Original Contribution

A comparison of bilevel and continuous positive airway pressure noninvasiveventilation in acute cardiogenic pulmonary edema!

Hui Li MM, Chunlin Hu MD, Jinming Xia MM, Xin Li MD, Hongyan Wei MM,Xiaoyun Zeng MM, Xiaoli Jing MD!Department of Emergency, The First Af!liated Hospital of Sun Yat-Sen University, Guangzhou 510080, China

a b s t r a c ta r t i c l e i n f o

Article history:Received 27 April 2013Received in revised form 26 May 2013Accepted 28 May 2013

Background: Whether bilevel positive airway pressure (BiPAP) is advantageous compared with continuouspositive airway pressure (CPAP) in acute cardiogenic pulmonary edema (ACPO) remains uncertain. The aim ofthe meta-analysis was to assess potential bene!cial and adverse effects of CPAP compared with BiPAP inpatients with ACPO.Methods: Randomized controlled trials comparing the treatment effects of BiPAP with CPAP wereidenti!ed from electronic databases and reference lists from January 1966 to December 2012. Tworeviewers independently assessed study quality. In trials that ful!lled inclusion criteria, we criticallyevaluate the evidence for the use of noninvasive ventilation on rates of hospital mortality, endotrachealintubation, myocardial infarction, and the length of hospital stay. Data were combined using ReviewManager 4.3 (The Cochrane Collaboration, Oxford, UK). Both pooled effects and 95% con!dence intervals(CIs) were calculated.Results: Twelve randomized controlled trials with a total of 1433 patients with ACPO were included. Thehospital mortality (relative risk [RR], 0.86; 95% CI, 0.65-1.14; P = .46; I2 = 0%) and need for requiringinvasive ventilation (RR, 0.89; 95% CI, 0.57-1.38; P = .64; I2 = 0%) were not signi!cantly differentbetween patients treated with CPAP and those treated with BiPAP. The occurrence of new cases ofmyocardial infarction (RR, 0.95; 95% CI, 0.77-1.17; P = .53, I2 = 0%) and length of hospital stay(RR, 1.01; 95% CI, !0.40 to 2.41; P = .98; I2 = 0%) were also not signi!cantly different between the2 groups.Conclusions: There are no signi!cant differences in clinical outcomes when comparing CPAP vs BiPAP.Based on the limited data available, our results suggest that there are no signi!cant differences in clinicaloutcomes when comparing CPAP with BiPAP.

© 2013 Elsevier Inc. All rights reserved.

1. Background

Acute cardiogenic pulmonary edema (ACPO) is one of the mostimportant medical emergencies. Conventional treatment with oxy-gen, diuretics, and vasodilators could improve the symptoms of mostof the patients with ACPO. However, assisted ventilation may beneeded in patients still presenting hypoxemia and respiratory distressafter the initial conventional medical therapy, with the associatedpotential for complications [1]. Continuous positive airway pressure(CPAP) and bilevel positive airway pressure (BiPAP) are widely usedas an adjunct to treatment in acute respiratory failure. Over the past10 years, application of noninvasive ventilation (NIV) has been

suggested in association with the conventional medical treatment asan effective therapeutic modality in ACPO. Acute pulmonary edema isconsidered the second most common indication for NIV in clinicalpractice [2,3]. The physiological effects of NIV include augmentation ofcardiac output and oxygen delivery, improved functional residualcapacity and respiratory mechanics, reduced effort in breathing, anddecreased left ventricular afterload [4–7]. As compared with CPAP,some studies have shown that BiPAP reduce the work of breathingand improve respiratory distress more effectively. Meanwhile, 2studies suggested that patients treated with BiPAP had a higherincidence of acute myocardial infarction (AMI) and a trend toward anincrease in creatine kinase [8,9]. A few recent meta-analyses could noteliminate all the doubts at this regard because of the paucity of dataavailable and of the presence of confounding factors [10-13].Considerable controversy remains over the short-term effect of NIVon ACPO. In light of this uncertainty, we conducted systematic reviewand meta-analysis, including several new randomized controlledtrials (RCTs), to investigate whether NIV improves survival andwhether BiPAP is more advantageous than CPAP.

