AIRWAY/ORIGINAL RESEARCH

Continuous Positive Airway Pressure Versus Bilevel Noninvasive Ventilation in Acute Cardiogenic Pulmonary Edema: A Randomized Multicenter Trial
Fabienne Moritz, MD, PhD Benoit Brousse, MD Bruno Gelle, MD Abdesslam Chajara, ScD Erwan LHer, MD, PhD Marie-France Hellot, ScD Jacques Bnichou, MD, PhD
From the Service dAccueil et dUrgences, CHU de Rouen, Hpital Charles Nicolle, University Hospital, Rouen, France (Moritz); Service dAccueil et dUrgences, CHG Elbeuf, Louviers, Val de Reuil, France (Brousse, Gelle); Dlgation la Recherche Clinique, CHU de Rouen, Hpital Charles Nicolle, University Hospital, Rouen, France (Chajara); Service dAccueil et dUrgences, CHU de Brest, France (LHer); and the Dpartement de Biostatistique et Unit Inserm 657, CHU de Rouen, Hpital Charles Nicolle, France (Benichou, Hellot).

Study objective: Patients with acute cardiogenic pulmonary edema may develop respiratory failure. Noninvasive respiratory support should be initiated rapidly to avoid tracheal intubation. The aim of this study is to compare the efcacy of continuous positive airway pressure (CPAP) delivered by the Boussignac CPAP device and bilevel positive airway pressure (bilevel PAP) in patients with acute respiratory failure caused by acute cardiogenic pulmonary edema. Methods: This prospective multicenter randomized study was conducted in 3 emergency departments. Patients were assigned to Boussignac CPAP through a facemask or to bilevel PAP, in addition to standard therapy. The main outcome was a combined criterion (tracheal intubation, death, or acute myocardial infarction). Complications, durations of ventilation, and hospitalization were also assessed. Results: After 1 hour of ventilation and at the end of the ventilation period, clinical parameters of respiratory distress and blood gas exchange signicantly improved in each treatment arm. No signicant differences were observed between the Boussignac CPAP and bilevel PAP arms for the combined criterion (5% versus 12%, respectively; odds ratio [OR] 0.4; 95% condence interval [CI] 0.0 to 1.9) and also for severe complications (9% versus 6%; OR 1.5; 95% CI 0.3 to 9.9), duration of ventilation (median for both groups 2 hours; interquartile range [IQR] 1.2 to 3.0 hours), duration of hospitalization (CPAP 8.5 [IQR 6 to 14] days; bilevel PAP 10 [IQR 7 to 16] days), or intrahospital mortality (8% versus 14%; OR 1.8 [IQR 0.4 to 8.8]). Similar results were obtained among hypercapnic patients (PaCO2 45 mm Hg). Whatever the ventilation support used, the combined criterion and severe complications were more frequently observed among hypercapnic patients. Conclusion: Both Boussignac CPAP and bilevel PAP appeared effective in rapidly improving respiratory distress even in hypercapnic patients, but they were not different in terms of patient outcome. [Ann Emerg Med. 2007;50:666-675.]
0196-0644/$-see front matter Copyright 2007 by the American College of Emergency Physicians. doi:10.1016/j.annemergmed.2007.06.488

INTRODUCTION
Acute cardiogenic pulmonary edema is the rst cause of acute respiratory distress worldwide.1 In patients with acute respiratory failure, standard treatment, including diuretics, nitroglycerin, morphine, and oxygen, may not be sufcient to reduce respiratory distress.2 In this setting, noninvasive 666 Annals of Emergency Medicine

ventilation support should be initiated rapidly, with the main goals to improve oxygenation, avoid invasive ventilation, and permit a sufcient period for medical therapy to decrease pulmonary vascular congestion.3 Moreover, noninvasive ventilation support, per se, has been shown to improve cardiogenic pulmonary edema by the establishment of a positive
Volume , . : December

Moritz et al
Editors Capsule Summary

Respiratory Support for Acute Cardiogenic Pulmonary Edema

What is already known on this topic Noninvasive ventilation support has benecial physiologic effects and reduces the rate of tracheal intubation in acute cardiogenic pulmonary edema. What question this study addressed The physiologic and clinical effects of 2 techniques for noninvasive ventilation support in acute cardiogenic pulmonary edema: continuous positive airway pressure through a Boussignac device and bilevel positive airway pressure. What this study adds to our knowledge In this 109-person randomized trial, both techniques improved clinical signs of respiratory distress and pulmonary gas exchange to a similar extent. How this might change clinical practice Either technique can be used to improve respiratory distress in patients with cardiogenic pulmonary edema, with an acceptably low rate of complications.
Figure 1. Boussignac CPAP device.

