.0030§
BMI, kg/m2
30.3⫾8.9
29.4⫾8.4
⬍0.0001§
29.3⫾8.3
29.7⫾8.4
⬍0.0001§
⬍0.0003§
⬍0.0001§
0.0147§
⬍0.0001§
⬍0.0001§
⬍0.0001§
0.2693
0.0551
0.0195§
Heart rate, bpm
(54 661)†
(24 488)†
(19 119)†
(5369)†
Presenting symptoms
Chest pain
10 630
20.2%
Dyspnea at rest and/or exertion
44 867
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
85.2%
Fatigue
9601
18.2%
Orthopnea
12 133
23.0%
Palpitations
1222
2.3%
Paroxysmal nocturnal dyspnea
5697
10.8%
3605
⬍0.0001§
15.4%
20 083
15.1%
0.0133§
85.9%
4943
⬍0.0001
0.5481
4228
23.2%
⬍0.0001
5.7%
2352
3884
21.3%
22.8%
1329
15 842
86.8%
21.1%
5341
2756
976
5.3%
0.0017§
10.1%
1889
10.4%
849
16.5%
4241
82.4%
1,059
20.6%
1113
21.6%
353
⬍0.0001§
⬍0.0001§
0.0001§
0.0042§
0.3948
0.7557
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
0.0030§
⬍0.0001§
0.7063
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
6.8%
463
9.0%
Presenting signs
Admission JVP, cm
15⫾7, 22*
(10 015)
Rales
24 575 35.9%
15⫾7, 22*
0.1859
(5020)
11 799 37.6%
⬍0.0001§
16⫾7, 22*
(4047)
(973)
9365
2434
38.0%
Lower-extremity edema
29 922
43.7%
14 742
⬍0.0001§
47.0%
15⫾7, 21*
11 684
47.4%
36.3%
3058
45.6%
Medical history
Smoking
13 680
20.0%
Hypertension
52 207
77.1%
Hyperlipidemia
30 665
45.3%
3447
⬍0.0001§
11.0%
23 244
10.2%
⬍0.0001§
74.7%
13 710
2525
18 650
75.6%
0.0005§
44.1%
11 216
45.5%
922
13.8%
4594
71.1%
2494
38.6%
Diabetes
Insulin-treated
15 252
22.5%
Non–insulin-treated
15 733
Valvular heart disease
7606
23.2%
11.2%
CAD
33 942
Prior MI
14 680
50.1%
21.7%
Prior PCI
7215
10.6%
4895
⬍0.0001§
15.7%
6725
16.0%
⬍0.0001§
5340
⬍0.0001§
4738
21.6%
5501
21.7%
17.7%
15 244
19.2%
0.0013§
12 442
⬍0.0001§
4896
49.0%
5990
50.5%
19.2%
2691
8.6%
3956
19.9%
⬍0.0001§
2229
9.0%
939
14.5%
1385
21.4%
763
11.8%
2802
43.4%
1094
16.9%
462
7.2%
(Continued)
Mountantonakis et al
Table 1.
Adverse Hospital Outcomes in HF Patients With AF
195
Continued
Variable
Prior CABG
Sinus Rhythm
(n⫽68 455)
9648
14.2%
Prior heart failure
44 501
65.7%
Ischemic HF
39 825
58.2%
CVA
9130
PVD
8209
13.5%
12.1%
Renal insufficiency
14 856
21.9%
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
Dialysis, chronic
3433
5.1%
COPD or asthma
20 429
Current AF
(n⫽31 355)
4488
P Value
0.4401
14.4%
21 699
57.0%
5181
⬍0.0001§
18 211
0.0007§
14 375
73.9%
58.3%
4351
0.5781
3140
17.6%
12.0%
6245
12.7%
⬍0.0001§
5154
⬍0.0001§
600
20.1%
782
20.9%
2.5%
9822
30.2%
31.6%
1417⫾1694
1141⫾1437
867*
696*
2.4%
⬍0.0001§
New-Onset
AF (n⫽6701)
P Value†
P Value‡
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
3510
⬍0.0001§
⬍0.0001§
830
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
0.0792
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
0.0284§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
⬍0.0001§
0.0113§
810
12.5%
3488
54.0%
52.4%
⬍0.0001§
16.6%
3731
3678
14.9%
69.7%
17 885
Preexisting AF
(n⫽24 654)
8089
32.8%
12.8%
591
9.2%
1091
16.9%
182
2.8%
1733
26.8%
Laboratory characteristics
BNP, pg/mL
Serum creatinine, mg/dL
(45 876)
(20 884)
1.9⫾4.8
1.6⫾2.6
1.30*
(56 135)
BUN, mg/dL
Troponin, abnormal
31⫾19, 26*
(24 666)
Ejection fraction, %
4552
39⫾17
42⫾17
(28 044)
1190⫾1454
686*
738*
(16 325)
(4559)
1.6⫾2.7
1.6⫾2.1
(19 518)
⬍0.0001§
31⫾19, 26*
(19 208)
⬍0.0001§
14.5%
(61 688)
1127⫾1432
1.30*
(24 984)
30⫾20, 25*
18.6%
⬍0.0001§
1.30*
(55 829)
12 703
⬍0.0001§
3336
13.5%
⬍0.0001§
1.30*
(5466)
30⫾20, 25*
(5458)
1216
18.2%
42⫾17
42⫾17
(22 104)
(5940)
Hospital characteristics
Academic hospital
38 984
56.9%
18 332
⬍0.0001§
58.5%
14 957
3375
60.7%
50.4%
⬍0.0001§
Hospital region
West
13.1%
13.2%
12.7%
14.8%
South
31.3%
25.4%
23.3%
33.2%
Midwest
23.3%
23.3%
24.1%
20.2%
Northeast
32.1%
37.7%
39.4%
Heart transplant–capable
7419
10.8%
PCI- and cardiac surgery–capable
46 770
68.3%
3060
⬍0.0001§
9.7%
21 095
67.3%
2581
10.5%
⬍0.0001§
16 679
67.6%
31.5%
479
7.2%
4416
65.9%
Actual numbers are presented on first row of each categorical variable and percentages on the second. For continues variables, means and standard deviations
are presented with the number of patients for each subgroup in parenthesis.
