Helmet Continuous Positive Airway Pressure vs Oxygen Therapy To Improve Oxygenation in Community-Acquired Pneumonia

CHEST Original Research COMMUNITY-ACQUIRED PNEUMONIA Helmet Continuous Positive Airway Pressure vs Oxygen Therapy To Improve Oxygenation in Community-Acquired Pneumonia A Randomized, Controlled Trial Roberto Cosentini, MD; Anna Maria Brambilla, MD; Stefano Aliberti, MD; Angelo Bignamini, PhD; Stefano Nava, MD; Antonino Maffei, MD; Renato Martinotti, MD; Paolo Tarsia, MD; Valter Monzani, MD; and Paolo Pelosi, MD Objective: Our objective was to evaluate the efficacy of noninvasive continuous positive airway pressure (CPAP) delivered by helmet in improving oxygenation in comparison with oxygen therapy in community-acquired pneumonia (CAP). Methods: This was a multicenter, randomized, controlled trial enrolling patients with CAP admitted to an ED with moderate hypoxemic acute respiratory failure (ARF) ( ratio 210 and 285). Patients were randomized to helmet CPAP or standard oxygen therapy (control group). The primary end point was the time to reach a ratio. 315. After reaching this value, patients randomized to CPAP were switched to oxygen, and the proportion of subjects who could maintain a ratio. 315 at 1 h was recorded. Results: Forty-seven patients were recruited: 20 randomized to CPAP and 27 to controls. Patients randomized to CPAP reached the end point in a median of 1.5 h, whereas controls reached the end point in 48 h ( P,.001). The proportion of patients who reached the primary end point was 95% (19/20) among the CPAP group and 30% (8/27) among controls ( P,.001). One hour after reaching the primary end point, 2/14 patients in the CPAP group maintained a value. 315. Conclusions: CPAP delivered by helmet rapidly improves oxygenation in patients with CAP suffering from a moderate hypoxemic ARF. This trial represents a proof-of-concept evaluation of the potential usefulness of CPAP in patients with CAP. Trial registration: clinicaltrials.gov; Identifier: NCT00603564. CHEST 2010; 138(1):114 120 Abbreviations: ABG 5 arterial blood gas analysis; ARF 5 acute respiratory failure; CAP 5 community-acquired pneumonia; CPAP 5 continuous positive airway pressure; NIV 5 noninvasive ventilation; PEEP 5 positive end-expiratory pressure; RR 5 respiratory rate Acute respiratory failure (ARF) is a well-known complication of community-acquired pneumonia (CAP), which may persist until antibiotics and the immune system have fully controlled the primary infection. In patients with CAP, time is needed for conventional therapy to show its effect. During this critical period, the maintenance of a satisfactory oxygenation represents the main goal in the management of the ARF. Oxygen administration may not always improve gas exchange, because it does not correct the alterations in lung structure and mechanics that may occur. 114 The alveolar recruitment obtained through the application of a noninvasive continuous positive airway pressure (CPAP) has been shown to improve oxygenation, but not outcomes, in patients affected by severe hypoxemic ARF. 1 Some anecdotal cases have been reported regarding immunocompetent patients with CAP successfully treated with noninvasive CPAP in a miscellaneous population. 2-4 Beneficial effects of CPAP have also been observed in immunosuppressed patients with Pneumocystis carinii pneumonia.5. Because of the paucity of data, international guidelines recommend only cautious Original Research

trials of noninvasive ventilation (NIV) in immunocompetent patients with ARF due to CAP. 6,7 In view of the previous limited data, a first step in analyzing the possible application of CPAP in patients with CAP would be to study the pathophysiologic effect of positive end-expiratory pressure (PEEP) in recruiting alveoli in a well-defined model of ARF due to CAP. No randomized control designs have been applied to test the effects of CPAP on gas exchange in patients with CAP admitted to an ED with an early phase of ARF. Therefore, the aim of this study was to evaluate the efficacy of noninvasive CPAP in improving oxygenation in comparison with standard oxygen therapy in patients admitted to the ED with moderate hypoxemic ARF due to CAP. Materials and Methods Patients with moderate hypoxemic ARF due to CAP were recruited from four EDs in Italy between January 2006 and February 2008. Patients were admitted and treated in the ED. CAP was defined as the presence of a new pulmonary infiltrate on chest radiograph with at least one of the following: new or increased cough, an abnormal temperature (, 35.6 C or. 37.8 C) or serum leukocyte count (leukocytosis. 