European Journal of Heart Failure (2010) 12, 1193–1202

doi:10.1093/eurjhf/hfq138

Acute pulmonary oedema: clinical

characteristics, prognostic factors, and
in-hospital management
John T. Parissis 1*, Maria Nikolaou 1, Alexandre Mebazaa 2, Ignatios Ikonomidis 1,
Juan Delgado 3, Fabio Vilas-Boas 4, Ioannis Paraskevaidis 1, Antony Mc Lean 5,
Dimitrios Kremastinos 1, and Ferenc Follath 6
1
Heart Failure Clinic and Second Cardiology Department, Attikon University Hospital, University of Athens, Athens, Greece; 2Department of Anesthesiology and Critical Care
Medicine, Hospital Lariboisière, APHP, Université Paris Diderot Paris 7, U942 Inserm, Paris, France; 3Heart Failure and Transplant Unit, Cardiology Department, Hospital Doce de
Octubre, Madrid, Spain; 4Cardiology Division and Heart Failure and Transplantation Program, Hospital Espanhol, Salvador, Bahia, Brazil; 5Department of Intensive Care Medicine,
Nepean Hospital, University of Sydney, Penrith, New South Wales, Australia; and 6Department of Internal Medicine, University Hospital Zurich, Zurich, Switzerland

Received 27 February 2010; revised 10 May 2010; accepted 14 May 2010; online publish-ahead-of-print 13 September 2010

Aims Acute pulmonary oedema (APE) is the second, after acutely decompensated chronic heart failure (ADHF), most fre-
quent form of acute heart failure (AHF). This subanalysis examines the clinical profile, prognostic factors, and manage-
ment of APE patients (n ¼ 1820, 36.7%) included in the Acute Heart Failure Global Survey of Standard Treatment
(ALARM-HF).
.....................................................................................................................................................................................
Methods ALARM-HF included a total of 4953 patients hospitalized for AHF in Europe, Latin America, and Australia. The final
and results diagnosis was made at discharge, and patients were classified according to European Society of Cardiology guidelines.
Patients with APE had higher in-hospital mortality (7.4 vs. 6.0%, P ¼ 0.057) compared with ADHF patients (n ¼ 1911,
38.5%), and APE patients exhibited higher systolic blood pressures (P , 0.001) at admission and higher left ventricular
ejection fraction (LVEF, P , 0.01) than those with ADHF. These patients also had a higher prevalence of diabetes
(P , 0.01), arterial hypertension (P , 0.001), peripheral vascular disease (P , 0.001), and chronic renal disease
(P , 0.05). They were also more likely to receive intravenous (i.v.) diuretics (P , 0.001), i.v. nitrates (P , 0.01), dopa-
mine (P , 0.05), and non-invasive ventilation (P , 0.001). Low systolic blood pressure (P , 0.001), low LVEF
(,0.05), serum creatinine ≥1.4 mg/dL (P , 0.001), history of cardiomyopathy (P , 0.05), and previous cardiovascu-
lar event (P , 0.001) were independently associated with increased in-hospital mortality in the APE population.
.....................................................................................................................................................................................
Conclusion APE differs in clinical profile, in-hospital management, and mortality compared with ADHF. Admission characteristics
(systolic blood pressure and LVEF), renal function, and history may identify high-risk APE patients.
-----------------------------------------------------------------------------------------------------------------------------------------------------------
Keywords Acute heart failure † Acute pulmonary oedema † Medical management † Prognosis

variety of AHF syndromes and the lack of a universally accepted

Introduction definition of AHF.
Acute heart failure (AHF) is a heterogeneous syndrome associated The international Acute Heart Failure Global Survey of Standard
with increased morbidity and mortality worldwide. Recent AHF Treatment (ALARM-HF) was planned to assess the utility of the
registries in the USA, Europe, and Japan have grouped various clini- European Society of Cardiology (ESC) guidelines7 for describing
cal situations and provided important information about patients’ and comparing the clinical presentation, management, and
clinical characteristics and prognosis.1 – 6 However, these registries outcome of AHF under ‘real life’ conditions in different countries
have shown a wide range of in-hospital mortality underlining the and regions of the world. AHF can present as acute de novo

* Corresponding author. Tel: +30 210 6123720, Fax: +30 210 5832195, Email: [email protected]
Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2010. For permissions please email: [email protected].
1194 J.T. Parissis et al.