American Journal of Emergency Medicine 31 (2013) 1322–1327

! This studywas supported by a grant from the TIANPU Fund (01200918). The fundershad no role in the study design, data collection, analysis, decision to publish, or thepreparation of the manuscript.! Correspondence author. Tel.: +86 20 87755766-8500; fax: +86 20 87755766

8773.E-mail address: totohui1984@163.com (X. Jing).

0735-6757/$ – see front matter © 2013 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.ajem.2013.05.043

Contents lists available at ScienceDirect

American Journal of Emergency Medicine

j ourna l homepage: www.e lsev ie r .com/ locate /a jem

2. Materials and methods

2.1. Search strategy and quality assessment

A literature search was conducted using EMBASE and MEDLINEdatabases (1966–December 2012). Only randomized controlledclinical trials comparing BiPAP with CPAP in adult patients withACPOwere included. During the electronic search for publications, thekey words used were as follows: “congestive heart failure,” “heartfailure,” “pulmonary edema,” “cardiogenic pulmonary edema,” “lungedema,” “positive pressure ventilation,” “noninvasive ventilation,”“noninvasive positive pressure ventilation,” “NPPV,” “intermittentpositive-pressure ventilation,” “positive pressure respiration,” “inter-mittent positive-pressure breathing,” “BiPAP,” and “CPAP.” Thesearches were limited to English language publications. Referenceswere examined to identify additional articles. Authors were contactedto clarify details of trials if necessary. Studies using both CPAP andBiPAP in the same group of patients in a crossover design wereexcluded. In reporting the results of this systematic review, the

authors have followed the recommended guidelines from the Qualityof Reporting of Meta-Analysis statement.

The selected articles were reviewed for the following studycharacteristics: (1) studies were prospective randomized trials, (2)adult patients undergoing ACPO, (3) documented inclusion andexclusion criteria and comparability of groups at baseline, (4) bilevelnoninvasive positive pressure ventilation was compared with CPAP,(5) documented criteria for invasive ventilation, and (6) a follow-upperiod lasting at least the duration of hospitalization. Data selectionand data extracted were performed among pairs of independentreviewers, and the results were con!rmed by a third reviewer.Discrepancies were discussed and adjudicated by the team consensus.

The quality assessment of the studies included based on 4 mainfactors, including description of randomization, allocation conceal-ment, description of withdrawals, and blinding outcome assessment[14-18]. The primary outcomes for the included trials were thehospital mortality and the proportion of patients requiring invasiveventilation (or intubation). The other outcomes assessed in the meta-analysis included the length of hospital stay and incidence of newmyocardial infarction.

2.2. Statistical analyses

Data were checked and entered into the Review Manager (version4.3; The Cochrane Collaboration, Oxford, UK) database for furtheranalysis. Using a !xed-effects model, we calculated the risk ratios(RRs) and 95% con!dence intervals for hospital mortality, need forintubation, and incidence of new AMI and weighted mean differencesfor the length of hospital stay. For continuous variables (length ofhospital stay), pooled estimates were only available if the mean andSD were reported. The presence of heterogeneity between trials wasassessed using the !2 statistics, and the extent of inconsistency wasassessed using I2 statistics.

3. Results

Twelve 12 articles with a total of 1433 patients ful!lled the criteriafor inclusion (Fig. 1) and compared CPAP with BiPAP, including 11studies that were conducted in the emergency departments [7-9,14–22]. Trial characteristics are summarized in Table 1. All of the articleswere published in English, including data from 7 countries. Six studieswere multiple-center trials, whereas the others were conducted in 1single center. The CPAP level used in these trials ranged from 5 to 15cm H2O, although the most frequent pressure was 10 cm H2O. Thelevel of BiPAP was variable. Average inspiratory pressure airway

2316 potentially relevant trials

Limit to RCT (n = 127)

Limit to English (n = 118)

Limit to Adult (n = 97)

Potentially appropriate trials

for meta-analysis (n = 14)

Trials Excluded:

Not heart failure (n = 25)

Not between CPAP and

BiPAP(n = 50)

Other: (n = 8)

Trials included in meta analysis (n = 12)

Trials Excluded:

Part of larger study (n = 1)

Crossover design (n = 1)

Fig. 1. Flow diagram of selection process to obtain articles chosen for meta-analysis.