10,000/mL or leukopenia, 4,000/mL). Chest radiograph was interpreted by different radiologist physicians in the ED who had clinical and research experience with CAP. The study was performed in compliance with the Declaration of Helsinki and was approved by the ethics committees (clinicaltrials.gov number: NCT00603564). A written informed consent from each patient was obtained at entry into the study. Study Population The inclusion criteria were age 18 years, diagnosis of CAP as the only cause of ARF, respiratory rate (RR) 35 breaths/min, Manuscript received September 27, 2009; revision accepted January 11, 2010. Affiliations: From the Emergency Medicine Department (Drs Cosentini, Brambilla, and Monzani), Gruppo NIV, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico; the Thoracopulmonar and Cardiovascular Department (Drs Aliberti and Tarsia), University of Milan, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico; the School of Specialization in Hospital Pharmacy (Dr Bignamini), University of Milan, Milan; the Respiratory Intensive Care Unit (Dr Nava), Fondazione S. Maugeri, Istituto Scientifico di Pavia, IRCCS, Pavia; the Emergency Medicine Unit (Dr Maffei), Ospedale San Paolo, Naples; the Emergency Medicine Department (Dr Martinotti), Ospedale San Carlo, Milan; and the Department of Ambient, Health and Safety (Dr Pelosi), University of Insubria, Varese, Italy. This work was presented as an oral communication at Società Italiana di Medicina d Emergenza-Urgenza (SIMEU), VI Annual Congress, Rimini, Italy, November 5, 2008. Correspondence to: Roberto Cosentini, MD, Emergency Medicine Department, Gruppo NIV, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, Via F. Sforza 35, 20122 Milan, Italy; e-mail: roberto.cosentini@policlinico.mi.it. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( http://www.chestpubs.org/ site/misc/reprints.xhtml ). DOI: 10.1378/chest.09-2290 and ratio 200 and 300 evaluated during oxygen therapy supplied for at least 15 min through a Venturi mask (F io 2 of 0.50, flow around 40 L/m). In order to reduce the potential bias associated with the enrollment of patients near the extremes of the inclusion range of the ratio, the criterion was limited to patients with ratio 210 and 285 (approximately the central 90% of the standard inclusion criterion). Patients were excluded from the study if at least one of the following was present: diagnosis of health care-associated pneumonia, immunosuppressive state, acute cardiogenic pulmonary edema, unstable angina or acute myocardial infarction, respiratory acidosis, failure of three or more organs, systolic BP, 90 mm Hg despite fluid resuscitation or vasopressors, severe arrhythmias, contraindications to CPAP treatment, 7 or pregnancy. Study Design This was a multicenter, randomized, open-label, controlled trial in parallel groups. Patients enrolled were randomized to CPAP or standard oxygen therapy (control group). Both groups received standard antibiotic therapy for CAP in accordance with American Thoracic Society guidelines. 8 Interventions CPAP was delivered through a high-flow generator (90-140 L/min; VitalSigns Inc; Totowa, NJ) using a helmet (StarMed, Italy) as interface with a PEEP valve (VitalSigns). CPAP was applied with an initial PEEP of 10 cm H 2 O and with an F io 2 set to maintain a pulse oximetry 92%. PEEP value was maintained at 10 cm H 2 O until CPAP removal. Standard oxygen therapy was supplied through a Venturi mask with an F io 2 delivered to maintain a pulse oximetry 92%. The Fio 2 was measured using an oxygen analyzer (MiniOX I; MSA; Pittsburgh, PA) through both the Venturi mask and the helmet. Demography, medical history, and clinical and laboratory data were recorded on admission. Arterial blood gas (ABG) analysis and vital signs were recorded on admission and at 1, 6, 12, 24, and 48 h until ratio 315. This threshold was chosen, as previously adopted for the inclusion criterion, to reduce the risk of false-positive results for values of ratio equal to or slightly greater than 300. Time to ABG assessment was