(new onset heart failure) or worsening of chronic heart failure Survey description
(CHF) and can be classified into several distinct clinical conditions The hospital sample was recruited to be representative according to
such as (i) acutely decompensated chronic heart failure (ADHF) geographical region, hospital size (by number of beds), sector (public
defined as signs and symptoms of AHF that do not fulfil the criteria vs. private), and type (university vs. non-teaching status). Paper-based
for cardiogenic shock, pulmonary oedema, or hypertensive crisis, data collection was conducted over the period from October 2006 to
(ii) hypertensive AHF defined as signs and symptoms of heart March 2007. Patient case report forms were completed for five to
failure accompanying high blood pressure, relatively preserved eight consecutive patients at or close to discharge, based on medical
ejection fraction, and chest X-ray compatible with acute pulmon- records. Diagnoses at discharge were classified according to ESC cat-
egories by the responsible cardiologists and/or intensive care unit
ary oedema (APE), (iii) APE defined as alveolar or interstitial
(ICU) physicians. Where two or more categories were indicated by a
oedema verified by chest X-ray and/or with O2 saturation ,90%
respondent, the final appropriate classification was adjudicated accord-
on room air prior to treatment accompanied by severe respiratory
ing to the additional clinical data supplied. The retrospective method-
distress, with crackles over the lungs and orthopnoea, (iv) cardio- ology also enabled the assessment of both primary (present on
genic shock defined as tissue hypoperfusion induced by heart admission) and secondary (occurring in hospital) causes of AHF. Clinical
failure after correction of pre-load, accompanied by systolic history, co-morbidities, precipitating factors, signs and symptoms, and
blood pressure ,90 mmHg, low urine output ,0.5 mL/kg/h medication (admission as well as discharge) were recorded. Details of
with a pulse rate .60 b.p.m., (v) high output heart failure, and intravenous (i.v.) drugs including timing and location of initiation as
(vi) right heart failure characterized by low output, hypotension, well as dosage and duration were also registered. Where available, echo-
increased jugular venous pressure, and liver enlargement. One of cardiography data were collected at diagnosis and/or prior to discharge.
the main problems in managing patients with AHF syndromes is Finally, there were no exclusion criteria in this survey.
the fact that each patient represents a different entity, a different
clinical scenario, derived from a different pathophysiological mech- Discrimination of acute pulmonary oedema
anism with a different outcome. As a result, grouping patients for and acute decompensated heart failure
educational purposes may only partly reflect reality. The clinical ESC definitions were used to classify APE and ADHF.7 APE is a diag-
classification was suggested by the ESC as a simple, easy, and nosis based on patient’s history (abrupt onset), clinical examination
rapidly applied classification that provides useful information (orthopnoea and rales), chest X-ray (alveolar oedema), oxygen desa-
turation, and treatment response. Unfortunately, there is no definite
since the patients’ admission, although overlapping cannot be
measurement that can distinguish the level of alveolar congestion,
avoided. This classification was only slightly modified in the most
since this is related not only to the pressures in the microcirculation,
recent ESC guidelines published in 2008, by omitting the entity but also to the membrane properties. Right catheterization was avail-
of high output heart failure and inserting the acute coronary able in only 12% and natriuretic peptides in 6.2% of the patient popu-
syndrome entity.8 lation and therefore did not help in the differential diagnosis of the two
Undoubtedly, APE represents the most dramatic condition, fol- conditions by the treating physicians.
lowing cardiogenic shock, although it is poorly described in the The distinction between APE and ADHF is not always clear, as these
majority of registries. Our study therefore has three main objec- two clinical entities represent a continuous range of the same patho-
tives: first to evaluate the clinical characteristics of APE and physiology, with no clear cut-offs. As a result, the discrimination is
compare them with ADHF, because these two AHF phenotypes based primarily on the severity and often on the rapidity of
are the most frequent, involving about 76% of the overall patient symptom evolution, as judged by the clinician, and secondarily by
the X-ray findings that may or may not support the diagnosis (e.g.
population in the ALARM-HF survey; secondly, to assess the prog-
flash pulmonary oedema).
nostic markers of APE; and thirdly, to describe the management of
APE patients registered in ALARM-HF.
Statistics
Statistical analysis was performed using the SPSS 13.0 statistical soft-
ware package (SPSS Inc., Chicago, IL, USA). Categorical variables are
Methods presented as counts and percentages and quantitative variables as
The ALARM-HF survey was an in-hospital observational survey that mean and standard deviation. Subgroups of patients were compared
included patients hospitalized for AHF in Europe, Latin America, and using x2 tests for categorical variables and t-test or analysis of variance
Australia. Anonymized data were collected for 4953 patients. This Wilcoxon rank-sum tests for quantitative variables. Age categories
database contains detailed patient characteristics from initial presen- were compared using the x2 test for trend. Receiver operating charac-
tation in the hospital or emergency department until discharge, trans- teristic (ROC) analysis was used to identify the best cut-off value for
fer, or in-hospital death. The ALARM-HF survey included patients the continuous variable of serum creatinine. Univariate and multivari-
hospitalized with AHF in community, tertiary, and academic centres. able logistic regression analysis was used to identify prognostic predic-
The database includes information on demography, medical history, tors. All variables with P , 0.1 at univariate analysis were included in
baseline clinical characteristics, initial evaluation, treatment received, the multivariable model. Both backward and forward stepwise analyses
procedures performed, and hospital course. were used to determine the independent predictors of outcome. All
Standardized definitions are used for all patient-related variables, interactions between covariates were assessed. The odds ratio and
clinical diagnoses, and hospital outcomes. Institutional Review Board corresponding 95% confidence intervals are given for each covariate.
approval was required for all participating centres; however, informed Kaplan– Meier in-hospital survival curves were constructed by dividing
consent of individuals was not required for survey entry. Diagnosis was patients into subgroups according to the quartiles of systolic blood
made at discharge, and patients were classified according to the defi- pressure or to the cut-off value of serum creatinine. The survival
nition and classification of the ESC guidelines 2005.7 rates between the various subgroups were compared using the log
APE management 1195

rank test. All tests were two-sided, and P-values less than 0.05 were as NYHA III, respectively]. Differences in symptoms and clinical
considered as indicating significant differences. signs between these two conditions are also described in Table 1.
APE patients exhibited higher systolic blood pressure (P ,
0.001), higher heart rate (P , 0.001), and lower levels of oxygen
Results saturation (P , 0.001) at admission. These patients also had a
slightly (but statistically significant) higher left ventricular ejection
The ALARM-HF survey included a total of 4953 patients (France
fraction (LVEF) (P , 0.01) (Table 2). About 14% of the APE
n ¼ 588, Germany n ¼ 617, Italy n ¼ 679, Spain n ¼ 700, UK
patients vs. 39% of the ADHF patients had no evidence of conges-
n ¼ 623, Greece n¼255, Turkey n ¼ 628, Mexico n ¼ 601, and
tion (interstitial or alveolar oedema) on chest radiography. In these
Australia n ¼ 262), admitted to 666 hospitals of different sizes
cases, the APE diagnosis was based on severe respiratory distress
and types and managed by either cardiologists (n ¼ 655) or ICU
accompanying low arterial saturation oxygen (,90%) and signs
physicians (n ¼ 287). The clinical characteristics of the whole
implying congestion as described previously.7
cohort of AHF patients have already been presented.9 Thirty-five
Cardiovascular and non-cardiovascular co-morbidities are also
per cent of the patients needed ICU/Coronary Care Unit (CCU)
summarized in Table 1. Acute coronary syndrome, new onset of
management. APE and ADHF were the most common presenta-
arrhythmia, and post-surgery clinical worsening were aetiologies
tions, affecting 37 and 39% of the total patient population, respect-
of exacerbation favouring APE, whereas poor compliance with
ively. Moreover, cardiogenic shock presented with a frequency of
treatment and dietary habits as well as infection favoured ADHF.
12%, hypertensive heart failure 7%, right ventricular failure 4%,
Consequently, APE patients more frequently had positive
and high output heart failure 1%. The percentage of APE and
markers of myocardial injury.
ADHF patients in the different countries is shown in Figure 1.
Differences in pre-admission medications are described in
Table 1 summarizes the differences in the clinical profile and lab-
Table 3. Although medication was optimized during hospitalization
oratory findings between the two conditions. No differences in
in both groups, the differences persisted at discharge. In-hospital
gender or age were observed.
treatment modalities also differed between the two conditions,
Concerning the transfer to hospital, APE patients were more
as shown in Table 4.
often already hospitalized compared with ADHF patients (14.6
vs. 9.3%) or admitted at the emergency room by ambulance
(48.6 vs. 41.3%), whereas ADHF patients were more frequently In-hospital outcome
admitted independently (16.4 vs. 11.0%) or from referral centres APE patients experienced worse in-hospital outcomes compared
(27.0 vs. 19.7% P , 0.001 for all comparisons). with ADHF patients, with increased in-hospital mortality (7.4 vs.
A higher percentage of APE patients had de novo heart failure 6%, P ¼ 0.057), increased ICU discharge (P , 0.05), and increased
(39.2 vs. 28.1% for ADHF, P , 0.001), and APE patients were transfer to rehabilitation centres (P , 0.05) (Table 5). Further-
more frequently diagnosed later during the hospitalization (24.9 more, 80.8% of the APE vs. 65.2% (P , 0.001) of the ADHF
vs. 17.7% of patients had a secondary diagnosis of APE and patients required ICU/CCU during hospitalization.
ADHF, respectively, P , 0.001). The onset of symptoms was Univariate analysis of prognosis for patients with APE is demon-
more abrupt in APE patients (,6 h in 42.9% of APE vs. 34.0% of strated in Table 6. The multivariate model including those par-
ADHF, 6 h to 5 days in 10.4% of APE vs. 26.5% of ADHF, P , ameters univariately associated with survival revealed that low
0.001) and the clinical status more dramatic [39.0% of APE vs. systolic blood pressure (P ¼ 0.019), low LVEF (P , 0.05), serum
30.1% of ADHF patients were classified as New York Heart creatinine ≥1.4 mg/dL (P , 0.001), history of cardiomyopathy
Association (NYHA) IV, whereas 31.6 vs. 42.8% were classified (P , 0.05) or previous cardiovascular event (P , 0.001),