AF indicates atrial fibrillation; BMI, body mass index; JVP, jugular venous pressure; CAD, coronary artery disease; MI, myocardial infarction; PCI, percutaneous
intervention; CABG, coronary artery bypass graft; HF, heart failure; CVA, cerebrovascular accident; CVA, cerebrovascular accident; COPD, chronic obstructive pulmonary
disease; BNP, B-type natriuretic peptide; BUN, blood urea nitrogen.
*Median values for variables that did not follow normal distribution.
†P comparison between sinus rhythm, preexisting AF, and new-onset AF.
‡P comparison between preexisting and new-onset AF.
§Statistically significant differences.
196
Circ Heart Fail
Table 2.
March 2012
Hospital Management Stratified by AF Status
Variable
Right cardiac catheterization
Sinus Rhythm
(n⫽68 455)
1716
2.8%
Mechanical ventilation
1611
Coronary angiography
5792
2.6%
9.4%
PCI
860
CABG
325
1.4%
0.5%
Pacemaker
362
CRT-P
123
0.6%
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
0.2%
CRT-D
2058
ICD
2550
3.3%
4.1%
Cardioversion
200
0.3%
Dialysis
2944
4.5%
Current AF
(n⫽31 355)
651
Preexisting
AF (n⫽24 654)
P Value
0.0002‡
2.3%
751
2.3%
0.4288‡
525
⬍0.0001‡
1198
2.7%
1795
2.4%
6.4%
279
5.5%
⬍0.0001‡
194
0.1973
64
1.0%
128
0.9%
0.5%
295
0.3%
⬍0.0001‡
222
0.1736
57
1.1%
68
1.0%
0.2%
801
0.3%
0.0004‡
685
⬍0.0001‡
629
2.9%
784
3.1%
2.8%
811
2.9%
⬍0.0001‡
556
⬍0.0001‡
552
2.9%
745
498
2.5%
2.7%
2.5%
New-Onset
AF (n⫽6701)
153
P Value*
0.0004‡
P Value†
0.2269
2.5%
226
⬍0.0001‡
⬍0.0001‡
⬍0.0001‡
⬍0.0001‡
⬍0.0001‡
0.0003‡
⬍0.0001‡
⬍0.0001‡
⬍0.0001‡
0.1835
3.8%
597
9.9%
85
1.4%
64
1.1%
73
1.2%
11
0.2021‡
0.2784‡
⬍0.0001‡
⬍0.0001‡
⬍0.0001‡
0.2123
⬍0.0001‡
⬍0.0001‡
⬍0.0001‡
0.0036‡
0.2%
116
1.9%
155
2.6%
255
4.2%
193
3.21%
Actual numbers are presented on first row of each categorical variable and percentages on the second.
AF indicates atrial fibrillation; PCI, percutaneous intervention; CABG, coronary artery bypass graft; CRT-P, cardiac resynchronization pacemaker; CRT-D, cardiac
resynchronization defibrillator; and ICD, implantable cardioverter-defibrillator.
*P comparison between sinus rhythm, preexisting AF, and new-onset AF.
†P comparison between preexisting and new-onset AF.
‡Statistically significant differences.
valvular disease and chronic obstructive pulmonary disease
and stroke were, as expected, more prevalent in the AF group.