intrathoracic pressure and the reduction of left ventricular afterload. This nonpharmacologic form of treatment of acute cardiogenic pulmonary edema results in an improvement in gas exchange, a decreased need for intubation, and a decreased relative risk of mortality in comparison to standard medical treatment.3-9 Patients with severe acute cardiogenic pulmonary edema may benet from noninvasive ventilation support either with continuous positive airway pressure (CPAP) or with bilevel positive airway pressure (PAP).9,10 CPAP increases functional residual capacity, which results in a reduction of atelectasis and intrapulmonary shunt.11 Efcacy of CPAP in reducing intubation and invasive ventilation was rst demonstrated by Bersten et al.4 Randomized controlled studies of acute cardiogenic pulmonary edema patients have demonstrated signicant improvements in vital signs and gas exchange, as well as a drastic reduction in tracheal intubation rates and mortality attributable to CPAP, in comparison with standard medical treatment.4,12-15 Despite these convincing results, the use of CPAP has gained interest only recently with the extensive use of the Boussignac CPAP device. This device is of major interest because it does not require specic technical skills or the use of a ow generator or a ventilator, as does other CPAP or bilevel PAP, which explains its low cost16,17 (See Appendix E1, available online at http://www.annemergmed.com; Figure 1). Bilevel PAP provides inspiratory pressure support coupled with positive end-expiratory pressure. Bilevel PAP theoretically confers the same benets as CPAP and inspiratory assistance that unloads the respiratory muscles in acute cardiogenic pulmonary edema.18 Compared to CPAP,
Volume , . : December

bilevel PAP is more effective at reducing the work of breathing, although its use in acute cardiogenic pulmonary edema remains controversial because results from reported studies are inconsistent.3,5,7,13,19,20 In a large multicenter study, Nava et al21 did not nd bilevel PAP superior to medical treatment in avoiding intubation, although it improved dyspnea, respiratory rate, and gas exchange. In contrast, a metaanalysis showed that bilevel PAP reduced the need for invasive mechanical ventilation.22 However, this meta-analysis underscored the potential increase of myocardial infarction in the bilevel PAP group, which may be explained by the inspiratory effort of the patient, patient-ventilator asynchrony, and the rapid correction of PaCO2 values with potential coronary vasoconstriction.22 Goals of This Investigation This study aimed to compare the efcacy of CPAP delivered by the Boussignac device, adjusted on a full facemask, with bilevel PAP in patients admitted to the emergency department (ED) with respiratory failure caused by acute cardiogenic pulmonary edema.

MATERIALS AND METHODS

Study Design This randomized, prospective, nonblinded, multicenter study compared the adjuvant effects of Boussignac CPAP to bilevel PAP on respiratory distress in acute cardiogenic pulmonary edema patients treated in the ED. The recommended treatment for pulmonary edema consisted of furosemide 40 mg intravenously, or twice the patients normal dose, and isosorbide dinitrate infusion 2 to 4 mg/h, with possible additional boluses of 1 mg isosorbide dinitrate and morphine 0.05 mg/kg. Investigators were free to adapt treatment to patient clinical presentation. Setting The study was simultaneously conducted in 3 EDs: Rouen University Hospital, Brest University Hospital, and Elbeuf
Annals of Emergency Medicine 667

Respiratory Support for Acute Cardiogenic Pulmonary Edema General Hospital, all in France. No specic training for ventilation support was carried out in any of the 3 EDs. Patients were enrolled from November 2002 to March 2005. The protocol was approved by the local ethics committee and registered in ClinicalTrials.gov with the identier number NCT00213681 (2005-12-29). Each patient or the next of kin gave written informed consent either before inclusion if possible or after clinical improvement. Selection of Participants Patients older than 18 years and treated in the ED for respiratory failure caused by acute cardiogenic pulmonary edema were eligible for inclusion. Acute cardiogenic pulmonary edema was dened by the association of sudden onset of dyspnea; presence of bilateral rales on auscultation, with no medical history suggesting pulmonary aspiration or infection; or congestion found on chest radiograph. Furthermore, patients had to present 2 of the following severity criteria of respiratory failure: respiratory frequency greater than 30 breaths/min; pulse oxymetry saturation (SpO2) below 90%, with oxygen at greater than 5 L per minute through reservoir facemask; and use of accessory muscles (Patrick scale 3).23 This scale assigns scores of 1 in case of taut neck muscles but with no respiratory modulation, 2 for mild respiratory modulation in neck muscle contraction, 3 for moderate phasic activity, 4 for vigorous phasic activity, and 5 for vigorous phasic activity with abdominal paradox.23 Exclusion criteria are listed in Figure 2. Allocation to treatment was stratied for center and based on block randomization of 10 consecutive study numbers as either CPAP or Bilevel PAP. Each of the 3 study centers had blocks of sealed envelopes. Patients were allocated to consecutive study numbers as they were enrolled. Enrollment was convenience based, at the discretion of the ED treating physician. Interventions Enrolled patients were randomized to receive, at arrival in the ED, a ventilation support by a full facemask, either with the Boussignac CPAP valve or with bilevel PAP. Under CPAP, the oxygen ow was adjusted to reach an end-expiratory pressure level of 10 cm H2O.16,24 For patients assigned to bilevel PAP, the pressure level was adjusted to obtain a tidal volume between 8 and 10 mL/kg and positive end-expiratory pressure at 5 cm H2O.7,19 CPAP was discontinued when the patients breathing frequency had decreased to less than 25 breaths/min and SpO2 exceeded 90% on oxygen at 6 L per minute. Bilevel PAP was discontinued when the same criteria were obtained on a 50% FIO2. These criteria were dened according to a preliminary study in which patients could breathe with oxygen alone when they were met.16 Methods of Measurement Enrolled patients received continuous electronic vital sign monitoring, with half-hourly automatic recording of pulse rate, 668 Annals of Emergency Medicine Out-of-hospital use of CPAP or of bilevel PAP Temperature above 39C Altered mental state