Figure 1 Percentages of acute pulmonary oedema (APE) vs. acutely decompensated chronic heart failure (ADHF) patients in countries of
ALARM-HF.
1196 J.T. Parissis et al.

Table 1 Differences in clinical presentation between Table 1 Continued

patients with acutely decompensated heart failure
(ADHF) and acute pulmonary oedema (APE) ADHF (n 5 APE (n 5 P-value
1911) 1820)
................................................................................
ADHF APE P-value
(n 5 1911) (n 5 1820) CRT (%) 3.2 1.2 0.001
................................................................................ ICD (%) 5.3 3.2 0.001
Age Smoker (%) 18.4 25.5 0.001
,50 years (%) 9.2 8.1 NS Ex-smoker (%) 37.9 35.9 NS
50–75 years (%) 46.8 47.3 NS ICU hospitalization (%) 65.2 80.8 0.001
.75 years (%) 44.0 47.7 NS
Cardiovascular co-morbidities CAD, coronary artery disease; CHF, chronic heart failure; CKMB, cardiac
Atrial fibrillation/ 27.9 24.7 0.015 isoenzyme of creatinine phosphokinase; ACS, acute coronary syndrome; CRT,
cardiac resynchronization therapy; ICD, implantable cardioverter defibrillator;
flutter (%)
ICU, intensive care unit.
Cardiomyopathy (%) 14.9 11.7 0.002
Peripheral vascular 6.5 9.9 ,0.001
disease (%)
CAD (%) 30.4 31.8 NS Table 2 Vital signs, left ventricular ejection fraction
CHF (%) 40.5 35.2 ,0.001 (LVEF), and chest radiographic findings in acutely
Obesity (%) 25.1 26.1 NS decompensated heart failure (ADHF) and acute
Diabetes mellitus (%) 44.1 48.4 0.005 pulmonary oedema (APE) patients
Dyslipidaemia (%) 38.3 44.8 ,0.001
Arterial hypertension 66.3 72.7 ,0.001 ADHF APE P-value
................................................................................
(%)
Vital signs
Non-cardiovascular co-morbidities
Systolic blood pressure 130.9 + 32.8 138.5 + 37.1 ,0.001
Anaemia (%) 13.9 15.9 0.054
(mmHg)
Benign tumour (%) 0.6 0.6 NS
Heart rate (b.p.m.) 104.7 + 28.8 109.5 + 25.2 ,0.001
Malignancy (%) 3.5 2.7 NS
Oxygen saturation (%) 89.1 + 17.1 87.3 + 7.6 ,0.001
Depression (%) 8.3 8.0 NS
Echocardiography
Dementia (%) 3.9 5.3 0.039
LVEF (%) 37.2 + 14.1 38.8 + 13.2 ,0.01
Hyponatraemia (%) 6.4 5.6 NS
Preserved EF (.45%) 22.3 25.3 0.07
Chronic obstructive 25.3 25.4 NS (%)
pulmonary disease (%)
Chest X-ray
Chronic renal disease 20.4 24.3 0.003
Interstitial or alveolar 61 86 ,0.001
(%)
oedema (%)
Liver disease (%) 3.7 3.4 NS
Symptoms and signs
Cool extremities (%) 18.8 26.0 ,0.001
Dyspnoea (at rest and/ 74.9 75.3 NS secondary diagnosis of APE (P , 0.001), and treatment with diure-
or exertional) (%) tics (P , 0.01) were independently associated with in-hospital
Fatigue (%) 49.1 42.3 ,0.001 outcome (Table 7).
Orthopnoea (%) 51.3 65.6 ,0.001 Table 8 summarizes the clinical characteristics, in-hospital man-
Peripheral oedema (%) 49.8 38.6 ,0.001 agement, and outcomes of patients with APE according to systolic
Jugular venous 40.1 39.6 NS blood pressure quartiles. For the stronger predictors of survival
distension (%) such as systolic blood pressure (relevant quartiles) and serum crea-
Rales (%) 55.4 72.0 ,0.001 tinine, Kaplan–Meier plots are constructed and depicted in
Weight gain (%) 28.5 23.7 ,0.001 Figures 2 and 3, respectively.
Positive cardiac markers
Troponin T (%) 23.9 33.6 0.001
CKMB (%) 17.5 24.1 0.001 Discussion
Precipitating factors
APE is a common manifestation of AHF syndromes and is associ-
ACS (%) 35.0 37.3 0.07
ated with dramatic clinical presentation and worse in-hospital
Arrhythmia (%) 26.6 29.9 0.012
outcome.10 – 12 The principal pathophysiological mechanism is the
Non-compliance (%) 14.5 12.0 0.013
abrupt increase in the pulmonary capillary wedge pressure that
Infection (%) 17.9 15.3 0.02
leads to pulmonary interstitial and alveolar oedema. Unlike
Post-surgery (%) 2.4 4.1 0.003
ADHF, which is characterized by milder and more slowly worsen-
Continued ing symptoms due to chronic cardiac dysfunction per se, APE has
been attributed to both cardiac and vascular failure, with low
APE management 1197