Subgroup analysis showed differences in clinical characteristics in patients with newly diagnosed AF. It appears that
those patients have fewer comorbidities than patients in
preexisting AF or sinus rhythm, which is in agreement with
prior reports.15 Specifically, almost half of patients who were
diagnosed with AF for the first time were simultaneously
diagnosed with HF. This is in contrast to patients with
preexisting AF, where only a quarter of patients with preexisting AF were de novo diagnosed with HF. This observation
brings the interesting notion of new onset AF being the
initiating mechanism of HF decompensation. The deleterious
effects of both increased heart rate and irregularity in combination with loss of atrial kick could result in adverse
hemodynamic consequences, namely, decrease in ventricular
filling time, increase in oxygen consumption, and ultimately
reduction in cardiac output.23 The effects of onset of AF in
patients with chronic HF have been previously studied in a
cohort of 344 patients. In that prospective study, onset of AF
was associated with clinical worsening of HF decrease in
peak oxygen consumption and cardiac index and development of mitral and tricuspid regurgitation all of which
translated to overall poorer prognosis.24
In terms of hospital management, adherence to quality
measures of the participating hospitals was relatively high
with patients discharged, with all-or-none composite measure
rate ⬎85%. Nevertheless, small statistically significant differences in the hospital management among the study subgroups were identified. Patients in AF received less diagnostic and therapeutic interventions related to coronary artery
disease, namely stress testing, coronary angiography, and
percutaneous interventions during hospitalization. An interesting difference was seen in device implantations. Patients in
AF received almost twice as many pacemakers and half as
many defibrillators as patients in sinus rhythm. The higher
incidence of pacemaker implantation could be explained by
the higher age in AF patients as well as the higher use of
atrioventricular nodal blocking and antiarrhythmic agents that
may exacerbate sinus node dysfunction in patients with
paroxysmal AF.25
In-hospital treatment of patients with new-onset AF was
also significantly different than both patients in sinus rhythm
and preexisting AF. Despite their more benign clinical profile
before admission they underwent significantly more invasive
procedures than other patients. Another important observation is that even though patients with newly diagnosed AF
underwent significantly more cardioversions, they did so only
Mountantonakis et al
Table 3.
Adverse Hospital Outcomes in HF Patients With AF
197
GWTG-HF Performance and Quality Measures by AF Status
Variable
Sinus Rhythm
(n⫽68 455)
Discharged with appropriate
instructions
47 731
Documented EF
61 213
90.3%
97.5%
ACEi or ARB for patients with
reduced EF
22 677
Smoking cessation instructions
12 456
92.3%
97.4%
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
Use of ␤-blockers for patients
with reduced EF
26 768
Composite performance
measure for 100% compliance
54 105
Anticoagulation for patients
with AF
Aldosterone antagonist for
patients with reduced EF
Use of evidenced ␤-blockers
for patients with reduced EF
94.4%
85.%
NA
Current AF
(n⫽31 355)
P Value
19 309
0.3685
15 125
0.0732
21 584
90.5%
90.4%
27 379
97.7%
7781
97.8%
⬍0.0001‡
90.3%
0.0266
96.7%
29.3%
22 397
77.8%
9704
0.7478
7615
⬍0.0001
19 278
94.3%
94.3%
86.8%
86.8%
16 880
13 843
Hydralazine and nitrates for
African American patients with
reduced EF
1973
DVT prophylaxis
8486
25.5%
56.7%
Use of lipid-lowering
medication
29 756
ICD for eligible HF patients
with EF ⱕ35%
10 337
65.5%
48.3%
72.3%
2948
0.0027‡
27.7%
7610
2349
28.1%
⬍0.0001‡
72.5%
5967
72.4%
420
0.4831
26.3%
333
26.3%
4215
0.0020‡
54.5%
11 763
2233
96.2%
69.3%
8271
6036
89.9%
3054
24 484
Preexisting
AF (n⫽24 654)
3231
53.3%
⬍0.0001‡
59.3%
3819
47.2%
9531
58.9%
0.0707
3248
49.4%
New-Onset
AF (n⫽6701)
4184
P Value*
P Value†
0.5279
0.4903
0.0012‡
0.0010‡
90.8%
5795
97.1%
1745
⬍0.0001‡
0.0271
0.0024‡
0.0147
0.9041
0.7552
⬍0.0001‡
0.8036
91.6%
821
98.0%
2089
94.1%
5206
86.7%
⬍0.0001
3037
58.5%
599
0.0040‡
0.1450
26.5%
1643
⬍0.0001
0.7931
0.7816
0.9762
72.7%
87
26.4%
984
⬍0.0001
⬍0.0001‡
⬍0.0001‡
0.0245
⬍0.0001‡
⬍0.0001‡
59.2%
2232
60.9%
571
37.4%
Actual numbers are presented on first row of each categorical variable and percentages on the second.
GWTG-HF indicates Get With The Guideline–Heart Failure; AF, atrial fibrillation; EF, ejection fraction; ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin
receptor antagonist; DVT, deep venous thrombosis; ICD, implantable cardioverter defibrillator; HF, heart failure.
*P comparison between sinus rhythm, preexisting AF, and new-onset AF.
†P comparison between preexisting and new-onset AF.
‡Statistically significant differences.
in 4.2% of cases. Although it is possible that cases admitted
in AF spontaneously converted to sinus rhythm after management of HF, and such practice cannot be captured by our
registry, this scenario could not fully explain the very small
number of cardioversion attempts in AF patients. Another
plausible explanation of low cardioversion rates is the belief
that cardioversion success could be lower in setting of
decompensated failure, and many clinicians may have elected
to perform cardioversion at a later time.