Moritz et al

Severe acute or chronic airow obstruction, with no evidence of pulmonary edema Chronic renal failure Evidence of pneumonia Acute myocardial infarction, necessitating thrombolysis or primary angioplasty Immediate indication for tracheal intubation: Cardiogenic shock: systolic blood pressure Cardiac or respiratory arrest SpO2 85% with 100% FIO2 90 mm Hg

Decreased alertness or major agitation requiring sedation Clinical signs of exhaustion: active contraction of the respiratory accessory muscles, with paradoxic abdominal or thoracic motion.
Figure 2. Exclusion criteria.

oxygen saturation, and noninvasive blood pressure. The attending physicians recorded data on a preformatted data collection sheet and in the case records. Baseline characteristics of the patients, occurrence of previous pulmonary edema, and causes of acute cardiogenic pulmonary edema were collected from the computerized patient medical records and on initial clinical examination. Acute myocardial infarction diagnosis was based on serial ECG evaluation and repeated troponin measurement, in accordance with the American Heart Association/European Society of Cardiology guidelines. Troponin Ic was measured with an Immulite 2000 automaton (Society Diagnostic Products Corporation, Los Angeles, CA). With this method of measurement, our validated institutional cutoff for myocardial infarction was 1 g/L.25 Durations of the ventilation period, hospital stay, and hospital mortality were also recorded. Medical treatment of acute cardiogenic pulmonary edema in the ED was also recorded. Before beginning the ventilation procedure, after 1 hour of ventilation, and at the end of the ventilation period, blood gas levels were measured. At these points, pulse rate, systolic and diastolic blood pressure, respiratory rate and accessory muscle use, evaluated by the Patricks scale, were also recorded. Patients were monitored for severe complications during the rst 24 hours, including cardiogenic shock, gastric liquid aspiration, and switch of ventilation mode because of patient worsening.
Volume , . : December

Moritz et al

Respiratory Support for Acute Cardiogenic Pulmonary Edema

Figure 3. Flow of participants through each study stage.

Outcome Measures The main outcome was a combined criterion, dened as the occurrence of intubation, acute myocardial infarction, or death during the rst 24 hours after the beginning of the ventilation period in the ED. Secondary outcomes included systolic and diastolic blood pressure, pulse rate, accessory muscle use evaluated with the Patrick scale, respiratory rate, SpO2, durations of the ventilation period and of hospital stay, and hospital mortality. To obtain 80% power to detect a 20% difference on the main outcome (9% incidence in the CPAP group4 versus 29% in the bilevel PAP19) with Pearsons 2 test and 2-sided 0.05 level, 120 patients had to be included. Primary Data Analysis Data collection and analysis were performed independently of clinical investigations. To assess changes in mean hemodynamic, respiratory, and blood gas parameters separately within each treatment arm, the paired Student t test was used. The comparative effects of the 2 ventilation modes on the primary and secondary outcomes were assessed on an intention-to-treat basis, using all available data. For continuous variables, results for each group are given as means (SD) or as median (interquartile range), according to the shape of their distribution. Treatment differences are expressed as median or mean differences (95% condence interval [CI]). For the main outcome (intubation, acute myocardial infarction, or death) and other dichotomous variables, odds ratios (ORs) (95% CIs) were calculated with exact inference to assess treatment differences, with bilevel PAP as the reference.
Volume , . : December

For the main outcome, the OR was also estimated while 5 prognostic characteristics were separately adjusted for, namely, acute myocardial infarction (yes, no), systolic blood pressure ( 160 mm Hg, 160 mm Hg), hypercapnia (yes if PaCO2 45 mm Hg; no otherwise), pH ( 7.25, 7.25), and SpO2 ( 92%, 92%), within the framework of exact logistic regression. The same analyses were carried out separately and among patients with hypercapnia on an exploratory basis. All analyses were performed using SAS version 8.02 (SAS Institute, Inc., Cary, NC), SatXact version 3.1, and LogXact version 7 (Cytel Software Corporation, Cambridge, MA).