Table 3 Differences in admission (pre-existing) and discharge oral medications between acutely decompensated chronic
heart failure (ADHF) and acute pulmonary oedema (APE) patients

Admission Discharge
............................................................. .............................................................
ADHF APE P-value ADHF APE P-value
...............................................................................................................................................................................
Diuretics 38.0 29.8 ,0.001 68.7 63.7 ,0.001
ACE-I/ARBs 47.5 53.2 ,0.5 63.1 70.1 ,0.01
BB 26.8 21.6 ,0.001 39.1 33.8 0.001
Nitrates 9.8 9.1 NS 20.9 22.3 NS
CCB 0.1 0.2 NS 0.3 0.5 NS
Digoxin 14.3 10.2 ,0.001 25.0 20.8 0.002
ASA 30.7 36.9 ,0.001 48.5 56.7 ,0.001
Clopidogrel 4.8 5.5 NS 13.3 16.8 0.003
AVKs 13.7 9.3 ,0.001 23.3 18.1 ,0.001

ACE-I, angiotensin-converting enzyme-inhibitor; ARB, angiotensin II receptor I blockers; BB, beta-blockers; CCB, calcium channel blockers; ASA, aspirin; AVK, vitamin K
antagonist.

Table 4 Therapeutic modalities given during Table 5 Differences in in-hospital outcome between
hospitalization in acutely decompensated heart failure acutely decompensated heart failure (ADHF) and acute
(ADHF) vs. acute pulmonary oedema (APE) patients pulmonary oedema (APE) patients

ADHF (%) APE (%) P-value ADHF APE P-value

................................................................................ (%) (%)
Diuretics 95.1 97.9 ,0.001 ................................................................................
Diuretics per os 73.9 71.4 0.04 Death 6.0 7.4 0.057
Diuretics i.v. 87.5 94.9 ,0.001 Discharged intensive care unit 8.5 13.5 ,0.05
ACE-I/ARBs 74.5 82.6 ,0.01 Discharged rehabilitation centres or 10.7 12.0 ,0.05
required healthcare facilities
BB 50.0 48.1 NS
Discharged home 70.1 61.9 ,0.05
Nitrates per os/TTS 25.7 29.9 0.002
Nitrates i.v. 37.9 50.3 ,0.001
CCB 0.7 1.6 0.004
Other vasodilators 1.6 2.6 0.021
Analgesics i.v. 23.8 33.2 ,0.001
cardiac contractility reserve, low tissue perfusion, and fluid
Digoxin 34.2 29.9 0.003
accumulation, along with increased arterial stiffness, neurohor-
ASA 52.7 61.5 ,0.001
monal activation, and fluid redistribution, resulting in pulmonary
Clopidogrel 15.4 20.8 ,0.001
congestion.13,14
AVKs 25.6 21.0 ,0.001
CPAP 6.6 10.8 ,0.001 Epidemiology
MV 9.3 15.8 ,0.001 Although APE is common, few data are available about this clinical
PCI 8.8 14.5 ,0.001 entity. Registries such as ADHERE and OPTIMIZE-HF recruited
CABG 2.4 3.4 0.04 few critically ill patients and therefore had low overall in-hospital
Valve surgery 2.9 4.0 0.05 mortality rates (about 4%).1,2 Thus, the percentages of APE and car-
IABP 1.6 3.8 ,0.001 diogenic shock were lower in these registries compared with
Dobutamine 19.3 18.2 NS ALARM-HF, which recruited more severe AHF patients with an
Dopamine 9.8 12.0 0.018 overall in-hospital death rate of 12%.9 The most recent EHFS II
Noradrenaline 2.2 3.1 0.05 reported that 16% of the AHF patients present with APE, which
Levosimendan 7.3 5.8 0.03 is lower than in ALARM-HF. This variation can probably be
explained by differences in the assessment of pulmonary congestion
ACE-I, angiotensin-converting enzyme-inhibitor; ARBs, angiotensin receptor in chest X-rays. As some degree of interstitial congestion is seen in
blockers; BB, beta-blockers; CCB, calcium channel blockers; ASA, aspirin; AVKs, most patients with acute decompensation, categorization will
vitamin K antagonist; MV, venturi mask; PCI, percutaneous coronary intervention;
CABG, coronary artery bypass graft surgery; IABP, intra-aortic balloon pump. depend on the individual judgement of the clinicians. This assump-
tion is in agreement with the finding that the overall incidence of
1198 J.T. Parissis et al.