The main finding of our analysis was the significant
difference in all primary outcome measures between patients
with sinus rhythm and AF. Patients in AF, despite having
fewer comorbidities on presentation, had lengthier hospital
course and were more likely to be discharged to a facility
other than home. More importantly, mortality in patients with
AF was significantly higher despite the fact that they had
lower prevalence of previously described predictors of mor-
tality including lower EF, higher BNP, abnormal troponins,
and presence of renal dysfunction. In the outpatient setting,
the importance of AF in HF population has been demonstrated in several studies. The importance of AF in patients
with chronic systolic HF has been shown in the Studies of
Left Ventricular Dysfunction (SOLVD) and Carvedilol or
Metoprolol European Trial (COMET) trials. SOLVD found
that AF was an independent predictor for all-cause mortality
in 6500 patients with EF ⱕ35%,10 whereas in a retrospective
analysis of the COMET, AF was associated with increased
unadjusted risk of death and HF hospitalization in 3029 with
depressed EF.14 However, after adjusting for covariates, AF
was not found to independently predict mortality. Similarly,
Middlekauff et al7 showed that patients in AF and advanced
HF had significantly worse 1-year survival than patients in
sinus rhythm. The importance of AF in patients developing
HF after acute myocardial infarction was shown in the
198
Table 4.
Circ Heart Fail
March 2012
Unadjusted and Adjusted Odds Ratios for In-Hospital Clinical Outcomes
Clinical Outcomes
Event Rate
AF, %
Event Rate
SR, %
Unadjusted
OR
Lower
95% CI
Upper
95% CI
P Value
Adjusted
OR
Lower
95% CI
Upper
95% CI
P Value
Current AF vs SR
In-hospital mortality
4.00
2.63
1.52
1.40
1.65
⬍0.0001
1.17
1.05
1.29
0.0029
LOS ⱕ4 d, median
48.84
41.49
1.39
1.34
1.44
⬍0.0001
1.29
1.22
1.36
⬍0.0001
Discharge: other than home
28.15
19.72
1.49
1.42
1.55
⬍0.0001
1.12
1.07
1.18
⬍0.0001
Preexisting AF vs SR
In-hospital mortality
3.88
2.63
1.47
1.33
1.62
⬍0.0001
1.12
0.99
1.25
0.0637
LOS ⱕ4 d, median
47.56
41.49
1.33
1.28
1.39
⬍0.0001
1.22
1.15
1.30
⬍0.0001
Discharge: other than home
28.41
19.72
1.50
1.43
1.58
⬍0.0001
1.10
1.05
1.17
0.0003
New AF vs SR
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
In-hospital mortality
4.43
2.63
1.70
1.52
1.89
⬍0.0001
1.29
1.10
1.52
0.0023
LOS ⱕ4 d, median
53.56
41.49
1.63
1.53
1.75
⬍0.0001
1.52
1.38
1.68
⬍0.0001
Discharge: other than home
27.15
19.72
1.44
1.36
1.51
⬍0.0001
1.15
1.07
1.24
0.0003
Variables included in adjustment: demographics: age (per 10 years), race (white/black/Hispanic/others), and sex. Medical history (imputed to no): chronic obstructive
pulmonary disease, stroke, depression, diabetes, renal failure, dyslipidemia, hypertension, peripheral vascular disease, smoking, ischemic etiology, and valve disease.
Vitals (imputed with group-specific medians): heart rate (splines with knots at 75 and 105), systolic blood pressure (per 5 mm Hg), and ejection fraction (dichotomized
at 40). Laboratories (not imputed): sodium (splines with knots at 35 and 40), hemoglobin (truncated at 12), creatinine (truncated at 1 and 3.5 mg/dL), and urea (per
5 units). Hospital (not imputed): region, academic status, and heart transplant capability.
AF indicates atrial fibrillation; SR, sinus rhythm; OR, odds ratio; CI, confidence intervals; LOS, length of stay.
Valsartan in Acute Myocardial Infarction (VALIANT) trial,
where AF was associated with greater long-term morbidity
and mortality in 14 703 patients.15 The results of these studies
are in contrast to smaller prior studies were AF was not
associated with worse long-term outcome. More specifically,
the Vasodilator Heart Failure Trial (V-HeFT) showed no
difference in mortality in 1427 outpatients with mild to
moderate HF.8 Similarly, in 2 smaller studies, AF was not
Figure 2. Hospital outcomes stratified by atrial fibrillation (AF) groups. Patients in new-onset AF had lengthier hospital stay and higher
hospital mortality than both patients with preexisting AF and patients in sinus rhythm. Patients in AF (new onset or preexisting) were
more likely to be discharged to a facility other than home.
Mountantonakis et al
Adverse Hospital Outcomes in HF Patients With AF
199
Table 5. Adjusted Odds Ratios for In-Hospital Outcomes for AF Subgroups Stratified by Preserved and
Reduced Ejection Fraction
Clinical Outcomes
EF
Subgroup
Adjusted
OR
Lower
95% CI
Upper
95% CI
Adjusted P Value
for Interaction
0.9654
Current AF vs SR
In-hospital mortality
LOS ⱕ4 d, median
Discharge: other than home
EF ⱖ40%
1.21
1.07
1.37
EF ⬍40%
1.22
1.01
1.46
EF ⱖ40%
1.25
1.18
1.34
EF ⬍40%
1.35
1.26
1.45
EF ⱖ40%
1.10
1.03
1.18
EF ⬍40%
1.17
1.09
1.26
EF ⱖ40%
1.18
1.04
1.35
EF ⬍40%
1.14
0.90
1.43
EF ⱖ40%
1.21
1.13
1.29
EF ⬍40%
1.26
1.17
1.36
EF ⱖ40%
1.08
1.01
1.16
EF ⬍40%
1.16
1.08
1.25
0.0521
0.1987
Preexisting AF vs SR
In-hospital mortality
LOS ⱕ4 d, median
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
Discharge: other than home
0.7915
0.2269
0.1529
New AF vs SR
In-hospital mortality
LOS ⱕ4 d, median
Discharge: other than home
EF ⱖ40%
1.22
0.99
1.50
EF ⬍40%
1.45
1.14
1.85
EF ⱖ40%
1.42
1.25
1.61
EF ⬍40%
1.72
1.51
1.97
EF ⱖ40%
1.14
1.04
1.25
EF ⬍40%
1.18
1.05
1.33
0.2683
0.0208
0.6439
AF indicates atrial fibrillation; EF, ejection fraction; SR, sinus rhythm; OR, odds ratio; CI, confidence intervals; and LOS, length of
stay.