recorded, and actual time intervals were used for estimating the results. After reaching this threshold value, patients randomized to CPAP were switched to oxygen, whereas those randomized to oxygen continued on oxygen. Adverse events were monitored throughout the whole treatment period. Patients could withdraw from the study at any time and be removed by the attending physician if any of the following occurred: ratio, 200 with an RR. 35 breaths/min, acute failure of three or more organs, or criteria for endotracheal intubation (at least one among: respiratory arrest, severe hemodynamic instability; or at least two among: RR. 35 breaths/min, ratio, 150, an increase of the Pa co 2. 20% in comparison with the previous ABG, mental status change due to respiratory failure). End Points The primary end point on which the sample size was computed was the time to reach an improvement in terms of gas exchange, defined as a ratio 315. Patients who did not reach this threshold level before the last planned ABG measurement at 48 h were considered as failures. In addition, the following data were monitored: the proportion of subjects who could maintain, once reached, a ratio 315 at 1 and 24 h after the qualifying measurement, the frequency of adverse events, and the in-hospital mortality. www.chestpubs.org CHEST / 138 / 1 / JULY, 2010 115

Sample Size Assuming a time to reach the preset level in the control group of 1.75 days and a time in the CPAP group of 0.75 days, both with a 95% CI of 6 0.25 days, this translates into a difference of 1 day with a common standard deviation of approximately 2.7. Thus, we hypothesized that in patients in the CPAP group the primary end point could have been reached 1 6 2.75 days earlier than in the control group. Based on 80% power to detect a significant difference with a error 0.05 two-tailed, 120 patients were required for each study arm. 9 Randomization Patients were randomized using a computer-generated randomization list, unique for each center. Sequentially numbered, opaque sealed envelopes were supplied to each center, containing the indication of the treatment to be applied. The block size was known only to the study statistician. Statistical Analysis All data were statistically analyzed with SPSS, version 15, statistical software for Windows (SPSS Inc; Chicago, IL). All patients were analyzed according to the intention-to-treat principle. Descriptives with 95% CI, univariate and repeatedmeasures analysis of variance, Kruskal-Wallis test, and contingency tables were used as needed. The end point was analyzed in terms of time by comparison of medians with 95% CI by bootstrapping (5,000 runs) and by Cox survival analysis, in terms of proportion by contingency tables. 10 Dropout patients (for any reason) and patients not reaching the primary end point during observation were considered failures; time to reach the end point was replaced with the maximum planned time (48 h) for median comparison and censored at the last observed time for Cox regression. The missing values of ratio were replaced with the last observation carried forward technique to analyze its time course and analyzed by repeated measures of the analysis of variance. Results The study was prematurely interrupted after recruiting 47 patients, 20 randomized to CPAP and 27 to controls, because it soon became apparent, during the course of the trial, that patients randomized to CPAP reached the end point more quickly than anticipated in the protocol ( Fig 1 ). The steering committee requested an interim analysis at 20% enrollment to monitor the criteria for actual equipoise of the two treatments. Because this was unplanned, the Lan-DeMets approach was followed, 11 and a manifest superiority of the CPAP treatment was identified, with a nominal a exceeding the criterion for interruption set at 0.00008. There were no appreciable differences in baseline characteristics between the CPAP and the control group ( Table 1 ). All patients enrolled had one single organ failure. No significant difference in empirical antibiotic treatment was detected between the two study groups. Patients randomized to CPAP reached the end point in a median of 1.5 h, whereas controls reached the end point in 48 h ( P,.001), with a median difference of 46.5 h (95% CI, 45.9-46.8 h). The proportion of patients who reached the primary end point was 95% (19/20; 95% CI, 85%-100%) among those randomized to CPAP, and 30% (8/27; 95% CI, 12%- 47%) among controls ( P,.001). The relative chance that a patient randomized to CPAP reached the end point was 3.21 (95% CI, 1.78-5.78) and the absolute probability difference was 65% (95% CI, 46%-86%). There was no detectable center effect ( P 5.462, Breslow-Day test). At the Cox analysis corrected for Figure 1. Flow chart of the study. CPAP 5 continuous positive airway pressure. 116 Original Research

