- Email: [email protected]
A prospective, randomized trial of BiPAP in severe acute congestive heart failure
The Journal of Emergency Medicine, Vol 21, No 4, pp 363–369, 2001 Copyright © 2001 Elsevier Science Inc. Printed in the USA. All rights reserved 0736-4679/01 $–see front matter
PII S0736-4679(01)00385-7
Original Contributions
A PROSPECTIVE, RANDOMIZED TRIAL OF BiPAP IN SEVERE ACUTE CONGESTIVE HEART FAILURE M. Andrew Levitt,
DO
Department of Emergency Medicine, Alameda County Medical Center–Highland Campus, Oakland, California, USA Reprint Address: M. Andrew Levitt, DO, Department of Emergency Medicine, Alameda County Medical Center–Highland Campus, 1411 East 31st Street, Oakland, CA 94602-1018
e Abstract—Noninvasive positive pressure ventilation has been found to be efficacious in the setting of acute respiratory failure, specifically in chronic obstructive pulmonary disease exacerbations. Its use in congestive heart failure (CHF) is less well established. Additionally, it has been reported that there is an increase in acute myocardial infarction (AMI) rate with the use of bilevel positive pressure ventilation (BiPAP) in CHF patients. This study examined whether BiPAP decreases the intubation rate or improves cardiopulmonary parameters in severe CHF patients compared to high flow O2 by mask (MASK), and whether there is an increase in AMI rate with the use of BiPAP. A prospective, randomized clinical trial at a county hospital teaching Emergency Department was conducted by enrolling 38 patients who were in severe CHF. Patients were randomized to receive either BiPAP or MASK in addition to adjunct therapy. Age and gender were not different between the groups. Heart rate, systolic blood pressure, diastolic blood pressure, respiratory rate, and pulse oximetry all showed no significant difference in change over time between groups, but there was a significant change over time within groups. Arterial pH, pCO2, and pO2 also showed no significant difference in change over time between groups, but there was a significant change over time within groups. The intubation rate for BiPAP was 23.8% (5) vs. MASK at 41.2% (7). The AMI rate was 19% (4) in the BiPAP group and 29.4% (5) in the MASK group. No true differences were detected between groups for increased oxygenation or a reduction in intuba-
tion rate. An increase in AMI rate with BiPAP was not found in this study as previously reported. This study provides support for a larger clinical trial assessing the safety and efficacy of BiPAP in acute CHF. © 2001 Elsevier Science Inc. e Keywords—noninvasive positive pressure ventilation; BiPAP; congestive heart failure; randomized trial
INTRODUCTION Acute cardiogenic pulmonary edema is a frequent cause of respiratory failure among patients presenting to the Emergency Department (ED). Initially, most of these patients can be treated with diuretics, oxygen, and nitrates; some require endotracheal intubation and mechanical ventilation to survive. These patients are generally in extremis, appearing tachypneic and diaphoretic, or are lethargic and too tired from rapid breathing to oxygenate themselves. Although a lifesaving procedure, artificial airways may lead to infectious complications and injury to the trachea (1–3). Additionally, endotracheal intubation and mechanical ventilation may prolong intensive care unit and hospital stays because additional time may be necessary for weaning or for managing the complications of intubation. Noninvasive positive pressure ventilation (NPPV), a possible alternative to endotracheal intubation, was first described more than 50 years ago (4). Today, there are
Presented at the 1999 Society for Academic Emergency Medicine Meeting, Boston, Massachusetts.
RECEIVED: 22 September 2000; FINAL ACCEPTED: 24 April 2001
SUBMISSION RECEIVED:
2 March 2001;
363
364
M. A. Levitt
basically two types of devices that can be employed for NPPV. The first type is referred to as a continuous positive airway pressure device (CPAP) and the second type, a newer version of CPAP, is referred to as a bi-level positive airway pressure device (BiPAP; Respironics, Inc., Murrysville, PA). Both types of devices have been studied as to their efficacy in acute respiratory failure. When this study was initiated there were three randomized controlled clinical trials (RCCT) of CPAP in acute pulmonary edema (5– 8). All three demonstrated a reduction in intubation rate. At the time this study was initiated, BiPAP had yet to be studied in a RCCT in cardiogenic pulmonary edema. The present study was designed to test the hypothesis that BiPAP could reduce the rate of endotracheal intubation and improve cardiopulmonary parameters of patients presenting to the ED in severe congestive heart failure (CHF). While this present study was ongoing, in April of 1997 a fourth RCCT of CPAP in acute pulmonary edema was published by Mehta et al. Their protocol not only evaluated CPAP, but studied it against BiPAP. The study was stopped early because of a larger prevalence of acute myocardial infarction (AMI) in the BiPAP group (9). The present study was interrupted when the author became aware of the Mehta et al. study, and the prevalence of AMI was compared between groups. Based on the findings of the present study, it was decided that it would be of interest to present these data as preliminary findings. Additionally, the different intubation rates found in this study led to a much higher sample size calculation than could be achieved at one institution. Therefore, the study was not restarted.