Variables included in adjustment: demographics: age (per 10 years), race (whites/blacks/Hispanic/others), and sex. Medical history
(imputed to no): chronic obstructive pulmonary disease, stroke, depression, diabetes, renal failure, dyslipidemia, hypertension,
peripheral vascular disease, smoking, ischemic etiology, and valve disease. Vitals (imputed with group-specific medians): heart rate
(splines with knots at 75 and 105) and systolic blood pressure (per 5 mm Hg). Laboratories (not imputed): sodium (splines with knots
at 35 and 40), hemoglobin (truncated at 12), creatinine (truncated at 1 and 3.5 mg/dL), and urea (per 5 units). Hospital (not imputed):
region, academic status, and heart transplant capability.
associated with worse mortality in patients with advanced
HF.11,12 In the Danish Investigations of Arrhythmia and
Mortality on Dofetilide (DIAMOND) study, Pedersen et al16
studied the association of AF with outcomes in 3587 patients
admitted with decompensated HF. In this study, in-hospital
mortality was similar between patients presenting in AF and
sinus rhythm.16
An interesting finding of our study is that the worse
hospital outcomes were in patients with newly diagnosed AF.
New-onset AF had been previously linked to long-term
mortality in the outpatient setting. Ahmed et al26 showed in a
retrospective analysis of 944 hospitalized elderly patients
with HF that new onset AF but not preexisting AF carried a
significantly higher risk for 4 year all-cause mortality. Similar
association between new-onset AF and mortality and absence
of association between preexisting AF and mortality was
reported in an analysis of COMET trial.14 Whether this
association between new-onset AF and mortality is causative
or development of AF is simply a marker of advanced disease
is unclear. In addition to the adverse hemodynamic consequences of AF, management of new-onset AF, including
initiation of anticoagulation and antiarrhythmic agents, espe-
cially in setting of HF, may be challenging and result in
higher morbidity and mortality.
In our study, the associations between AF and outcomes in
the adjusted analyses were similar for patients with preserved
and low EF (EF ⱖ40% and ⬍40%). One could have expected
worse outcomes in patients with AF and preserved EF;
however, our finding helps to reinforce the overall findings
and shows the relationship between AF and outcomes applies
irrespective of type of HF. Clear associations between AF and
mortality in either hospitalized patients with preserved EF HF
has not, to the best of our knowledge, been previously
demonstrated.
Limitations
The retrospective nature of the analysis is an inherent
limitation of the study. The registry is dependent on accuracy
of data and the completeness of data abstraction from medical
charts. Certain variables were not routinely collected or
missing from the database. Limitation in some types of data
also precluded the assessment of important variables such as
evolution of rhythm during hospitalization. Therefore, we
were unable to report the number of patients who although
they presented in AF, subsequently converted spontaneously
200
Circ Heart Fail
March 2012
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
to sinus rhythm and the number of patients who developed
AF during the course of their hospital stay. In addition, we
were not able to evaluate the association of restoration of
sinus rhythm with hospital outcomes. It is possible that
patients with asymptomatic chronic AF may have been
included in the new-onset AF group. Determination of EF
could have largely been affected by the underlying rhythm as
well as rate and therefore the timing of echocardiogram could
have also influenced differences in EF. Given the observational nature of the study, residual measured and unobserved
variables may have confounded some or all of the findings.
Because of the large number of patients in this study, some
small differences might lead to statistical significance but
lack clinical relevance. Postdischarge data, including mortality and readmissions, were not available and will require
further study. There is also the possibility that the cohort
derived from a voluntary registry may differ from hospitals
not willing to participate, but prior research evaluating this
hypothesis found participants in registries to have in-hospital
mortality similar to that in nonparticipants.27
Conclusions
Among hospitals participating in GWTG-HF, AF is present in
a third of patients who were hospitalized with HF. Patients in
AF were older and more likely to have stroke and valvular
heart disease. AF, particularly newly diagnosed, was independently associated with adverse in-hospital outcomes
among patients hospitalized with HF, including a significantly higher mortality rate. This association appears to be
present for both patients with reduced as well as preserved
left ventricular EF. Whether AF or its related treatment is the
mediator of these adverse outcomes or whether AF represents
a marker of greater HF severity cannot be addressed by this
analysis. Future studies addressing the importance of prompt
restoration of sinus rhythm in patients in AF are important in
better defining the optimal treatment of these patients.