Table 1 Demographics and Baseline Characteristics of the Two Study Groups Characteristic CPAP Controls P Value Patients, No. 20 27 Demographics Age, mean 6 SD, y 65617 72613.133a Female, No. (%) 6 (30) 11 (41).449 b Comorbidities Cardiovascular disease, No. (%) 10 (50) 15 (58).604 b COPD, No. (%) 5 (25) 5 (19).591 b Liver disease, No. (%) 2 (10) 1 (3.7).383 b Diabetes mellitus, No. (%) 3 (15) 5 (19).751 b Current smokers, No. (%) 9 (45) 12 (44).970 b Severity of the disease Glasgow Coma Scale 15 15 SAPS II score 6 SD 2167.4 2165.7.848c Physical findings Systolic BP, mm Hg 6 SD 132626 135622.689a Diastolic BP, mm Hg 6 SD 78614 73612.181a Heart rate, beats/min 6 SD 89615 94616.316a Respiratory rate, breaths/min 6 SD 2764.5 2764.4.799a Temperature, C 6 SD 37.361.1 37.761.0.186a Arterial blood gas analysis, mm Hg 6 SD 125613 123610.582a ratio 6 SD 249625 246620.680a ph 6 SD 7.4660.05 7.4560.06.341a Pa co 2, mm Hg 6 SD 3466 3665.150a Bicarbonates, mmol/l 6 SD 24 63 2563.370a Radiologic findings Multilobar pneumonia, No. (%) 1 (5) 3 (11).626 b Pleural effusion, No. (%) 3 (15) 6 (23).494 b CPAP 5 continuous positive airways pressure; SAPS 5 Simplified Acute Physiology Score. a t test. b x 2 test. cmann-whitney U test. center, age, and baseline ratio, CPAP was the sole significant predictor for reaching the end point, with hazard ratio of 11.3 (95% CI, 3.51-36.32; P,.001). The substantial effect of the CPAP, as well as the rapidity at which the effect diverges from controls, is well depicted by the hazard function plot at the average of the covariates ( Fig 2 ). Among those who reached the end point, the value was significantly higher among the patients randomized to CPAP in comparison with controls (373 6 47, n 5 19 vs 342 6 24, n 5 8, respectively; P 5.031). Respiratory and cardiovascular findings after 1 h of treatment, after adjustment for sex, age, and baseline value, are given in Table 2. In accordance with the improvement of ratio, a significant decrease of the alveolar-arterial oxygen tension difference and an increase of the arterial oxygen content were detected in the CPAP group. ratio change in the first hour of treatment was compared between the two groups ( Fig 3 ). Patients who reached the primary end point during observation were monitored for maintenance of the achieved gas exchange after 1 and 24 h. In the CPAP Figure 2. Relative rapidity (cumulative hazard) for reaching ratio 315 with the two considered techniques. See Figure 1 legend for expansion of abbreviation. group, a total of 14/19 patients were monitored after 1 h, whereas 16/19 patients were monitored at 24 h. One hour after reaching the primary end point, 2/14 patients in the CPAP group maintained a value 315 ( Fig 4 ). Twenty-four hours after reaching the primary end point, 4/16 patients in the CPAP group maintained a value above the threshold. Although different from original planning, it was not possible to compare the two study groups at 6, 12, 24, and 48 h, because 19 out of 20 patients randomized to CPAP reached the primary end point within 1.5 h. No patient required intubation or died during treatment and no serious adverse events were seen. Overall, two patients had to be withdrawn for treatment because of intolerance to the device, one among the controls and one in the CPAP group. No other adverse event was observed. Discussion The main finding of this study is that in moderate hypoxemic ARF due to CAP, CPAP improved oxygenation faster and in a greater proportion of patients in comparison with standard oxygen therapy. During CPAP treatment neither cardiovascular effects nor significant adverse events related to the method were observed. Moreover, we found that the improvement of oxygenation due to CPAP is no longer evident after discontinuation of the technique. The role of CPAP in improving oxygenation in hypoxemic patients has been recently addressed in an elegant physiologic study. 12 Delclaux and colleagues 1 showed a beneficial effect of CPAP on the oxygenation www.chestpubs.org CHEST / 138 / 1 / JULY, 2010 117