MATERIALS AND METHODS The study was approved by the hospital’s Institutional Review Board. Because of their state of extremis, patients at first gave verbal consent, and then written consent was obtained as soon as possible. Subjects were recruited from patients presenting to the ED with severe respiratory distress and suspected CHF. Severe respiratory distress was evidenced by tachypnea (generally 30 breaths/min), diaphoresis, or accessory muscle use. Congestive heart failure was suspected by clinical findings of pulmonary rales, distended neck veins, peripheral edema, or a history of CHF. Patients were enrolled only when either the author or one of the research personnel and one of the respiratory therapists participating in the study were present in the ED. This was to ensure the protocol and data collection were done correctly. The research personnel or the author obtained consent from the patient and collected data for the study.
Patients were not eligible for study entry if they met the above criteria but required immediate intubation; this was based on a decision by the senior physician that the patient had no respiratory reserve and required immediate mechanical ventilatory support. The senior physician was either an attending or senior level Emergency Medicine resident. Patients who were initially enrolled in the study, but subsequently were found not to have CHF on chest radiograph, were excluded from data analysis. The radiographical findings included enlarged cardiac silhouette, pulmonary venous distension and possible redistribution to the apices, pleural or interlobar effusions, or aveolar edema. When patients met the above entry criteria on initial evaluation on arrival in the ED, they were randomized from a previously computer-generated list to receive either BiPAP or high-flow oxygen by mask (MASK). Patients additionally were treated at the senior physician’s discretion with sublingual, cutaneous, or i.v. nitroglycerin, i.v. lasix, and i.v. morphine. Baseline measurements included blood pressure, heart rate, respiratory rate, Borg dyspnea scale score (10), pulse oximetry, and arterial blood gas (ABG). If the patient was randomized to the BiPAP group, a mask was placed on the patient and the machine engaged by a respiratory technician. The spontaneous/timed mode was employed. This mode allowed the machine to be triggered by the patient’s spontaneous breathing. If the patient failed to breathe, the unit would cycle to the inspiratory phase based on a preset interval. The machine was started with an initial inspiratory positive airway pressure (IPAP) of 8 cm H2O and an initial expiratory positive airway pressure of 3 cm H2O. These initial pressures could be adjusted in 2 cm H2O increments, maintaining a pressure support of 5 cm H2O (IPAP-expiratory positive airway pressure). The mask could be either a face type or a nose type depending on which worked the best for the patient. Because most patients were mouth breathing due to severe respiratory distress, the face mask was more commonly employed. The intention to treat principle was maintained in this study. This meant that in the case where a patient could not tolerate BiPAP and was switched to a simple high flow oxygen mask, the patient still would be analyzed as being in the BiPAP group. The reverse situation (switching from ambient pressure mask to BiPAP) was not permitted. The previously mentioned baseline cardiopulmonary measurements were repeated for study data collection at 30 min, 60 min, and 120 min from patient entry into the study protocol. The exception was the ABG, which was obtained at 60 min and 120 min after the baseline ABG. No further physiologic data were collected after 120 min. If the patient was intubated before the 2 h end of the study, the study was considered concluded for that pa-
BiPAP In Congestive Heart Failure
tient. It was then up to the managing physician to decide how to further treat the patient. The decision to endotracheally intubate the patient was made by the senior physician treating the patient based on clinical or ABG findings. The decision was made without input from the research personnel present. These criteria included severe unrelenting respiratory distress, a deterioration in mental status, or further deterioration in vital signs with increasing heart rate or respiratory rate or a decrease in blood pressure. A decrease in pO2 below 60 or a rise in pCO2 above 50 along with the above signs of clinical deterioration warranted intubation. The primary outcome measurement was intubation rate within the first 24 h of study entry. Secondary outcome measurements included blood pressure, heart rate, respiratory rate, pulse oximetry, Borg dyspnea scale score, arterial pH, arterial oxygen, and arterial CO2. As previously stated, the prevalence rates of AMI for each group were compared post hoc. The presence of an AMI was determined from review of the medical records and based on the World Health Organization (WHO) criteria. Acute myocardial infarction was considered present if it occurred during the first 24 h of hospitalization.