Sources of Funding
The GWTG-HF program is provided by the American Heart Association. The GWTG-HF program is currently supported in part by
Medtronic, Ortho-McNeil, and the American Heart Association
Pharmaceutical Roundtable. GWTG-HF has been funded in the past
through support from GlaxoSmithKline.
Disclosures
Dr Bhatt received research grants (significant) from Amarin, Astra
Zeneca, Bristol-Myers Squibb, Eisai, Ethicon, Medtronic, Sanofi
Aventis, and The Medicines Company. Dr Hernandez received
support from Research Johnson & Johnson (significant) and Amylin
(significant) and served as a consultant for Corthera (modest). Dr
Grau-Sepulveda is an employee of DCRI. Dr Peterson serves as the
principal investigator of the analytic center for the Get With The
Guidelines Program. Dr Fonarow received support from Research
National Heart, Lung, and Blood Institute (significant); served as a
consultant for Novartis (significant) and Scios (modest); and received honorarium from Medtronic (modest).
References
1. Braunwald E. Shattuck Lecture: cardiovascular medicine at the turn of the
millennium: triumphs, concerns, and opportunities. N Engl J Med. 1997;
337:1360 –1369.
2. Stevenson WG, Stevenson LW. Atrial fibrillation and heart failure-five
more years. N Engl J Med. 2004;351:2437–2440.
3. Koelling TM, Chen RS, Lubwama RN, L’Italien GJ, Eagle KA. The
expanding national burden of heart failure in the United States: the
influence of heart failure in women. Am Heart J. 2004;147:74 –78.
4. Go AS, Hylek EM, Phillips KA, Chang Y, Henault LE, Selby JV, Singer
DE. 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 –2375.
5. Maisel WH, Stevenson LW. Atrial fibrillation in heart failure: epidemiology, pathophysiology, and rationale for therapy. Am J Cardiol. 2003;
91:2D– 8D.
6. Wang TJ, Larson MG, Levy D, Vasan RS, Leip EP, Wolf PA,
D’Agostino RB, Murabito JM, Kannel WB, Benjamin EJ. Temporal
relations of atrial fibrillation and congestive heart failure and their joint
influence on mortality: the Framingham Heart Study. Circulation. 2003;
107:2920 –2925.
7. Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance
of atrial fibrillation in advanced heart failure: a study of 390 patients.
Circulation. 1991;84:40 – 48.
8. Carson PE, Johnson GR, Dunkman WB, Fletcher RD, Farrell L, Cohn JN.
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(suppl):VI-102–VI-110.
9. The Digitalis Investigation Group. The effect of digoxin on mortality and
morbidity in patients with heart failure. N Eng J Med. 1997;336:525–533.
10. Dries DL, Exner DV, Gersh BJ, Domanski MJ, Waclawiw MA, Stevenson LW. 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.
11. Mahoney P, Kimmel S, DeNofrio D, Wahl P, Loh E. Prognostic significance of atrial fibrillation in patients at a tertiary medical center referred
for heart transplantation because of severe heart failure. Am J Cardiol.
1999;83:1544 –1547.
12. Crijns HJ, Tjeerdsma G, de Kam PJ, Boomsma F, van Gelder IC, van den
Berg MP, van Veldhuisen DJ. Prognostic value of the presence and
development of atrial fibrillation in patients with advanced chronic heart
failure. Eur Heart J. 2000;21:1238 –1245.
13. Wyse DG, Love JC, Yao Q, Carlson MD, Cassidy P, Greene LH, Martins
JB, Ocampo C, Raitt MH, Schron E, Stamato NJ, Olarte A. Atrial
fibrillation: a risk factor for increased mortality: an AVID registry analysis. J Interv Card Electrophysiol. 2001;5:267–273.
14. Swedberg K, Olsson LG, Charlesworth A, Cleland J, Hanrath P, Komajda
M, Metra M, Torp-Pedersen C, Poole-Wilson P. Prognostic relevance of
atrial fibrillation in patients with chronic heart failure on long-term
treatment with beta-blockers: results from COMET. Eur Heart J. 2005;
26:1303–1308.
15. Køber L, Swedberg K, McMurray JJ, Pfeffer MA, Velazquez EJ, Diaz R,
Maggioni AP, Mareev V, Opolski G, Van de Werf F, Zannad F, Ertl G,
Solomon SD, Zelenkofske S, Rouleau JL, Leimberger JD, Califf RM.
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;8:591–598.
16. Pedersen OD, Søndergaard P, Nielsen T, Nielsen SJ, Nielsen ES, FalstieJensen N, Nielsen I, Køber L, Burchardt H, Seibaek M, Torp-Pedersen C;
DIAMOND study group investigators. Atrial fibrillation, ischaemic heart
disease, and the risk of death in patients with heart failure. Eur Heart J.
2006;27:2866 –2870.
17. Hernandez AF, Fonarow GC, Liang L, Al-Khatib SM, Curtis LH,
LaBresh KA, Yancy CW, Albert NM, Peterson ED. Sex and racial
differences in the use of implantable cardioverter-defibrillators among
patients hospitalized with heart failure. JAMA. 2007;298:1525–1532.