Table 2 Respiratory and Cardiovascular Findings at Baseline and During the First Hour of Treatment Findings Time CPAP (Patients) Controls (Patients) P Valuea Respiratory findings, mm Hg Baseline 125613 (20) 123610 (27).001 1 h 164643 (20) 117630 (24) ratio Baseline 249625 (20) 246620 (27),.001 1 h 349669 244651 D (A 2 a)o2, mm Hg Baseline 189614 (20) 188611 (25).002 1 h 151641 (20) 191630 (22) Arterial O 2 content, ml/dl Baseline 18.861.7 (20) 17.762.4 (25).009 1 h 19.161.8 (20) 17.662.4 (21) ph Baseline 7.4660.05 (20) 7.4560.06 (25).998 1 h 7.4660.05 (20) 7.4460.06 (22) Pa co 2, mm Hg Baseline 3465.8 (20) 3665.2 (25).380 1 h 3365.6 (20) 3766.7 (22) Bicarbonates, mmol/l Baseline 2463 (20) 2563 (25).544 1 h 2464 (20) 2563 (22) Respiratory frequency, breaths/min Baseline 2764.4 (25) 2764.5 (20).186 1 h 2264.2 (13) 2565.1 (13) Cardiovascular findings Systolic BP, mm Hg Baseline 132626 (18) 135622 (25).407 1 h 127613 (9) 123616 (11) Diastolic BP, mm Hg Baseline 78614 (18) 73612 (25).794 1 h 7467.4 (9) 7568.9 (11) Heart rate, beats/min Baseline 89615 (20) 94616 (25).189 1 h 84612 (12) 93618 (13) D (A 2 a)o2 5 alveolar-arterial oxygen tension difference: [(F io ) 3 (atmospheric pressure 2 H O pressure) 2 (Paco /0.8) ] 2 Pao 2 2 2 2. See Table 1 for expansion of other abbreviation. amultivariate P for technique from the repeated-measures analysis of variance using sex as cofactor and age as covariate. of patients with hypoxemic ARF due to different causes, including those with CAP. In an observational prospective study, Hilbert et al 13 showed that CPAP was efficient in preventing intubation in immunocompromised neutropenic patients with fever. Recent literature suggests differentiating the use of CPAP and NIV in ARF due to different clinical conditions, 14 but robust clinical data are still lacking. This is the first randomized controlled trial, to our knowledge, evaluating CPAP in moderate hypoxemic ARF due to CAP. We found that CPAP application in patients with moderate hypoxemic ARF due to CAP improves oxygenation. Moreover, the most impressive finding was the rapidity of the CPAP effect in comparison with standard oxygen therapy. These results led the steering committee to prematurely stop the trial. It is likely that the improvement of oxygenation obtained during the application of CPAP may be because of the recruitment of the collapsed alveoli surrounding the inflammatory process, as well as part of those involved by the exudative flooding. This could lead to a reduction of the shunt and the ventilation-perfusion ratio mismatch due to the pneumonia, followed by an increase of the functional residual capacity and the pulmonary compliance. The hypothesis of the recruitment of the collapsed alveoli during CPAP application is indirectly suggested in our study by the vanishing of the oxygenation improvement once CPAP is discontinued. This phenomenon could be defined as the on-off effect of CPAP on pneumonia. Our finding is supported by data from Jolliet and colleagues 15 who found in patients with severe CAP treated with NIV a return of the gas exchange improvement to baseline after discontinuation of the technique. Because of this, in order to obtain not only a physiologic but also a favorable clinical effect, it could be necessary to open and keep open the lungs of patients with pneumonia through the application of CPAP while waiting for the antibiotic effect. We found no hemodynamic consequences of positive pressure application in our patients with pneumonia, in accordance with previously published data. 16 In light of this and the absence of adverse events observed in our population, CPAP could be considered a safe technique in patients with moderate hypoxemic ARF with pneumonia. The effects of CPAP/NIV have been studied in patients with CAP admitted to the ICU with severe ARF in both miscellaneous 16,17 and selected populations. 18 Data showed an improvement in terms of gas exchange, but conflicting results were present regarding clinical outcomes. The population we selected was treated outside the ICU and was composed of patients with CAP with early, moderate, and hypoxemic ARF. We decided to test CPAP instead of NIV because in the early phase of ARF due to CAP, the main problem is the flooding in the interstitial space and alveoli. 118 Original Research