Table 6 Univariate analysis for in-hospital outcome in Table 6 Continued

acute pulmonary oedema patients
P-value EXP (B) 95.0% CI for
P-value EXP (B) 95.0% CI for EXP (B)
EXP (B) .....................
..................... Lower Upper
Lower Upper bound bound
bound bound ................................................................................
................................................................................ ARB 0.002 0.481 0.304 0.763
Demographic characteristics Nitrate 0.180 0.605 0.291 1.26
Sex (male) 0.105 0.755 0.537 1.060 Digoxin 0.052 1.645 0.996 2.719
Age 0.348 1.034 0.964 1.109 CCB 0.999 0.000 0.000 9.99
BMI 0.467 1.015 0.976 1.056 ASA 0.190 0.776 0.531 1.134
First evaluation CPAP 0.030 1.707 1.053 2.767
Heart rate 0.015 1.009 1.002 1.016 MV 0.000 6.434 4.457 9.287
LVEF 0.000 0.957 0.939 0.976 Treatment at hospitalization
SBP 0.000 0.957 0.939 0.976 PCI 0.107 0.617 0.343 1.111
Oxygen saturation 0.835 0.997 0.973 1.022 CABG 0.780 0.864 0.309 2.416
Sodium 0.438 1.010 0.985 1.035 Valve surgery 0.890 0.937 0.371 2.365
Cr ≥ 1.4 mg/dL 0.013 1.613 1.108 2.410 Adrenaline 0.000 6.362 3.213 12.598
Troponin negative 0.000 0.430 0.293 0.632 Dobutamine 0.000 3.371 2.339 4.859
CKMB negative 0.000 0.487 0.333 0.712 Noradrenaline 0.000 5.057 2.721 9.401
Co-morbidities Levosimendan 0.626 1.192 0.588 2.419
Atrial fibrillation 0.664 1.081 0.760 1.537 Dopamine 0.000 2.398 1.561 3.684
Atrial flutter 0.778 0.943 0.629 1.415
Cardiomyopathy 0.016 1.773 1.111 2.830 BMI, body mass index; LVEF, left ventricular ejection fraction; SBP, systolic blood
PVD 0.008 1.925 1.184 3.131 pressure; CKMB, cardiac isoenzyme of creatinine phosphokinase; PVD, peripheral
vascular disease; CAD, coronary artery disease; CHF, chronic heart failure; ACS,
Heart valve disease 0.75 1.081 0.66 1.757 acute coronary syndrome; CRT, cardiac resynchronization therapy; ICD,
CAD 0.844 1.039 0.712 1.516 implantable cardioverter defibrillator; ICU, intensive care unit; ACE-I,
CHF 0.073 1.389 0.969 1.990 angiotensin-converting enzyme-inhibitor; ARB, angiotensin receptor blockers;
CCB, calcium channel blockers; ASA, aspirin; MV, venturi mask; PCI, percutaneous
Obesity 0.034 0.611 0.387 0.963 coronary intervention; CABG, coronary artery bypass graft surgery.
Diabetes 0.211 0.796 0.557 1.138
Arterial hypertension 0.686 0.922 0.623 1.365
Dyslipidaemia 0.018 0.639 0.441 0.926 ADHF and APE was almost identical in both studies (76% in
Smoking 0.698 0.920 0.605 1.399 ALARM-HF and 81% in EHFS II). In addition, the ALARM-HF data
Anaemia 0.367 1.238 0.778 1.970 were based on the final discharge diagnosis and consequently
Depression 0.160 0.548 0.237 1.268 included events occurring in hospital. Twenty-four per cent of the
Dementia 0.000 2.922 1.646 5.188 patients in ALARM-HF were categorized as ‘secondary’ AHF, and
CRD 0.003 1.778 1.214 2.603 the majority of these deteriorations were due to APE or cardiogenic
Chronic obstructive 0.072 0.657 0.416 1.039 shock in patients hospitalized for acute coronary syndrome. Thus, a
pulmonary disease considerable proportion of patients would not have been included if
No co-morbidities 0.034 0.659 0.449 0.968 only the admission diagnosis had been considered. Furthermore, the
Previous CV event 0.000 3.016 1.811 5.023 relatively high proportion of ICU/CCU hospitalizations (83%) shifts
Precipitating factors the present study between EHFS II and the EFICA study which
Primary/secondary 0.001 1.889 1.308 2.728 included only ICU/CCU admissions, and presented very high per-
ACS 0.064 1.98 0.980 1.992 centages of APE (82%) and cardiogenic shock (29%). Additionally,
Arrhythmia 0.825 0.957 0.650 1.410 the Italian nationwide survey also reported a high percentage of
Compliance 0.098 0.573 0.296 1.109 APE patients (49.6%). Finally, although chest radiography is the cor-
Post-surgery 0.001 2.862 1.530 5.355 nerstone of diagnosis, approximately one in every five patients with
Valvular heart disease 0.32 0.761 0.44 1.304 AHF have no signs of congestion on radiography. Alveolar oedema
ICU stay 0.047 1.754 1.008 3.053 appears to be a very specific (99%) but poorly sensitive (6%) marker
Length of ICU stay 0.001 1.002 1.001 1.003 of heart failure, and clinicians should consider the prevalence of
Treatment at admission negative radiography, when evaluating dyspnoeic patients.15
Diuretics 0.005 0.529 0.338 0.828
BB 0.018 0.541 0.325 0.900 Clinical presentation
ACE-I 0.002 0.509 0.330 0.787 APE presents with an abrupt onset of symptoms (,6 h in about 50%
Continued of individuals), dramatic clinical status, requiring direct admission by
ambulance to the emergency room in half of the cases, whereas a
APE management 1199