Table 1Study characteristics, quality scores, and primary outcome data

Year of study publication Country Quality score Mortalitya Intubationb Incidence of AMIc Length of hospital stayd

Mehta et al [8] 1997 US 1 (13)/1 (14) 1 (13)/1 (14) 4 (13)/10 (14) N/APark et al [7] 2001 Brazil 1 (9)/0 (7) 3 (9)/1 (7) 1 (9)/0 (7) N/ACross et al [21] 2003 Australia 5 (36)/3 (35) 9 (36)/3 (35) N/A N/ACrane et al [9] 2004 UK 0 (20)/5 (20) N/A 6 (20)/9 (20) N/APark et al [20] 2004 Brazil 1 (27)/2 (27) 2 (27)/2 (27) N/A 11 ± 8/10 ± 7Bellone et al [19] 2004 Italy 2 (22)/0 (24) 1 (22)/2 (24) 3 (22)/2 (24) N/ABellone et al [14] 2005 Italy 1 (18)/0 (18) 1 (18)/2 (18) N/A N/AMoritz et al [15] 2007 France 4 (50)/8 (50) 2 (50)/1 (50) 3 (50)/2 (50) N/AFerrari et al [18] 2007 Italy 2 (27)/3 (25) 0 (27)/1 (25) 6 (27)/4 (25) 12.9±23.7/9.9±7.4Gray et al [17] 2008 UK 53 (346)/54 (356) 8 (346)/12 (356) 94 (346)/95 (356) N/AFerrari et al [16] 2010 Italy 2 (40)/7 (40) 0 (40)/3 (40) 0 (40)/2 (40) 10.5±15.41/10.06±8.13Nouira et a [22] 2011 Tunisia 3 (101)/5 (99) 7 (101)/10 (99) 2 (101)/4 (99) 11± 4/9± 3

N/A, not available. All data are expressed as CPAP/BiPAP.a Data are number of patients who died (total).b Data are number of patients receiving ventilation (total).c Data are number of new cases of AMI (total).d Data are mean length of hospital stay as mean ± SD (in days).

1323H. Li et al. / American Journal of Emergency Medicine 31 (2013) 1322–1327

pressure ranged from 8 to 20 cm H2O, with 15 cm H2O being the mostrepeated value. Conversely, expiratory pressure airway pressure wasset at 5 cm H2O in most trials. All studies had points deducted forblinding because of the inherent dif!culty of blinding the treatingphysicians to whether the patient is receiving CPAP or BiPAP.

Twelve studies compared the risk of hospital mortality betweenpatients treated with CPAP compared with BiPAP (Fig. 2).There wereno differences in hospital mortality identi!ed between the groups. Apooled analysis of 1433 patients con!rmed no differences betweengroups (RR, 0.86 [0.65-1.14]). The !2 test of heterogeneity (!2 =10.79) was not signi!cant (P = .46).

Eleven studies (n = 1385) assessed the risk of intubation inpatients treated with CPAP to BiPAP (Fig. 3). There was no statisticaldifference between groups in either the individual studies or thepooled analysis (RR, 0.89 [0.57-1.38]). The !2 test for heterogeneity(!2 = 7.39) was not signi!cant (P = .64).

Nine studies compared CPAP with BiPAP for the occurrence of newcases of myocardial infarction (Fig. 4). One of the studies demon-strated that higher rate of myocardial infarctions was associated withthe use of BiPAP [8]. The remainder of the studies showed littleevidence for a favorable trend between CPAP and BiPAP. The pooledeffect of 1272 patients suggested that there was an insigni!cant trendtoward an increase in new-onset AMI in patients treated with BiPAPwhen compared with CPAP (RR, 0.95 [0.77-1.17]). The !2 test forheterogeneity (!2 = 7.03) was not signi!cant (P = .53).

Data on patients with length of hospital stay after initiation ofCPAP or BiPAP were limited (Fig. 5). Based on the pooled analysis ofthe 4 trials (n = 386), there was an insigni!cant trend toward anincrease in length of hospital stay in patients between the 2 assistedventilatory modes (RR, 1.01 [!0.40 to 2.41]). The !2 test forheterogeneity (!2 = 0.22) was not signi!cant (P = .98).

Meanwhile, the funnel plot indicated that the publication bias ofthe review was acceptable (Fig. 6). The heterogeneity of all studiesincludedwas not signi!cant (P N .10 in all heterogeneity assessments).All studies achieved high-quality standard based on the qualityassessment (Table 2). Therefore, the evidence provided by the meta-analysis was convincing.