RESULTS
Characteristics of Study Subjects Among the 357 patients presenting with acute cardiogenic pulmonary edema during the study period in the 3 centers, 237 were not included, 68 of them because they had ventilation support with CPAP or with bilevel PAP before their arrival to the hospital, 154 because they had mild forms of acute cardiogenic pulmonary edema, and 15 because they were already intubated (Figure 3). Thus, 120 patients were enrolled in the study. Eleven patients were withdrawn from the study analysis because no informed consent was obtained (n 7) or because they were lost to follow-up (n 4). Thus, 109 patients were analyzed, 59 in the Boussignac CPAP arm and 50 in the bilevel PAP arm. For these 109 patients, baseline characteristics and causes of cardiac failure are given in Table 1; pulmonary treatment and dosages administered during the ventilation periods, in Table 2.
Annals of Emergency Medicine 669

Respiratory Support for Acute Cardiogenic Pulmonary Edema

Table 1. Baseline characteristics of the patients and etiologies of pulmonary edema.
Patient Characteristics Age, mean (SD) Weight, mean (SD) Height, mean (SD) Male/female sex ratio Troponin Ic, g/L, median (IQR) Heart disease etiologies, No. (%) Ischemic heart failure Diabetes mellitus Hypertension Valvular disease Other cardiac disease Causes of acute pulmonary edema, No. (%) Treatment noncompliance Respiratory tract infection Myocardial ischemia Acute myocardial infarction Hypertensive emergency Previous pulmonary edema Bilevel PAP (n 50) 77.7 (9.2) 72.2 (15.4) 165.3 (6.1) 29/21 0.1 (0.00.4) B-CPAP (n 59) 77.6 (9.4) 76.8 (22.9) 165.0 (8.6) 28/31 0.1 (0.00.4)

Moritz et al results were not materially altered for the main outcome (OR 0.4). Results were similar for the subgroup of hypercapnic patients (Table 4; Figure 5). Whatever the ventilation group, the combined criterion (12% versus 4%; OR 3.36 [95% CI: 0.6 to 34.4]) and the complications (12% versus 2%; OR 6.7 [95% CI: 0.8 to 308.9]) were more frequently observed among hypercapnic patients (mean PCO2 61 16 mm Hg) compared with nonhypercapnic patients (mean PCO2 38 4 mm Hg).

24 (48) 14 (28) 23 (46) 10 (20) 5 (10)

34 (58) 10 (17) 28 (47) 12 (20) 7 (12)

LIMITATIONS
The fact that we did not have a group with oxygen alone, applied with a facemask, might be considered a limitation. This point was discussed during the planning phase of the study, but it was not deemed ethical to include a group with oxygen alone because CPAP and bilevel PAP had already been shown to improve respiratory distress and to reduce the intubation rate in patients with acute cardiogenic pulmonary edema.4,19 The 2 treatment arms were slightly imbalanced with respect to potential prognostic factors such as hypercapnia. However, adjustment on these factors produced negligible changes compared with the unadjusted analysis. The French legislation requires the informed consent to be signed by the patient or by the next of kin of the patient before the beginning of the study. In an emergency situation, if the patient is not able to sign and if the family is absent, we have the possibility to begin the study and to ask patients consent and authorization to use the results when they are able to understand the explanation. In our study, 7 patients of the bilevel PAP group refused to give their informed consent after the beginning of the study, so it was unlawful to use their data. This was the main reason in the asymmetric dropout of patients. The sample size of our study was calculated according to results of initial studies that assessed CPAP and bilevel PAP in acute cardiogenic pulmonary edema.4,19 Our study did not permit us to conclude that one ventilation mode is more effective than the other, because the rate of the combined criterion and each of its elements was extremely low in both groups.

8 (16) 14 (28) 13 (26) 9 (18) 9 (18) 12 (24)

5 (8) 16 (27) 11 (19) 11 (19) 9 (15) 16 (27)

B-CPAP, Boussignac CPAP; IQR, interquartile range.

Table 2. Drug used in each study group during the ventilation period.*
Drug Used Furosemide, mg No. Isosorbide dinitrate, mg No. Morphine, mg No. Bilevel PAP (n 50) 110.2 (57.9) 50 4.4 (3.4) 47 4.3 (2.4) 15 B-CPAP (n 59) 121.0 (56.5) 59 4.7 (3.6) 53 4.0 (2.0) 16

*Results are expressed as mean (SD) among patients who received the drug. No. indicates the number of patients who received the reported treatment.