Figure 3. Plot of the mean change ( 6 95% CI) of ratio in the first hour of treatment of the two considered techniques. See Figure 1 legend for expansion of abbreviation. Thus, the sole application of PEEP, leading to a recruitment of the collapsed alveoli, could sustain and improve the oxygenation without additional pressure support. Moreover, when compared with other noninvasive techniques requiring a mechanical ventilator, CPAP has proven not only to be easier to use and quicker to implement in clinical practice but also to carry smaller associated costs. The choice of helmet in the present study has been supported by data showing fewer complications and better tolerance of this interface when compared with facemask.19 Our data support these preliminary findings, showing no serious adverse events in the CPAP group. Patients were able to expectorate throughout the interface or during the short period of discontinuation of the technique during meals or family visits, and only one patient had intolerance due to claustrophobia. In view of this, the helmet could be the appropriate device to be used in patients with pneumonia who may need long periods of CPAP treatment. Is CPAP effective in improving outcomes in patients with CAP who have severe respiratory failure and are at high risk of requiring intubation? We still do not know. Our study proved the effect of CPAP in improving gas exchange in a well-defined model of hypoxemic ARF, such as CAP. The present study should be considered as a proof-of-concept study that could sustain the development of future trials analyzing more meaningful clinical outcomes and possible adverse events in a population of patients with more severe CAP treated with CPAP. These trials would be needed to evaluate the possible role of CPAP as a valid and a safe tool in the management of patients with CAP. Until further randomized con- Figure 4. Time course of the ratio in patients randomized to CPAP, at CPAP removal upon reaching the primary end point, and 1 h later. See Figure 1 legend for expansion of abbreviation. trolled trials define this issue, we do not recommend the use of this technique in patients with CAP with severe respiratory failure. Limitations of our study are mainly related to the early reaching of the end point of the CPAP group. Because we prematurely stopped the study, this led to a reduction of number enrolled and may lead to misleading conclusions. However, it is clear that noninvasive helmet CPAP was effective to achieve a faster improvement of oxygenation that was not maintained when treatment was stopped. In the CPAP group, one-fourth of patients missed the reevaluation after 1 h of treatment. This study was an unsponsored, reallife investigation, in which protocol procedures could not interfere with the normal duties of the ED. A few patients refused some arterial blood samplings, including some for reevaluations. However, we believe that this missing information because of operational reasons did not negatively affect the overall results. In conclusion, CPAP rapidly improves gas exchange in patients with CAP suffering from a moderate hypoxemic ARF. This trial represents a proof-of-concept evaluation of the potential usefulness of CPAP in patients with CAP and warrants future trials evaluating the impact of CPAP on relevant clinical and safety outcomes in patients with CAP. Acknowledgments Author contributions: Dr Cosentini: contributed to design of the study and writing of the Dr Brambilla: contributed to design of the study, enrollment of the patients, and writing of the Dr Aliberti: contributed to design of the study and writing of the Dr Bignamini: contributed to statistical analysis. Dr Nava: contributed to design of the study, enrollment of the patients, and writing of the www.chestpubs.org CHEST / 138 / 1 / JULY, 2010 119