4. Discussion

This systematic review and meta-analysis investigates whetherthere was a difference between CPAP and BiPAP with respect toendotracheal intubation rate, mortality, incidence of AMI, and hospitalstay. Noninvasive ventilationwas recommended in treatment of acuteheart failure by the European Society of Cardiology and AmericanHeart Association [23,24]. It would be advantageous to identify whichNIV strategy offers the most optimal therapeutic advantage. However,one large randomized trial suggests no mortality bene!t associatedwith NIV. The results of our systematic review are consistent with this

Fig. 2. Effect of BiPAP and CPAP on mortality.

Fig. 3. Effect of BiPAP and CPAP on intubation.

1324 H. Li et al. / American Journal of Emergency Medicine 31 (2013) 1322–1327

RCT, which shows no bene!t of NIV between ventilatory modes in themanagement of patients presenting with ACPO.

In this meta-analysis, we did not compare a group of standardoxygen therapywith CPAP or BiPAP, because it had been reported thatNIV delivered by either CPAP or BiPAP safely provides earlierimprovement and resolution of dyspnea, respiratory distress, andmetabolic abnormalities than standard oxygen therapy [9,17]. Thesigni!cant bene!ts of both CPAP and BiPAP are the reduction in theneed for intubation and a decrease in mortality and length of hospitalstay in ACPO, as seen in previous systematic review andmeta-analysis[25-27].

Themain pathophysiologic mechanism of ventilatory pump failurein patients with pulmonary edema is related to an increase ininspiratory muscle energy demands [28]. Work of breathing may beheld responsible for increasing energy demands and elastic load(alveolar congestion and, consequently, decreased lung compliance).It is well known that administration of CPAP in patients with severedecompensated heart failure reduced left ventricular !lling pressureand mitral regurgitation and increased left ventricular ejectionfraction [29-32]. Also, within 2 hours of administration, CPAPimproves systolic function, myocardial energy consumption, andrespiratory muscles workload in stable patients with heart failure[32,33]. After a few weeks of administration, CPAP improves leftventricular ejection fraction, left ventricular transmural pressure, andleft ventricular hypertrophy and cardiac work index [14]. In thecomparison of NIV modalities, BiPAP has the potential advantage overCPAP of assisting the respiratory muscles during inspiration, whichwould result in faster alleviation of dyspnea and exhaustion [34].Bilevel positive airway pressure was more effective at unloading therespiratory muscles than CPAP, while avoiding unnecessary airwaypressure and an excessive reduction in cardiac output [35]. Despitethe theoretical additional bene!ts of BiPAP as compared with CPAP,we observed no differences in therapeutic ef!cacy between the 2noninvasive treatment methods. One study demonstrated that therewas no difference between the 2 ventilatory modes in a patient

diagnosed as having ACPOwith hypercapnia [10]. The results suggeststhat 15/5 cm H2O in the mode of BiPAP may equally have effects onimproving respiratory and hemodynamics with 10 cm H2O pressurein CPAP, perhaps helping to explain our negative results on intubationrate and mortality.

Although a preliminary study described a higher rate of AMI withBiPAP and prematurely terminated their trial comparing CPAP withBiPAP, Moritz et al [15] found a higher proportion of myocardialinfarction in patients who did not respond to treatment andtherefore did not recommend BiPAP in patients with myocardialinfarction. Nevertheless, Ferrari et al [16] described a good responseto nasal CPAP in patients with ACPO secondary to myocardialinfarction. No other trial !ndings showed the same result betweenboth techniques. Several studies considered that the potentialadvantage of BiPAP on faster alleviation of dyspnea and exhaustionover CPAP was offset by the tendency toward an increased risk ofnew myocardial infarction [18].

However, patients recruited in these studies were, by nature, athigh risk for developing AMI, either before or during the early phase ofhospitalization. Whether the infarcts preceded or were a consequenceof therapy should be identi!ed. One possible explanation for thehigher rate of AMI results in the study of Mehta et al [8] is that theetiologies of acute pulmonary edema are unbalanced at entrance. Forsome studies, patients with AMI, presented increased concentrationsof troponin I or necessitating thrombolysis or primary angioplasty,were excluded from the study [9,19,36,37]. Also, they had similarrates of treatment-related myocardial infarction in patients receivingCPAP and BiPAP. In our meta-analysis, BiPAP was proved a safemodality for treating patients with ACPO in comparison with CPAP.