Main Results During the ventilation period, expiratory pressure was 4.9 0.9 cm H2O and pressure support was 12.0 3.2 cm H2O in the bilevel PAP group. In the CPAP group, expiratory pressure was 7.7 2.1 cm H2O, obtained with a mean oxygen ow of 26.0 3.6 L/min. All clinical and biological characteristics signicantly improved in the 2 ventilation modes at the end of the rst hour and at the end of the ventilation period compared to baseline measurements (Figure 4). Respiratory rate, respiratory distress, and pH also improved signicantly in both groups between the end of the rst hour and the end of the ventilation period (Figure 4). However, there were no signicant differences between the 2 groups for the main endpoint, with an OR 0.4 (95% CI 0.0 to 1.9), time on device, hospital-stay duration, hospital mortality, or the outcome rate of severe complications (Table 3). On separate adjustment for acute myocardial infarction, hypercapnia, systolic blood pressure, pH, and SpO2, 670 Annals of Emergency Medicine

DISCUSSION
The present multicenter randomized study compares for the rst time, to our knowledge, the efcacy of the Boussignac CPAP to bilevel PAP in the treatment of patients with acute cardiogenic pulmonary edema. After 1 hour of ventilation and at the end of the ventilation period, clinical characteristics of respiratory distress and blood gas exchange improved in each study group. Boussignac CPAP and bilevel PAP are both effective modalities for rapidly improving respiratory distress in patients with acute cardiogenic pulmonary edema, even if they are hypercapnic. Although duration of ventilation periods was the same, there was a trend for the combined criterion
Volume , . : December

Moritz et al

Respiratory Support for Acute Cardiogenic Pulmonary Edema

Figure 4. Hemodynamic, respiratory, and blood gas results at baseline (H0), after 1 hour (H1) and at the end (End) of the ventilation support period with Boussignac CPAP (n 59) or bilevel PAP (n 50) for all patients. Data are presented as box plots with median lines, 25th and 75th percentile boxes, and 10th and 90th percentiles error bars. The circles (for Boussignac CPAP) and the squares (for bilevel PAP) represent data outside the 10th to 90th percentile range.

(intubation, death, myocardial infarction) to occur more often in the bilevel PAP group than in the CPAP group. Because CPAP and particularly the Boussignac device are cheap and easy to implement in clinical practice (no ow generator or ventilator is necessary), their possible use should be kept in mind in the immediate treatment of patients with respiratory failure caused by acute cardiogenic pulmonary edema, even before arrival to the hospital. Approximately 50% of the patients in each of our study groups had hypercapnia. We found that in those patients with
Volume , . : December

hypercapnia, the combined criterion but also other severe complications more often occurred, conrming the more severe evolution of this subgroup of patients. Hypercapnia is a major criterion to promptly begin noninvasive ventilation to prevent tracheal intubation.21,26 As reported by Bellone et al,27 we demonstrated that bilevel PAP and Boussignac CPAP lowered carbon dioxide concentrations within the same range. This important result may explain the low tracheal intubation rate we observed, ie, 3%. This low tracheal intubation rate compared to that of other studies may also be explained by the early
Annals of Emergency Medicine 671

Respiratory Support for Acute Cardiogenic Pulmonary Edema

Table 3. General outcomes of the study groups.
General outcomes Time on device, h Median (IQR) Mean (95% CI) Troponin Ic, 24 h after arrival All patients, g/L, median (IQR) Mean (95% CI) Patients without MI, g/L, median (IQR) Mean (95% CI) Intubation, No. (%) Acute myocardial infarction, No. (%) Death in the rst day, No. (%) Combined criteria, No. (%) Complications, No. (%) Hospital stay (days) among nondeceased patients Median (IQR) Mean (95% CI) Hospital mortality, No. (%) Bilevel PAP (n 50) 2.0 (1.23.0) 2.8 (2.13.6) 0.6 (0.22.1) 19.9 (0.052.8) 0.4 (0.11.0) 0.8 (0.41.2) 2 (4) 3 (6) 1 (2) 6 (12) 3 (6) 10.0 (7.016.0) 13.6 (10.516.7) 4 (8) B-CPAP (n 59) 2 (1.23.0) 2.3 (1.92.8) 0.3 (0.04.8) 7.3 (1.413.2) 0.15 (0.00.6) 1.1 (0.41.8) 1 (2) 2 (3) 0 3 (5) 5 (9) 8.5 (6.014.0) 11.1 (8.813.4) 8 (14)

Moritz et al

Treatment Difference* 0.0 0.5 ( 1.3 to 0.3) 0.3 12.7 (42.316.9) 0.3 ( 0.51.2) 0.4 (0.08.4) 0.5 (0.03.4) 0 (033) 0.4 (0.01.9) 1.5 (0.39.9)

2.5 ( 6.3 to 1.2) 1.8 (0.48.8)

MI, Myocardial infarction. *Treatment difference: OR (95% CI) for dichotomous variables, median and mean difference (95% CI) for continuous variables, with bilevel PAP as the reference. Combined criterion: acute myocardial infarction, death or intubation during the rst 24 hours after the beginning of the ventilation period.