18. Horwich TB, Hernandez AF, Liang L, Albert NM, Labresh KA, Yancy
CW, Fonarow GC; Get With The Guidelines Steering Committee and
Hospitals. Weekend hospital admission and discharge for heart failure:
association with quality of care and clinical outcomes. Am Heart J.
2009;158:451– 458.
19. Senni M, Tribouilloy CM, Rodeheffer RJ, Jacobsen SJ, Evans JM, Bailey
KR, Redfield MM. Congestive heart failure in the community: a study of
all incident cases in Olmsted County, Minnesota, in 1991. Circulation.
1998;98:2282–2289.
20. Deedwania PC, Singh BN, Ellenbogen K, Fisher S, Fletcher R, Singh SN.
Spontaneous conversion and maintenance of sinus rhythm by amiodarone
in patients with heart failure and atrial fibrillation: observations from the
Mountantonakis et al
21.
22.
23.
24.
Adverse Hospital Outcomes in HF Patients With AF
Veterans Affairs Congestive Heart Failure Survival Trial of Antiarrhythmic Therapy (CHF-STAT): the Department of Veterans Affairs
CHF-STAT Investigators. Circulation. 1998;98:2574 –2579.
Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, Redfield
MM. Trends in prevalence and outcome of heart failure with preserved
ejection fraction. N Engl J Med. 2006;355:251–259.
Lloyd-Jones DM, Wang TJ, Leip EP, Larson MG, Levy D, Vasan RS,
D’Agostino RB, Massaro JM, Beiser A, Wolf PA, Benjamin EJ. Lifetime
risk for development of atrial fibrillation: the Framingham Heart Study.
Circulation. 2004;110:1042–1046.
Zipes DP. Atrial fibrillation: a tachycardia-induced atrial cardiomyopathy. Circulation. 1997;95:562–564.
Pozzoli M, Cioffi G, Traversi E, Pinna GD, Cobelli F, Tavazzi L.
Predictors of primary atrial fibrillation and concomitant clinical and
201
hemodynamic changes in patients with chronic heart failure: a prospective study in 344 patients with baseline sinus rhythm. J Am Coll
Cardiol. 1998;32:197–204.
25. Weinfeld MS, Drazner MH, Stevenson WG, Stevenson LW. Early
outcome of initiating amiodarone for atrial fibrillation in advanced heart
failure. J Heart Lung Transplant. 2000;19:638 – 643.
26. Ahmed A, Perry GJ. Incident atrial fibrillation and mortality in older
adults with heart failure. Eur J Heart Fail. 2005;7:1118 –1121.
27. Curtis LH, Greiner MA, Hammill BG, Di Martino LD, Shea AM, Hernandez AF, Fonarow GC. Representativeness of a national heart failure
quality-of-care registry: comparison of OPTIMIZE-HF and nonOPTIMIZE-HF Medicare patients. Circ Cardiovasc Qual Outcomes.
2009;2:377–384.
CLINICAL PERSPECTIVE
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
Atrial fibrillation and heart failure are both major causes of cardiovascular morbidity and mortality. Although there have
been extensive studies of the association of atrial fibrillation with outcomes in patients with chronic heart failure, far less
is known about the association of atrial fibrillation, particularly newly diagnosed atrial fibrillation, with acute outcomes in
patients hospitalized with heart failure. Using data from 99 810 patients enrolled in 255 hospitals participating in Get With
The Guidelines–Heart Failure, this study evaluated the clinical characteristics, management, length of stay, and mortality
of hospitalized heart failure patients with and without atrial fibrillation. Of patients hospitalized with heart failure, 31.4%
presented with atrial fibrillation, of which 21.3% were newly diagnosed. Patients in atrial fibrillation were older, were more
likely to have history of stroke and valvular heart disease, and had higher left ventricular ejection fraction. Heart failure
patients with atrial fibrillation were more likely to have longer length of stay, to be discharged to a facility other than home,
and to have higher hospital mortality. Atrial fibrillation, particularly newly diagnosed, was independently associated with
higher in-hospital mortality. These associations applied to both patients with reduced as well as preserved left ventricular
ejection fraction. Atrial fibrillation, or its related treatment, may be the cause of these adverse outcomes or atrial fibrillation
may just represent a marker of greater heart failure severity. Whether prompt restoration of sinus rhythm would improve
outcomes in patients hospitalized with heart failure and new-onset or paroxysmal atrial fibrillation is unclear and requires
further study.
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 11, 2017
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
Stavros E. Mountantonakis, Maria V. Grau-Sepulveda, Deepak L. Bhatt, Adrian F. Hernandez,
Eric D. Peterson and Gregg C. Fonarow
Circ Heart Fail. 2012;5:191-201; originally published online February 23, 2012;
doi: 10.1161/CIRCHEARTFAILURE.111.965681
Circulation: Heart Failure is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX
75231
Copyright © 2012 American Heart Association, Inc. All rights reserved.