The other causes for the similar result in the incidence of AMIbetween BiPAP and CPAP may be the use of nitrate. It was consideredthat nitrates had con!rmed ef!cacy in patients with ACPO [36]. Onestudy found that nitrate dose and use of prehospital nitrates wereassociated with survival to discharge [9]. Furthermore, some in-vestigators have reported that high-dose intravenous isosorbide-

Fig. 4. Effect of BiPAP and CPAP on AMI.

Fig. 5. Effect of BiPAP and CPAP on hospital stay.

1325H. Li et al. / American Journal of Emergency Medicine 31 (2013) 1322–1327

dinitrate administration was more effective than BiPAP ventilation incontrolling pulmonary edema [37]. In our meta-analysis, the dose ofnitrate in ACPO was not reported in half of the pooled studies,whereas in the other half of studies, patients received remarkablydifferent amounts of nitrate treatment. It is hard to relate the use ofnitrate combined with BiPAP or CPAP to a lower mortality.

5. Limitations

In this meta-analysis, although the inclusion criteria, interventionprotocols, and baseline characteristics were similar between the 2groups, the inability to adjust statistically for differences in otherfactors beyond treatment group assignment could have in"uenced the!nal results. First, meta-analyses are prone to bias. The quality of trialscan affect the direction and magnitude of treatment effect in meta-analyses. A second limitation is that the treatment was not blinded.Because the physicians who decided when to initiate endotrachealintubation or to cease NIV were not blinded, introducing thepossibility of investigator bias might have affected the results. FutureNIV studies should consider blinding the physician to the mode of NIVused. Most of these studies were powered for intubation and not formortality, AMI, and length of hospital stay. Most of the includedstudies have small numbers of participants, thereby reinforcing thenecessity of our meta-analysis.

6. Conclusion

According to all available data, there is no evidence to suggestsuperiority of either CPAP or BiPAP in patients with acute cardiogenic

pulmonal edema. Our data can be added to the accumulating evidenceunderscoring the same effect of BiPAP compared with CPAP onmortality, intubation rate, and length of stay in hospital in ACPO.Importantly, no true difference or trend toward an increase in AMIrates was found with the use of BiPAP.

References

[1] British Thoracic Society Standards of Care Committee. Non-invasive ventilation inacute respiratory failure. Thorax 2002;57:192–211.

[2] Antonelli M, Conti G, Moro ML, Esquinas A, Gonzalez-Diaz G, Confalonieri M, et al.Predictors of failure of noninvasive positive ventilation in patients with acutehypoxemic respiratory failure. Intensive Care Med 2001;27:1718–28.

[3] Burns KE, Sinuff T, Adhikari NK, Meade MO, Heels-Ansdell D, Martin CM, et al.Bilevel noninvasive positive pressure ventilation for acute respiratory failure. CritCare Med 2005;33:1477–83.

[4] Baratz DM, Westbrook PR, Shah PK, Mohsenifar Z. Effect of nasal continuouspositive airway pressure on cardiac output and oxygen delivery in patients withcongestive heart failure. Chest 1992;102:1397–401.

[5] Lenique F, Habis M, Lofaso F, Dubois-Randé JL, Harf A, Brochard L. Ventilatory andhemodynamic effects of continuous positive airway pressure in left heart failure.Am J Respir Crit Care Med 1997;155:500–5.

[6] International Consensus Conference in Intensive Care Medicine. Noninvansivepositive pressure ventilation in acute respiratory failure. Am J Respir Crit CareMed2001;163:283–91.

[7] Park M, Lorenzi-Filho G, FeltrimMI, Viecili PR, Sangean MC, Volpe M, et al. Oxygentherapy, continuous positive airway pressure, or noninvasive bilevel positivepressure ventilation in the treatment of acute cardiogenic pulmonary edema. ArqBras Cardiol 2001;76:221–30.

[8] Mehta S, Jay GD, Woolard RH, Hipona RA, Connolly EM, Cimini DM, et al.Randomized, prospective trial of bilevel versus continuous positive airwaypressure in acute pulmonary edema. Crit Care Med 1997;25:620–8.

[9] Crane SD, Elliott MW, Gilligan P, Richards K, Gray AJ. Randomized controlledcomparison of continuous positive airway pressure, bilevel non-invasiveventilation and standard treatment in emergency department patients withacute cardiogenic pulmonary oedema. Emerg Med J 2004;21:155–61.