Table 4. General outcomes of the hypercapnic subgroup of patients.

General outcomes PaCO2 (mm Hg), mean SD Time on device, h Median (IQR) Mean (95% CI) Troponin Ic, 24 h after arrival, g/L Median (IQR) Mean (95% CI) Intubation, No. (%) Acute myocardial infarction, No. (%) Death in the rst day, No. (%) Combined criterion, No. (%) Complications, No. (%) Hospital stay among nondeceased patients, days Median (IQR) Mean (95% CI) Hospital mortality, No. (%) Bilevel PAP (29) 62.8 (17.7) 2.0 (1.44.7) 3.6 (2.44.8) 0.5 (0.21.9) 30.0 (0.086.1) 2 (7) 1 (3) 1 (3) 4 (14) 3 (10) 9 (715) 12.4 (9.015.7) 4 (14) B-CPAP (28) 59.8 (13.9) 2.0 (1.43) 2.4 (1.73.2) 0.6 (0.05.8) 7.6 (0.016.6) 1 (4) 2 (7) 0 3 (11) 4 (15) 7 (614) 10.2 (7.413.3) 4 (15) 0 1.2 ( 2.6 to 0.2) 0.1 22.3 ( 77.9 to 33.2) 0.5 (0.010.2) 2.17 (0.1131.5) 0 (0.040.4) 0.7 (0.14.9) 1.5 (0.211.3) 2 2.2 ( 6.4 to 2.1) 1.1 (0.26.6) Treatment Difference*

*Treatment difference: OR (95% CI) for dichotomous variables, median and mean difference (95% CI) for continuous variables, with bilevel PAP as the reference. Combined criterion: infarction, death, and tracheal intubation during the 24 rst hours after the beginning of the ventilation period.

beginning of ventilation assistance, as soon as diagnosis was established and medical treatment begun. Unlike other reported studies, we found no greater occurrence of myocardial infarction in the bilevel PAP group than in the CPAP group.11,22 Mehta et al11 interrupted their trial comparing CPAP (13 patients) to bilevel PAP (14 patients) at an early stage when a high proportion of patients with myocardial infarction had been detected in the bilevel PAP group. However, whether the infarcts preceded or were a consequence of therapy was not clear.3 Moreover, in studies that compared bilevel PAP with conventional oxygen therapy or with CPAP, no increase in acute myocardial infarction was observed in the bilevel PAP group.21,28,29 Nevertheless, caution 672 Annals of Emergency Medicine

is still advised for the use of pressure support ventilation in patients with acute coronary syndrome.30 In our study, mean duration of ventilation support was short, ie, approximately 3 hours in both ventilation modes, and similar to the 4 hours in the study by Masip et al7 and 2 hours in that by Cross et al.31 This duration differed from that reported by Nava et al,21 who evaluated bilevel PAP in the treatment of acute cardiogenic pulmonary edema in the ED, with an average period of 11.4 hours of noninvasive ventilation. This major discrepancy may be explained by our use of a simple protocol, serving as an objective indicator for weaning patients from airway pressure support. The duration of ventilation is of major importance, specically in the
Volume , . : December

Moritz et al

Respiratory Support for Acute Cardiogenic Pulmonary Edema

Figure 5. Subgroup of hypercapnic patients. Hemodynamic, respiratory, and blood gas results at baseline (H0), after 1 hour (H1), and at the end (End) of the ventilation support period with Boussignac CPAP (n 29) or bilevel PAP (n 28) for hypercapnic patients. Data are presented as box plots with median lines, 25th and 75th percentile boxes, and 10th and 90th percentiles error bars. The circles (for Boussignac CPAP) and the squares (for bilevel PAP) represent data outside the 10th to 90th percentile range.

setting of the ED, where procedures must be short and effective. Moreover, prolonged noninvasive ventilation may be associated with reduced care to the patient and dangerously delaying unavoidable tracheal intubation and invasive ventilation.32,33 To our knowledge, this is the rst multicenter study conducted in the ED that evaluated the use of 2 noninvasive ventilation modes in the treatment of acute cardiogenic
Volume , . : December

pulmonary edema. We observed a low occurrence of tracheal intubation, death, myocardial infarction, or complications, although no previous specic training was performed in each center. Our results, conrmed by the recent meta-analysis by Collins et al,10 are in favor of a widespread use of noninvasive ventilation support in the emergency care of severe acute cardiogenic pulmonary edema to avoid tracheal intubations.
Annals of Emergency Medicine 673