Print ISSN: 1941-3289. Online ISSN: 1941-3297
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circheartfailure.ahajournals.org/content/5/2/191
Data Supplement (unedited) at:
http://circheartfailure.ahajournals.org/content/suppl/2012/02/23/CIRCHEARTFAILURE.111.965681.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation: Heart Failure can be obtained via RightsLink, a service of the Copyright Clearance Center, not
the Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation: Heart Failure is online at:
http://circheartfailure.ahajournals.org//subscriptions/
Supplemental Material
Appendix Table 1. Prevalence of atrial fibrillation based on prior history of heart failure
Atrial Fibrillation
Current Atrial Fibrillation
Type of Atrial Fibrillation
Newly Diagnosed HF Prior History of HF
9,414
21,699
28.8%
32.8%
No
23,253
44,501
71.2%
67.2%
New Onset AF
2,971
3,488
9.1%
5.3%
Pre-existing AF
6,443
18,211
19.7%
27.5%
Sinus Rhythm
23,253
44,501
71.2%
67.2%
P
Yes
<.0001
<.0001
Actual numbers are presented on first row of each variable and percentages on the second.
AF atrial fibrillation; HF heart failure.
Appendix Table 2. Medical therapy prior to admission stratified by AF status
Variable
Sinus
Rhythm
Current
AF
ACE inhibitor
24,809
39.9%
10,554
37.2%
Aldosterone antagonist
6,344
10.2%
Angiotensin receptor
blocker
Pre-Existing
AF
New Onset
AF
<.0001
8,449
37.6%
2,105
35.4%
<.0001
0.0017
2,801
9.9%
0.1187
2,338
10.4%
463
7.8%
<.0001
<.0001
9,522
15.3%
4,212
14.8%
0.0630
3,313
14.8%
899
15.1%
0.1384
0.4760
Antiarrhythmic
4,410
7.1%
3,591
12.6%
<.0001
3,117
13.9%
474
8.0%
<.0001
<.0001
Aspirin
30,188
48.5%
11,955
42.1%
<.0001
9,337
41.6%
2,618
44.0%
<.0001
0.0006
Beta-Blocker
34,671
55.7%
15,959
56.2%
0.2027
13,136
58.5%
2,823
47.5%
<.0001
<.0001
11,377 6,90
21.4%
<.0001
4,935
22.0%
1,155
19.4%
<.0001
<.0001
Ca channel blocker
18.3%
P
P#
P^
2 Digoxin
8,276
13.3%
7,597
26.7%
<.0001
6,559
29.2%
1,038
17.5%
<.0001
<.0001
Diuretic
39,637
63.7%
20,216
71.2%
<.0001
16,503
73.5%
3,713
62.5%
<.0001
<.0001
Hydralazine
5,392
8.7%
1,709
6.0%
<.0001
1,411
6.3%
298
5.0%
<.0001
0.0002
Statin
28,300
45.5%
11,857
41.7%
<.0001
9,504
42.3%
2,353
39.6%
<.0001
0.0001
Nitrate
12,356
19.9%
4,840
17.0%
<.0001
3,931
17.5%
909
15.3%
<.0001
<.0001
Warfarin
8,974
14.4%
13,945
49.1%
<.0001
12,403
55.2%
1,542
25.9%
<.0001
<.0001
Actual numbers are presented on first row of each variable and percentages on the second.
P# comparison between sinus rhythm, pre-existing AF and new onset AF. P^ comparison between pre-existing
and new onset AF. Statistical significant differences are highlighted with bold P values.
AF atrial fibrillation; ACE angiotensin converting enzyme.
Appendix Table 3. Medications prior to admission among patients with prior history of
HF with reduced ejection fraction (EF <40%).
Variable
Sinus
Rhythm
Current
AF
(N=22,226)
(N=8,865)
ACE inhibitor
10,015
48.1%
3,725
45.5%
Aldosterone antagonist
3,758
18.1%
Angiotensin receptor
blocker
P
Pre-Existing
AF
New Onset
AF
(N=7,421)
(N=1444)
<.0001
3,144
46.0%
1313
16.0%
<.0001
2,950
14.2%
1150
14.1%
Beta-Blocker
13,621
65.4%
Digoxin
4,717
22.7%
P#
P^
581
42.9%
<.0001
0.0375
1,134
16.6%
179
13.2%
<.0001
0.0020
0.7721
974
14.3%
176
13.0%
0.4635
0.2266
5203
63.5%
0.0021
4,437
64.9%
766
56.6%
<.0001
<.0001
2,711
33.1%
<.0001
2,384
34.9%
327
24.2%
<.0001
<.0001
3 Diuretic
14,992
72.0%
6,273
76.6%
<.0001
5,297
77.5%
976
72.1%
<.0001
<.0001
Actual numbers are presented on first row of each variable and percentages on the second.
Appendix Tables 4. Adjusted odds ratios for left and right heart catheterization procedures for
different subgroups based on presenting rhythm.
Procedures
Adjusted Lower Upper
OR
95% CI 95% CI P-value
Current AF. vs. SR
Right Cardiac
Catheterization
0.87
0.77
0.98
0.0236
Left Cardiac
Catheterization
0.75
0.67
0.83
<.0001
Pre-exist. AF vs. SR
Right Cardiac
Catheterization
0.77
0.68
0.89
0.0002
Left Cardiac
Catheterization
0.62
0.56
0.69
<.0001
New AF. vs. SR
Right Cardiac
Catheterization
1.21
0.93
1.57
0.1497
Left Cardiac
Catheterization
1.15
0.92
1.44
0.2260
AF atrial fibrillation; SR sinus rhythm.