Fig. 6. Funnel plot for publication bias.

Table 2Main quality assessment of included studies

Year Author Study design Description of randomization Allocation concealment Description of withdrawals Blinding outcome assessment

2007 Ferrari et al [18] RCT + + + !2008 Gray et al [17] RCT + + + !2004 Bellone et al [19] RCT + + + !2005 Bellone et al [14] RCT + + + !2001 Park et al [7] RCT ! ! ! !2004 Crane et al [9] RCT + + + !1997 Mehta et al [8] RCT + + + !2004 Park et al [20] RCT + + + !2003 Cross et al [21] RCT ! ! ! !2010 Ferrari et al [16] RCT + + + !2011 Nouira et a [22] RCT + + + !2007 Moritz et al [15] RCT + + + !

1326 H. Li et al. / American Journal of Emergency Medicine 31 (2013) 1322–1327

[10] Peter JV, Moran JL, Phillips-Hughes J, Graham P, Bersten AD. Effect of non-invasivepositive pressure ventilation (NIPPV) on mortality in patients with acutecardiogenic pulmonary oedema: a meta-analysis. Lancet 2006;367:1155–63.

[11] Ho KM,Wong K. A comparison of continuous and bi-level positive airway pressurenon-invasive ventilation in patients with acute cardiogenic pulmonary oedema: ameta-analysis. Crit Care 2006;10:R49.

[12] Masip J, Roque M, Sánchez B, Fernández R, Subirana M, Expósito JA. Noninvasiveventilation in acute cardiogenic pulmonary edema. JAMA 2005;294:3124–30.

[13] Collins SP, Mielniczuk LM, Whittingham HA, Boseley ME, Schramm DR, StorrowAB. The use of noninvasive ventilation in emergency department patients withacute cardiogenic pulmonary edema: a systematic review. Ann Emerg Med2006;48:260–9.

[14] Bellone A, Vettorello M, Monari A, Cortellaro F, Coen D. Noninvasive pressuresupport ventilation vs continuous positive airway pressure in acute hypercapnicpulmonary edema. Intensive Care Med 2005;31:807–11.

[15] Moritz F, Brousse B, Gellée B, Chajara A, L'Her E, Hellot MF, et al. Continuouspositive airway pressure versus bilevel noninvasive ventilation in acutecardiogenic pulmonary edema: a randomized multicenter trial. Ann Emerg Med2007;50:666–75.

[16] Ferrari G, Milan A, Groff P, Pagnozzi F, Mazzone M, Molino P, et al. Continuouspositive airway pressure vs. pressure support ventilation in acute cardiogenicpulmonary edema: a randomized trial. J Emerg Med 2010;39(5):676–84.

[17] Gray A, Goodacre S, Newby DE, Masson M, Sampson F, Nicholl J. 3CPO trialists.Noninvasive ventilation in acute cardiogenic pulmonary edema. N Engl J Med2008;359:142–51.

[18] Ferrari G, Olliveri F, De Filippi G, Milan A, Aprà F, Boccuzzi A, et al. Noninvasivepositive airway pressure and risk of myocardial infarction in acute cardiogenicpulmonary edema. Chest 2007;132:1804–9.

[19] Bellone A, Monari A, Cortellaro F, Vettorello M, Arlati S, Coen D. Myocardialinfarction rate in acute pulmonary edema: noninvasive pressure supportventilation versus continuous positive airway pressure. Crit Care Med2004;32(9):1860–5.

[20] Park M, Sangean MC, Volpe Mde S, Feltrim MI, Nozawa E, Leite PF, et al.Randomized, prospective trial of oxygen, continuous positive airway pressure, andbilevel positive airway pressure by face mask in acute cardiogenic pulmonaryedema. Crit Care Med 2004;32(12):2407–15.

[21] Cross AM, Cameron P, Kierce M, Ragg M, Kelly AM. Non-invasive ventilation inacute respiratory failure: a randomised comparison of continuous positive airwaypressure and bi-level positive airway pressure. Emerg Med J 2003;20:531–4.

[22] Nouira S, Boukef R, Bouida W, Kerkeni W, Beltaief K, Boubaker H, et al. Non-invasive pressure support ventilation and CPAP in cardiogenic pulmonary edema:a multicenter randomized study in the emergency department. Intensive CareMed 2011;37:249–56.