Respiratory Support for Acute Cardiogenic Pulmonary Edema The sustained improvement in the medical treatment of chronic heart failure has resulted in decreased incidence of severe acute cardiogenic pulmonary edema. To identify which noninvasive ventilation strategy offers the most therapeutic advantage, a large international multicenter study comparing CPAP to bilevel PAP, even before arrival to the ED, is warranted. The authors thank Richard Medeiros, Rouen University Hospital medical editor, for his valuable advice in editing the article and Julien Blot, ScD, for his technical assistance.
Supervising editor: J. Stephan Stapczynski, MD Author contributions: FM, BB, and JB conceived and designed the study. AC performed the data collection and data entry. MFH and JB performed the statistical analysis. FM, BB, BG, and EL were responsible for patient inclusions. FM and JB were responsible for the overall direction of the text and discussion. FM and JB takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, nancial, and other relationships in any way related to the subject of this article, that might create any potential conict of interest. See the Manuscript Submission Agreement in this issue for examples of specic conicts covered by this statement. Financial support was provided by grant 01/059HP (French Department of Health). This nancial support provided the study logistics and the equipment for continuous positive airway pressure treatment in each hospital, with the Boussignac CPAP, the specic owmeters, and manometers used. The authors do not report any nancial connection or conict of interest with the manufacturers, in particular Vygon Society. Financial support also provided the independent statistical analysis. Publication dates: Received for publication September 26, 2006. Revisions received March 1, 2007 and May 14, 2007. Accepted for publication June 25, 2007. Available online August 30, 2007. Reprints not available from the authors. Address for correspondence: Fabienne Moritz, MD, PhD, Dpartement dAnesthsie Ranimation, CHU de Rouen, Hpital Charles Nicolle, 1 rue de Germont, 76031 Rouen Cedex, France; 33-232888261, fax 33-232888326; E-mail Fabienne.Moritz@chu-rouen.fr.

Moritz et al
4. Bersten AD, Holt WH, Vedig AE, et al. Treatment of severe cardiogenic pulmonary edema with continuous positive airway pressure delivered by face mask. N Engl J Med. 1991;325:18251830. 5. Park M, Sangean MC, de S Volpe M, et al. Randomized, prospective trial of oxygen, continuous positive airway pressure, and bilevel positive airway pressure by face mask in acute cardiogenic pulmonary edema. Crit Care Med. 2004;32:24072415. 6. LHer E, Moriconi M, Texier F, et al. Non-invasive continuous positive airway pressure in acute hypoxaemic respiratory failure: experience of an emergency department. Eur J Emerg Med. 1998; 5:313-318. 7. Masip J, Betbes AJ, Paez J, et al. Non-invasive pressure support ventilation versus conventional oxygen therapy in acute cardiogenic pulmonary edema: a randomised trial. Lancet. 2000; 356:2126-2132. 8. Heinrichs W. Positive end-expiratory pressure. Anaesthesist. 1992;10:653-669. 9. Winck JC, Azevedo LF, Costa-Pereira A, et al. Efcacy and safety of non-invasive ventilation in the treatment of acute cardiogenic pulmonary edemaa systematic review and meta-analysis. Crit Care Med. 2006;2:1-18. 10. Collins SP, Mielniczuk LM, Wittingham HA, et al. The use of noninvasive ventilation in emergency department patients with acute cardiogenic pulmonary edema: a systematic review. Ann Emerg Med. 2006;48:260-269. 11. Mehta S, Jay G, Woolard RH, et al. Randomized, prospective trial of bilevel versus continuous positive airway pressure in acute pulmonary edema. Crit Care Med. 1997;25:620-628. 12. Pang D, Keenan SP, Cook DJ, et al. The effect of positive pressure airway support on mortality and the need for intubation in cardiogenic pulmonary edema: a systematic review. Chest. 1998;114:1185-1192. 13. Crane SD, Elliott MW, Gilligan P, et al. Randomised controlled comparison of continuous positive airways pressure, bilevel noninvasive ventilation, and standard treatment in emergency department patients with acute cardiogenic pulmonary oedema. Emerg Med J. 2004;21:155-161. 14. Lin M, Yang YF, Chiang HT, et al. Reappraisal of continuous positive airway pressure therapy in acute cardiogenic pulmonary edema. Chest. 1995;107:1379-1386. 15. L Her, Duquesne F, Girou E, et al. Noninvasive continuous positive airway pressure in elderly cardiogenic pulmonary edema patients. Int Care Med. 2004;5:882-888. 16. Moritz F, Benichou J, Vanheste M, et al. Boussignac continuous positive airway pressure device in the emergency care of acute cardiogenic pulmonary oedema: a randomized pilot study. Eur J Emerg Med. 2003;3:204-208. 17. Templier F, Dolveck F, Baer M, et al. Laboratory testing measurements of O2 delivered by Boussignac CPAP system with an input of 100% oxygen. Ann Fr Anesth Reanim. 2003;2:103107. 18. Chadda K, Annane D, Hart N, et al. Cardiac and respiratory effects of continuous positive airway pressure and non-invasive ventilation in acute cardiac pulmonary edema. Crit Care Med. 2002;30:2457-2461. 19. Rusterholtz T, Kempf J, Berton C, et al. Non-invasive pressure support ventilation (NIPSV) with face mask in patients with acute cardiogenic pulmonary edema (ACPE). Int Care Med. 1999;25: 21-28. 20. Sharon A, Shpirer I, Kaluski E, et al. High-dose intravenous isosorbide-dinitrate is safer and better than Bi-PAP ventilation combined with conventional treatment for severe pulmonary edema. J Am Coll Cardiol. 2000;36:832-837.