Statistical Analysis Group comparisons of AMI rates, intubation rates, gender distributions, and mortality rates were analyzed using 2 testing to detect proportional differences between groups. Comparisons of group intubation rates and AMI rates are also reported as the mean difference in proportions with a 95% confidence interval (95% CI). Age and hospital length of study between groups were analyzed by using an unpaired Student’s t test. Secondary outcome measurements were all analyzed by using repeated measures ANOVA testing for between group and within group differences. A sample size calculation was performed. Based on a previously published trial of CPAP in severe cardiogenic pulmonary edema, a sample size of 21 patients per group was calculated to detect a 30% difference in intubation rates at a two-tailed alpha level of .05 and a beta level of .20 (6).
365
Figure 1. The systolic and diastolic blood pressure changes for the BiPAP and MASK groups.
were randomized to the BiPAP group and 17 patients to the MASK group. Age was 67.4 ⫾ 15 years for the BiPAP group and 68.5 ⫾ 15 years for the MASK group, p ⫽ 0.8173. The gender distribution for the BiPAP group was 52.4% women and for the MASK group was 82.4% women, p ⫽ 0.0528. Because no true difference was detected in gender distribution between groups, no further analysis was done. Four patients did not tolerate the BiPAP treatment and were switched to an ambient pressure mask. There was no statistical difference between groups in the use of nitroglycerin (p ⫽ 0.2064), morphine (p ⫽ 0.4194), or lasix (p ⫽ 0.7775). These data included prehospital and in hospital medication administration. Specifically, 33.3% of the patients in both groups received i.v. nitroglycerin. No further medication analysis was done. Systolic blood pressure, p ⫽ 0.3327 (Figure 1); diastolic blood pressure, p ⫽ 0.0826 (Figure 1); heart rate, p ⫽ 0.671 (Figure 2); respiratory rate, p ⫽ 0.4495 (Figure 3); Borg dyspnea scale score (Figure 4), p ⫽
RESULTS The study period commenced in December 1995 and finished in June 1997. During this period of time, four patients who did not subsequently demonstrate CHF on chest radiograph were initially entered into the study and were removed from data analysis, resulting in a convenience sample of 38 patients satisfying study criteria and being included in the data analysis. Twenty-one patients
Figure 2. Heart rate changes for the BiPAP and MASK groups.
366
M. A. Levitt
Figure 3. Changes in respiratory rate for the BiPAP and MASK groups. Figure 5. The pulse oximetry measurements the BiPAP and MASK groups.
0.7247; and pulse oximetry, p ⫽ 0.7427 (Figure 5) all showed no difference in change over time between groups, but did show significant change over time within groups. Arterial pH, p ⫽ 0.1954 (Figure 6); arterial CO2, p ⫽ 0.2192 (Figure 7); and arterial oxygen, p ⫽ 0.4876 (Figure 8) also showed no significant difference in change over time between groups, but did show a significant change over time within groups. Three patients died in the BiPAP group (15.0%), and three patients died in the MASK group (21.4%). No patients died during the study period. The time to death ranged from 1 day to 38 days with a median of 10 days. The mean length of hospital stay was 8.1 ⫾ 6.4 days for the MASK group and 7.3 ⫾ 8.0 days for the BiPAP group, p ⫽ 0.7446. The intubation rate was 5 (23.8%, 95% CI ⫽ 5.6 – 42.0%) in the BiPAP group vs. 7 (41.2%, 95% CI ⫽ 17.8 – 64.6%) in the MASK group. All intubations took place during the 2 h study period. The mean difference in proportions between intubation rates was 17.4% (95% CI ⫽ –12.3 to 47.0%). The AMI prevalence was four patients (19%, 95% CI ⫽ 2.2–35.8%) in the BiPAP group and five patients (29.4%, 95% CI ⫽ 7.7–51.1%) in the MASK group. The mean difference in proportions
Figure 4. Measurements in Borg Dyspnea Scale Scores for the BiPAP and MASK groups.
between groups rates was –10.4% (95% CI ⫽ –37.8 to 17%). No patients had an electrocardiogram, which was consistent with AMI on presentation. Cardiac markers were not analyzed other than to be used to determine the presence of AMI as a discharge diagnosis.