[23] ESC guidelines for the diagnosis and treatment of acute and chronic heart failure2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic HeartFailure 2012 of the European Society of Cardiology. Developed in collaborationwith the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail 2012;14(8):803–69.

[24] Weintraub NL, Collins SP, Pang PS, Levy PD, Anderson AS, Arslanian-Engoren C,et al. American Heart Association Council on Clinical Cardiology and Council onCardiopulmonary, Critical Care, Perioperative and Resuscitation. Acute heartfailure syndromes: emergency department presentation, treatment, and disposi-tion: current approaches and future aims: a scienti!c statement from theAmerican Heart Association. Circulation 2010;122(19):1975–96.

[25] Bersten AD, Holt AW, Vedig AE, Skowronski GA, Baggoley CJ. Treatment of severecardiogenic pulmonary edema with continuous positive airway pressuredelivered by face mask. N Engl J Med 1991;325:1825–30.

[26] Chadda K, Annane D, Hart N, Gajdos P, Raphaël JC, Lofaso F. Cardiac and respiratoryeffects of continuous positive airway pressure and noninvasive ventilation inacute cardiac pulmonary edema. Crit Care Med 2002;30:2457–61.

[27] Bellone A, Barbieri A, Ricci C, Iori E, Donateo M, Massobrio M, et al. Acute effects ofnon-invasive ventilatory support on functional mitral regurgitation in patients withexacerbation of congestive heart failure. Intensive Care Med 2002;28:1348–50.

[28] Bendjelid K, Schütz N, Suter PM, Fournier G, Jacques D, Fareh S, et al. Doescontinuous positive airway pressure by face mask improve patients with acutecardiogenic pulmonary edema due to left ventricular diastolic dysfunction? Chest2005;127:1053–8.

[29] Yoshinaga K, Burwash IG, Leech JA, Haddad H, Johnson CB, deKemp RA. The effectsof continuous positive airway pressure on myocardial energetics in patients withheart failure and obstructive sleep apnea. J Am Coll Cardiol 2007;49:450–8.

[30] Naughton MT, Rahman MA, Hara K, Floras JS, Bradley TD. Effect of continuouspositive airway pressure on intrathoracic and left ventricular transmuralpressures in patients with congestive heart failure. Circulation 1995;91:1725–31.

[31] Tkacova R, Rankin F, Fitzgerald FS, Floras JS, Bradley TD. Effects of continuouspositive airway pressure on obstructive sleep apnea and left ventricular afterloadin patients with heart failure. Circulation 1998;98:2269–75.

[32] Wysocki M. Noninvasive ventilation in acute cardiogenic pulmonary edema:better than continuous positive airway pressure? Intensive Care Med 1999;25:1–2.

[33] Nava S, Carbone G, DiBattista N, Bellone A, Baiardi P, Cosentini R, et al. Noninvasiveventilation in cardiogenic pulmonary edema: a multicenter randomized trial. Am JRespir Crit Care Med 2003;168:1432–7.

[34] Rusterholtz T, Kempf J, Berton C, Gayol S, Tournoud C, Zaehringer M, et al.Noninvasive pressure support ventilation (NIPSV) with face mask in patients withacute pulmonary edema (ACPE). Intensive Care Med 1999;25:21–8.

[35] Takeda S, Nejima J, Takano T, Nakanishi K, TakayamaM, Sakamoto A, et al. Effect ofnasal continuous positive airway pressure on pulmonary edema complicatingacute myocardial infarction. Japan Circ J 1998;62:553–8.

[36] Cotter G, Metzkor E, Kaluski E, Faigenberg Z, Miller R, Simovitz A, et al.Randomised trial of high-dose isosorbide dinitrate plus low-dose furosemideversus high-dose furosemide plus low-dose isosorbide dinitrate in severepulmonary oedema. Lancet 1998;351:389–93.

[37] Sharon A, Shpirer I, Kaluski E, Moshkovitz Y, Milovanov O, Polak R, et al. High-doseintravenous isosorbide-dinitrate is safer and better than Bi-PAP ventilationcombined with conventional treatment for severe pulmonary edema. J Am CollCardiol 2000;36(3):832–7.

1327H. Li et al. / American Journal of Emergency Medicine 31 (2013) 1322–1327