REFERENCES
1. Kannel WB. Incidence and epidemiology of heart failure. Heart Fail Rev. 2000;5:167-173. 2. Cotter G, Metzkor E, Kaluski E, et al. Randomised trial of highdose furosemide dinitrate plus low-dose furosemide versus highdose furosemide plus low-dose isosorbide dinitrate in severe pulmonary oedema. Lancet. 1998;351:389-393. 3. Mehta S. Continuous versus bilevel positive airway pressure in acute pulmonary edema? a good question. Crit Care Med. 2004; 32:2546-2548.

674 Annals of Emergency Medicine

Volume , . : December

Moritz et al
21. Nava S, Carbone G, DiBattista N, et al. Noninvasive ventilation in cardiogenic pulmonary edema. Am J Respir Crit Care Med. 2003; 168:1432-1437. 22. Peter JV, Moran JL, Phillips-Higues J, et al. Effect of non-invasive positive pressure ventilation (NIPPV) on mortality in patients with acute cardiogenic pulmonary oedema. Lancet. 2006;367:11551163. 23. Patrick W, Webster K, Ludwig L, et al. Non-invasive positivepressure ventilation in acute respiratory distress without prior chronic respiratory failure. Am J Respir Crit Care Med. 1996;153: 1005-1011. 24. Lenique F, Habis M, Lofaso F, et al. Ventilatory and hemodynamic effects of continuous positive airway pressure in left heart failure. Am J Respir Crit Care Med. 1997;155:500-505. 25. Lavoinne A, Cauliez B, Eltchaninoff H, et al. Analytical and clinical performance of the Immulite cardiac troponin I assay. Clin Chem. 2000;46:1989-1990. 26. Masip J, Paez J, Montserrat M, et al. Risk factors for intubation as a guide for non-invasive ventilation in patients with severe acute cardiogenic pulmonary edema. Intensive Care Med. 2003; 29:1921-1928. 27. Bellone A, Vettorello M, Monari A, et al. Noninvasive pressure support ventilation versus continuous positive airway pressure in

Respiratory Support for Acute Cardiogenic Pulmonary Edema

acute hypercapnic pulmonary edema. Crit Care Med. 2005;31:807-811. Bellone A, Monari A, Cortellaro F, et al. Myocardial infarction rate in acute pulmonary edema: noninvasive pressure support ventilation versus continuous positive airway pressure. Crit Care Med. 2004;32:1860-1865. Levitt MA. A prospective, randomised trial of BiPAP in severe acute congestive heart failure. J Emerg Med. 2001;21:363369. BTS Guideline. Non-invasive ventilation in acute respiratory failure. Thorax. 2002;57:192-211. Cross AM, Cameron P, Kierce M, et al. Non-invasive ventilation in acute respiratory failure: a randomised comparison of continuous positive airway pressure and bilevel positive airway pressure. Emerg Med. 2003;20:531-534. Wood KA, Lewis L, Von Harz B, et al. The use of noninvasive positive pressure ventilation in the emergency department: results of a randomized clinical trial. Chest. 1998;113:13391346. Giacomini M, Iapichino G, Gigaga M, et al. Short-term noninvasive pressure support ventilation prevents ICU admittance in patients with acute cardiogenic pulmonary edema. Chest. 2003;123: 2057-2061.

28.

29.

30. 31.

32.

33.

Did you know?

You can personalize the new Annals of Emergency Medicine Web site to meet your individual needs.

Visit www.annemergmed.com today to see what else is new online!

Volume , . : December

Annals of Emergency Medicine 675

APPENDIX E1
The Boussignac device is a 5.5 cm tube with a weight of 10 grams, connected to the patient via a facial mask (Figure 1). The continuous positive pressure is generated by the oxygen high ow injected via microchannels inserted into the inner wall of the tube. This oxygen high ow has the effect of a virtual diaphragm by the way of a turbulence that transforms speed into pressure. The pressure level depends on the amount of ow delivered and gen-

erated by the entrainment mechanism. An oxygen ow of 25-30 liters/min produces a pressue of 8.5-10 cmH2O.17 The owmeter used is specic, graduated to reach an oxygen pressure of 30 liters/ min. A manometer is placed inline in order to monitor pressure in the mask. This device is marketed in Europe by VYGON, Ecouen, France, and in the United States and Canada by Vitaid, New York (http://www.vitaid.com).

675.e1 Annals of Emergency Medicine

Volume , . : December