Dr Maffei: contributed to enrollment of the patients. Dr Martinotti: contributed to enrollment of the patients. Dr Tarsia: contributed to design of the study and writing of the Dr Monzani: contributed to design of the study and writing of the Dr Pelosi: contributed to design of the study and writing of the Financial/nonfinancial disclosure: The authors have reported to CHEST the following conflicts of interest: Dr Cosentini received a consultancy fee in 2006 from StarMed. Drs Brambilla, Aliberti, Bignamini, Nava, Maffei, Martinotti, Tarsia, Monzani, and Pelosi have reported that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this Other contributions: We thank Tommaso Maraffi, MD; Federico Piffer, MD; Margherita Arioli, MD; Francesca Tantardini, MD; and the Gruppo NIV Policlinico Milan for their assistance. This work won the Noninvasive Ventilation Award of the Respiratory Intensive Care Assembly, European Respiratory Society Meeting, Vienna, Austria, September 12-16, 2009. This work also won the award for the Italian Emergency Medicine Society assembly at the SIMEU Annual Congress, November 15, 2008. References 1. Delclaux C, L Her E, Alberti C, et al. Treatment of acute hypoxemic nonhypercapnic respiratory insufficiency with continuous positive airway pressure delivered by a face mask: A randomized controlled trial. JAMA. 2000 ;284(18):2352-2360. 2. Brett A, Sinclair DG. Use of continuous positive airway pressure in the management of community acquired pneumonia. Thorax. 1993 ;48(12):1280-1281. 3. Manoury B, Daumal M, Gillon JC, Cassetto B, Fockenier F. Severe Chlamydia psittaci pneumopathy in an adult [in French ]. Ann Fr Anesth Reanim. 1984 ;3(6):456-457. 4. Pillans P. Chickenpox pneumonia. A case report. S Afr Med J. 1983 ;63(22):861-862. 5. Prevedoros HP, Lee RP, Marriot D. CPAP, effective respiratory support in patients with AIDS-related Pneumocystis carinii pneumonia. Anaesth Intensive Care. 1991 ; 19 ( 4 ): 561-566. 6. Mandell LA, Wunderink RG, Anzueto A, et al ; Infectious Diseases Society of America ; American Thoracic Society. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of communityacquired pneumonia in adults. Clin Infect Dis. 2007 ;44(suppl 2 ): S27-S72. 7. British Thoracic Society Standards of Care Committee. Non-invasive ventilation in acute respiratory failure. Thorax. 2002 ;57 (3 ):192-211. 8. Niederman MS, Mandell LA, Anzueto A, et al ; American Thoracic Society. Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med. 2001 ;163 (7 ):1730-1754. 9. Friedman LM, Furberg CD, DeMets DL. Fundamentals of Clinical Trials. 3rd ed. New York, NY : Springer-Verlag ; 1998 :109-114. 10. Lunnebor CE. Data Analysis By Resampling: Concepts and Applications. Pacific Grove, CA : Brooks/Cole/Duxbury ; 2000. 11. Lan KKG, Reboussin DM, DeMets DL. Information and information fractions for design and sequential monitoring of clinical trials. Commun Stat Theory Methods. 1994 ;23 (2):403-420. 12. L Her E, Deye N, Lellouche F, et al. Physiologic effects of noninvasive ventilation during acute lung injury. Am J Respir Crit Care Med. 2005 ;172 (9 ):1112-1118. 13. Hilbert G, Gruson D, Vargas F, et al. Noninvasive continuous positive airway pressure in neutropenic patients with acute respiratory failure requiring intensive care unit admission. Crit Care Med. 2000 ; 28 ( 9 ): 3185-3190. 14. Nava S, Carlucci A. Non-invasive pressure support ventilation in acute hypoxemic respiratory failure: common strategy for different pathologies? Intensive Care Med. 2002 ; 28 ( 9 ): 1205-1207. 15. Jolliet P, Abajo B, Pasquina P, Chevrolet JC. Non-invasive pressure support ventilation in severe community-acquired pneumonia. Intensive Care Med. 2001 ;27 (5 ):812-821. 16. Ferrer M, Esquinas A, Leon M, Gonzalez G, Alarcon A, Torres A. Noninvasive ventilation in severe hypoxemic respiratory failure: a randomized clinical trial. Am J Respir Crit Care Med. 2003 ;168 (12 ):1438-1444. 17. Hilbert G, Gruson D, Vargas F, et al. Noninvasive ventilation in immunosuppressed patients with pulmonary infiltrates, fever, and acute respiratory failure. N Engl J Med. 2001 ;344 (7 ):481-487. 18. Confalonieri M, Potena A, Carbone G, Porta RD, Tolley EA, Umberto Meduri G. Acute respiratory failure in patients with severe community-acquired pneumonia. A prospective randomized evaluation of noninvasive ventilation. Am J Respir Crit Care Med. 1999 ;160 (5 pt 1 ):1585-1591. 19. Principi T, Pantanetti S, Catani F, et al. Noninvasive continuous positive airway pressure delivered by helmet in hematological malignancy patients with hypoxemic acute respiratory failure. Intensive Care Med. 2004 ;30 (1 ):147-150. 120 Original Research