the rapidity of symptoms results more often in orthopnoea. Concern-

Table 7 Multivariate analysis of in-hospital outcome ing the clinical examination, APE patients present with higher systolic
in acute pulmonary oedema patients blood pressure along with cold extremities, underlining the mechan-
p-value EXP (B) 95.0% CI for ism of vasoconstriction, increased peripheral resistance, and ventricu-
EXP (B) lar–vascular mismatch, leading to alveolar oedema, clinically
..................... evidenced by orthopnoea, rales, and low oxygen saturation, despite
Lower Upper
bound bound an almost preserved LVEF.
................................................................................ In conjunction with the EHFS II findings, ALARM-HF confirmed the
Demographic characteristics role of cardiovascular co-morbid conditions in APE and showed that
Sex (male) 0.149 0.640 0.349 1.174 the most common precipitating factors were acute coronary syn-
Age 0.302 1.061 0.948 1.189 drome, new onset of arrhythmia, and post-surgery aetiology.
First evaluation Concerning the chest radiography findings, one in five APE
Heart rate 0.595 1.003 0.992 1.014 patients had no signs of alveolar or interstitial oedema, and this
LVEF 0.017 0.972 0.949 0.995 cannot be attributed to the abrupt evolution of the syndrome.
SBP 0.019 0.986 0.974 0.998 The proportion is consistent with recently published findings
Creatinine 0.000 3.064 1.690 5.556 from the ADHERE registry, in which almost 20% of the patients
≥1.4 mg/dL deemed to have heart failure during their inpatient stay had nega-
Troponin 0.396 1.347 0.677 2.680 tive chest radiography for signs of congestion in the emergency
Co-morbidities room.15 Furthermore, two out of five ADHF patients had no
Cardiomyopathy 0.033 2.192 1.063 4.520 signs of congestion on chest radiography, possibly due to the
PVD 0.422 1.438 0.593 3.484 chronic elevation of LV filling pressures, leading to interstitial fibro-
CHF 0.918 0.965 0.489 1.905 sis and prevention of fluid extravasation.
Obesity 0.214 0.618 0.289 1.321
Dyslipidaemia 0.315 0.726 0.388 1.357 Treatment
Chronic obstructive 0.124 0.559 0.266 1.174 As expected, the first-line treatment of APE patients was i.v. diure-
pulmonary disease
tics and vasodilators, compatible with the pathophysiology and
Dementia 0.008 4.020 1.428 11.316
treatment guidelines of this condition.7,16 – 18 Almost all (97.9%)
Previous CV event 0.000 4.966 2.270 10.864
APE patients in ALARM-HF received i.v. diuretics and half
Reason for hospitalization
(50.3%) received i.v. nitrates; respective percentages in EHFS II
ACS 0.945 0.971 0.420 2.245
were 94 and 70.6%.3 The high percentage of i.v. nitrates in EHFS
Compliance 0.186 0.327 0.062 1.717
II is surprising, considering that almost 40% of the population
Post-surgery 0.574 1.443 0.401 5.188
required i.v. inotropes. In ALARM-HF, one-fifth of the patients
Primary/secondary 0.000 3.320 1.797 6.134
required dobutamine, whereas almost 12, 6, and 3% required
Treatment
dopamine, levosimendan, or nor adrenaline, respectively.
Diuretics 0.006 0.369 0.183 0.747
Severe respiratory distress was underlined by the relatively high
BB 0.125 0.509 0.215 1.207
percentage of patients requiring a venturi mask (15.8%) or non-
ACE-I 0.404 0.725 0.340 1.544
invasive ventilation (10.8%). It has already been elucidated that
ARB 0.635 1.184 0.590 2.376
non-invasive ventilation relieves symptomatology and reduces the
Digoxin 0.060 2.293 0.965 5.446
need for intubation, but its impact on mortality is still under
debate.19 – 22 APE patients were also treated more frequently
BMI, body mass index; LVEF, left ventricular ejection fraction; SBP, systolic blood
pressure; CKMB, cardiac isoenzyme of creatinine phosphokinase; PVD, peripheral with interventional procedures and were more likely to receive
vascular disease; CAD, coronary artery disease; CHF, chronic heart failure; ACS, an intra-aortic balloon pump due to the higher incidence of an
acute coronary syndrome; CRT, cardiac resynchronization therapy; ICD,
acute coronary syndrome in this condition.
implantable cardioverter defibrillator; ICU, intensive care unit; ACE-I,
angiotensin-converting-enzyme-inhibitor; ARB, angiotensin receptor blockers; BB, Another important finding from ALARM-HF is that hospitaliz-
beta-blockers; CCB, calcium channel blockers; ASA, aspirin; MV, venturi mask; PCI, ation results in an almost doubling of treatment at discharge com-
percutaneous coronary intervention; CABG, coronary artery bypass graft surgery.
pared with treatment at admission, despite the relative low usage
of beta-blockers at discharge (33% for APE and 39% for ADHF).

quarter of APE cases occur during hospitalization for another reason. Prognosis
Dyspnoea was observed in a similar percentage of patients in both Our data illustrate the clinical profile of an APE patient who is a
groups. Although this may seem an unexpected finding, one should middle-aged male, smoker, hypertensive, diabetic, transferred by
bear in mind that the mechanism of dyspnoea differs between these the paramedics in a sitting position with a venturi mask, without
two groups. CHF patients presenting with decompensation have a a previous history of hospitalization, who was well until the pre-
restrictive pattern of pulmonary disease due to chronic interstitial vious day, and who usually ‘forgets’ to mention his chest discom-
oedema and increased cardiothoracic ratio, pulmonary muscle myo- fort. This patient is an individual from the general population
pathy, and remodelling of the pulmonary vascular bed. As a result, the with a 7.4% risk of death during his hospitalization, despite optim-
symptom of dyspnoea is as common as it is in APE patients, in which ization of AHF management. This study has identified that history
1200 J.T. Parissis et al.

Table 8 Characteristics and in-hospital outcomes of acute pulmonary oedema patients according to systolic blood
pressure (SBP) quartiles

SBP < 100 mmHg, SBP 5 100–120 mmHg, SBP 5 120–160 mmHg, SBP > 160 mmHg, P-value
n (%) n (%) n (%) n (%)
...............................................................................................................................................................................
Age (years)
,50 3.5 2.4 2.2 1.1 ,0.001
51–70 11.4 10.6 17.1 7.6 ,0.001
.71 8.5 8.2 18.5 9.0 ,0.001
Echocardiography
LVEF . 45% 19.8 22.7 24.9 31.8 ,0.005
In-hospital treatment
ACE-I/ARBs 71.1 80.5 84.2 86.3 ,0.0001
Beta-blockers 44.5 55.4 48.0 47.0 ,0.05
i.v. nitrates 20.5 26.3 33.8 34.9 ,0.001
i.v. diuretics 96.8 98.8 98.7 97.2 NS
Dobutamine 34.7 20.8 14.2 13.3 ,0.001
Dopamine 24.0 14.7 6.5 7.7 ,0.001
Adrenaline 5.3 3.4 1.0 0.8 ,0.001
Noradrenaline 6.4 3.7 2.3 1.2 ,0.001
In-hospital outcome
Death 14.1 8.0 5.4 3.4 ,0.001

LVEF, left ventricular ejection fraction; ACE-I, angiotensin-converting enzyme-inhibitor; ARBs, angiotensin receptor blockers.

Figure 2 Kaplan– Meier survival curves for systolic blood pressure quartiles in acute pulmonary oedema patients. Patients with a systolic
blood pressure ≤100 mmHg had higher in-hospital mortality (survival rate 79%) compared with those with systolic blood pressure from
101 to 119 mmHg (survival rate 86%), systolic blood pressure from 120 to 159 mmHg (survival rate 90%), or systolic blood pressure
≥160 mmHg (survival rate 99.8%; log rank ¼ 37.8, P , 0.001).

of previous cardiovascular event, cardiomyopathy, LVEF, systolic Few data are available concerning the mortality risk of patients with
blood pressure, serum creatinine at presentation, and treatment APE.23 – 31 The ALARM-HF survey supports the strong prognostic
with diuretics are independent prognostic factors associated with significance of LVEF and systolic blood pressure at admission, which
outcome in APE patients. has previously been proposed by small retrospective trials.25 – 29
APE management 1201

Figure 3 Kaplan– Meier survival curves for creatinine levels in acute pulmonary oedema patients (receiver operating characteristic derived
best cut-off level 1.4 mg/dL). Patients with a serum creatinine ≥1.4 mg/dL had an increased mortality (survival rate 85%) compared with those
with serum creatinine ,1.4 mg/dL (survival rate 92%; log rank ¼ 4.3, P ¼ 0.03).

The present survey also demonstrates the linear relationship between

these parameters. Additionally, impaired renal function as estimated Table 9 Differences and similarities of acute
by a serum creatinine ≥1.4 mg/dL is another parameter associated pulmonary oedema (APE) and acutely decompensated
with higher in-hospital mortality. Concerning patient history, a chronic heart failure (ADHF)
previous cardiovascular event is associated with a five-fold risk
APE ADHF
of in-hospital death, known cardiomyopathy with a two-fold ................................................................................
risk, whereas a secondary diagnosis of APE is associated with a Onset of symptoms Rapid Over weeks
four-fold risk. Pathophysiological Acute event in usually Volume overload
The ALARM-HF survey also showed that treatment with a diure- mechanism euvolaemic
tic on admission reduces the risk of in-hospital death, although a patients
trend for increased risk was observed for chronic digoxin use in Left ventricular filling Increased Increased
pressures
APE patients, possibly due to pro-arrhythmic actions.
Left ventricular Preserved Reduced
ejection fraction
Limitations Treatment Vasodilators Vasodilators or
Data collection in the participating centres of ALARM-HF was non-invasive inotropes
limited to five to eight consecutive admissions with a clinical diag- ventilation diuretics
nosis of AHF, so the cases included may not be representative of
the overall AHF population in all countries. Additionally, the differ-
ential diagnosis between AHF syndromes was made retrospec-
Based on epidemiological data from the main AHF registries as
tively at discharge. The final diagnosis of AHF, although made at
well as the ALARM-HF survey, Table 9 summarizes the differences
discharge and consequently a ‘true positive’, was not confirmed
and similarities in the clinical profile and the management of these
centrally. The overlapping of different AHF syndromes may also
two clinical entities, which may help clinicians to diagnose and treat
provoke extra difficulties, underlining the need for a more clear
each condition more effectively in routine clinical practice.
classification. Finally, as ALARM-HF was designed to collect anon-
ymized data, the lack of long-term outcome data is inevitable.
Conflict of interest: J.T.P., A.M., J.D., F.V.-B., and F.F. report being
consultants for and receiving honoraria from Abbott USA.
Conclusion
Our data suggest that a patient with APE can be stratified very
early in his clinical course, even from the time of diagnosis and sub- Funding
jected to intense monitoring according to the presence of the Abbott USA funded the ALARM-HF survey and data were acquired by
prognostic factors mentioned earlier. IMS, UK.
1202 J.T. Parissis et al.

References 14. Cotter G, Metra M, Milo-Cotter O, Dittrich HC, Gheorghiade M. Fluid overload
in acute heart failure—redistribution and other mechanisms beyond fluid accumu-
1. Adams KF Jr, Fonarow GC, Emerman CL, LeJemtel TH, Costanzo MR,
lation. Eur J Heart Fail 2008;10:165 –169.
Abraham WT, Berkowitz RL, Galvao M, Horton DP; ADHERE Scientific Advisory
15. Collins SP, Lindsell CJ, Storrow AB, Abraham WT. Prevalence of negative chest
Committee and Investigators. Characteristics and outcomes of patients hospital-
radiography results in the emergency department patient with decompensated
ized for heart failure in the United States: rationale, design, and preliminary obser-
heart failure. Ann Emerg Med 2006;47:13 –18.
vations from the first 100,000 cases in the Acute Decompensated Heart Failure
16. Shin D, Brandimarte F, de Luca L, Sabbah H, Fonarow G, Filippatos G, Komajda M,
National Registry (ADHERE). Am Heart J 2005;149:209 –216.
Gheorghiade M. Review of current and investigational pharmacologic agents for
2. Fonarow GC, Abraham WT, Albert NM, Gattis WA, Gheorghiade M,
acute heart failure syndromes. Am J Cardiol 2007;99(suppl.):4A –23A.
Greenberg B, O’Connor CM, Yancy CW, Young J. Organized Program to Initiate
Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF): 17. Kirk D, Parissis JT, Filippatos G. Pharmacologic stabilization and management of
rationale and design. Am Heart J 2004;148:43 –51. acute heart failure syndromes in the emergency department. Heart Fail Clin
3. Nieminen MS, Harjiola VP, Hochadel M, Drexler H, Komajda M, Brutsaert D, 2009;5:43– 54.
Dickstein K, Ponikowski P, Tavazzi L, Follath F, Lopez-Sendon JL. Gender 18. De Luca L, Fonarow G, Adams K, Mebazaa A, Tavazzi L, Swedberg K,
related differences in patients presenting with acute heart failure. Results from Gheorghiade M. Acute heart failure syndromes: clinical scenarios and pathophy-
Euro Heart Failure Survey II. Eur J Heart Fail 2008;10:140 –148. siologic targets for therapy. Heart Fail Rev 2007;12:97–104.
4. Zannad F, Mebazaa A, Juillière Y, Cohen-Solal A, Guize L, Alla F, Rougé P, Blin P, 19. Peter JV, Moran JL, Phillips-Hughes J, Graham P, Bersten AD. Effect of non-
Barlet MH, Paolozzi L, Vincent C, Desnos M, Samii K, for the EFICA Investigators. invasive positive pressure ventilation (NIPPV) on mortality in patients with
Clinical profile, contemporary management and one-year mortality in patients acute cardiogenic pulmonary oedema: a meta-analysis. Lancet 2006;367:
with severe acute heart failure syndromes: the EFICA study. Eur J Heart Fail 1155 –1163.
2006;8:697 –705. 20. Masip J, Roque M, Sanchez B, Fernandez R, Subirana M, Exposito J. Non-invasive
5. Nicol E, Fittall B, Roughton M, Cleland JK, Dargie H, Cowie MR. National Health ventilation in acute cardiogenic pulmonary oedema systematic review and
Service Heart Failure Survey: a survey of acute heart failure admissions in England, meta-analysis. JAMA 2005;294:3124 –3130.
Wales and Northern Ireland. Heart 2008;94:172–177. 21. Gray A, Goodacre S, Newby D, Masson M, Sampson F, Nicholl J, for the 3CPO
6. Tsutsui H, Tsuchihashi-Makaya M, Kinugawa S, Goto D, Takeshita A, the JCARE- Trialists. Non-invasive ventilation in acute cardiogenic pulmonary oedema. N Engl
CARD Investigators. Clinical characteristics and outcome of hospitalized patients J Med 2008;359:142 –151.
with heart failure in Japan. Circ J 2006;70:1617 – 1623. 22. Masip J, Mebazaa A, Filippatos GS. Noninvasive ventilation in acute cardiogenic
7. Nieminen MS, Bohm M, Cowie MR, Drexler H, Filippatos GS, Jondeau G, Hasin Y, pulmonary edema. N Engl J Med 2008;359:2068 –2069.
Lopez-Sendon J, Mebazaa A, Metra M, Rhodes A, Swedberg K, Priori SG, 23. Collins S, Storrow AB, Kirk JD, Pang PS, Diercks DB, Gheorghiade M. Beyond pul-
Garcia MA, Blanc JJ, Budaj A, Cowie MR, Dean V, Deckers J, Burgos EF, monary oedema: diagnostic, risk stratification, and treatment challenges of acute
Lekakis J, Lindahl B, Mazzotta G, Morais J, Oto A, Smiseth OA, Garcia MA, heart failure management in the emergency department. Ann Emerg Med 2008;51:
Dichstein K, Albuquerque A, Conthe P, Crespo-Leiro M, Ferrari R, Follath F, 45 –57.
Gavazzi A, Janssens U, Komajda M, Morais J, Moreno R, Singer M, Singh S, 24. Fiutowski M, Waszyrowski T, Krzeminska-Pakula M, Kasprzak J. Pulmonary
Tendera M, Thegesen K. Executive summary of the guidelines on the diagnosis oedema prognostic score predicts in-hospital mortality risk in patients with
and treatment of acute heart failure: the Task Force on Acute Heart Failure of acute cardiogenic pulmonary oedema. Heart Lung 2008;37:46– 53.
the European Society of Cardiology. Eur Heart J 2005;26:384 –416. 25. Di Marco F, Tresoldi S, Maggiolini S, Bozzano A, Bellani G, Pesenti A, Fumagalli R.
8. Dickstein K, Cohen-Solal A, Filippatos G, McMurray JJ, Ponikowski P,
Risk factors for treatment failure in patients with severe acute cardiogenic pul-
Poole-Wilson PA, Stromberg A, van Veldhuisen DJ, Atar D, Hoes AW,
monary oedema. Anaesth Intensive Care 2008;36:351 –359.
Keren A, Mebazaa A, Nieminen M, Priori SG, Swedberg K. ESC guidelines for
26. Edoute Y, Roguin A, Behar D, Reisner S. A prospective evaluation of pulmonary
the diagnosis and treatment of acute and chronic heart failure 2008: the Task
oedema. Crit Care Med 2000;28:330 –335.
Force for the diagnosis and treatment of acute and chronic heart failure 2008
27. Nunez J, Nunez E, Fonarow G, Sanchis J, Bodi V, Bertomeu-Gonzalez V,
of the European Society of Cardiology. Developed in collaboration with the
Minana G, Merlos P, Bertoeu-Martinez V, Redon J, Chorro F, Liacer A. Differential
Heart Failure Association of the ESC (HFA) and endorsed by the European
prognostic effect of systolic blood pressure on mortality according to left-
Society of Intensive Care Medicine (ESICM). Eur J Heart Fail 2008;10:933 –989.
ventricular function in patients with acute heart failure. Eur J Heart Fail 2010;12:
9. Follath F, Delgado JF, Mebazaa A, Parissis JT, Vilas Boas F, Smith H, Mclean A. Clas-
sifying patients with acute heart failure: presentation, treatment and outcome. A 38 –44.
nine country survey of acute heart failure management (ALARM-HF). Eur J Heart 28. Figueras J, Pena C, Soler-Soler J. Thirty day prognosis of patients with acute pul-
Fail 2008;7(suppl.):64. Abstract 263. monary oedema complicating acute coronary syndromes. Heart 2005;91:
10. Ezekowitz JA, Bakal J, Kaul P, Westerhout C, Armstrong P. Acute heart failure in 889 –893.
the emergency department: short and long-term outcomes of elderly patients 29. Shamagian L, Roman A, Ramos P, Acuña J, Veloso P, Gonzalez-Juanatey J. Acute
with heart failure. Eur J Heart Fail 2008;10:308 –314. pulmonary oedema in patients with decompensated heart failure. Role of under-
11. Nieminen MS, Harjola V. Definition and epidemiology of acute heart failure syn- lying cardiomyopathy on the prognosis. Int J Cardiol 2007;121:302 –305.
dromes. Am J Cardiol 2005;96(suppl.):G5 –G10. 30. Le Conte P, Coutant V, N’Guyen JM, Baron D, Touze MD, Potel G. Prognostic
12. Dar O, Cowie M. Acute heart failure in the intensive care unit: epidemiology. Crit factors in acute cardiogenic pulmonary oedema. Am J Emerg Med 1999;17:
Care Med 2008;36(suppl.):S3–S8. 329 –332.
13. Cotter G, Felker M, Adams KF, Milo-Cotter O, O’Connor CM. The pathophysiol- 31. Ilva T, Lassus J, Siirila-Waris K, Melin J, Peuhkurinen K, Pulkki K, Nieminen MS,
ogy of acute heart failure—is it all about fluid accumulation? Am Heart J 2008;155: Mustonen H, Porela P, Harjiola VP. Clinical significance of cardiac troponins I
9 –15. and T in acute heart failure. Eur J Heart Fail 2008;10:772–778.