DISCUSSION This present study demonstrated no true difference in intubation rate with the addition of BiPAP in patients presenting in acute severe CHF. An improvement in arterial oxygenation and intubation rate would be consistent with previous randomized controlled trials of CPAP in cardiogenic pulmonary edema (5– 8). Perhaps of greatest interest is that there was not an increase in AMI rate found with the use of BiPAP. Since the first report of the use of what has generally been termed NPPV, the majority of research has been with the CPAP version. This device delivers a constant
Figure 6. The arterial pH changes over time for the BiPAP and MASK groups.
BiPAP In Congestive Heart Failure
Figure 7. The arterial CO2 changes for the BiPAP and MASK groups.
positive airway pressure that can be adjusted to the desired level. The chosen pressure then would be maintained during the entire respiratory cycle. Continuous positive airway pressure has been studied mostly in the treatment of sleep apnea and chronic obstructive pulmonary disease (COPD). To date there are four RCCT’s of CPAP in acute cardiogenic pulmonary edema. The first, by Rasanen et al., was published in 1985 (5). This study of 40 patients found during the first 10 min of treatment a slight increase in arterial blood oxygen partial pressure and a slight decrease in respiratory rate, systolic arterial pressure, and heart rate in the CPAP group. The ambient airway pressure group showed only a slight change in respiratory rate. The improvement in arterial blood oxygenation persisted throughout the investigation. Thirteen patients (65%) in the control group and seven patients (35%) in the CPAP group met the authors’ criteria for treatment failure during the study (p ⫽ 0.068). In 1991, Berstein et al. reported on a study of 39 patients who were randomized to receive CPAP or oxygen alone. At 30 min of treatment, the CPAP group showed greater improvements in respiratory rate, arterial CO2 tension, arterial pH, and the ratio of arterial oxygen
Figure 8. Changes in arterial oxygenation over time in the BiPAP and MASK groups.
367
tension to the fraction of inspired oxygen (6). After 24 h, however, there were no significant differences between the two treatment groups in any of these respiratory indexes. Seven (35%) of the patients who received oxygen alone, but none who received oxygen plus CPAP, required intubation and mechanical ventilation (p ⫽ 0.005). However, no significant difference was found in in-hospital mortality (oxygen alone, four of 20 patients; oxygen plus CPAP, two of 19; p ⫽ 0.36). Berstein et al. concluded that CPAP in severe cardiogenic pulmonary edema can result in early physiologic improvement and reduce the need for intubation and mechanical ventilation. They further concluded that their short-term study could not establish whether CPAP has any long-term benefit or whether a larger study would have shown a difference in mortality between the treatment groups. Lin and Chiang in 1991 and 1995 reported in two papers on an RCCT of CPAP vs. face-mask oxygen therapy in acute cardiogenic pulmonary edema (7,8). The first report included 80 patients, and the second report included a final count of 100 patients. The authors reported that at 3 h, PaO2 in the CPAP group showed a significant increase, whereas the intrapulmonary shunt and alveolar-arterial oxygen tension gradient was significantly reduced (p ⬍ 0.005). The CPAP group had significantly lower rate-pressure product and higher stroke volume index compared with the control group. The therapeutic failure rate over 6 h was 24% in the CPAP group and 50% in the control group (p ⬍ 0.01). The CPAP group had a lower incidence of tracheal intubation, at eight of 50 patients (16%) vs. the control group, 18 of 50 patients (36%), p ⬍ 0.01. There was no significant difference in short-term mortality and hospital stay between the study groups. The authors concluded that CPAP resulted in physiologic cardiovascular and pulmonary function improvement and significantly reduced the need for intubation. Mortality was not significantly decreased, but this may have been because of a small sample size. Finally, Keenan et al. reported a meta-analysis in 1997 on the effect of NPPV on patients in acute respiratory failure (11). The study found that NPPV was associated with a decreased mortality (odds ratio ⫽ 0.29; 95% CI ⫽ 0.15– 0.59) and a decreased need for endotracheal intubation (odds ratio ⫽ 0.20; 95% CI ⫽ 0.11 to 0.36). However, the study concluded that these effects were restricted to patients whose cause of acute respiratory failure was an exacerbation of COPD. The study further stated that future research is warranted to determine whether NPPV confers benefit in patients without COPD who have acute respiratory failure. More recently, a modification of the CPAP apparatus was developed and named BiPAP. The device was so named because, unlike CPAP, the pressure during inspi-
368
ration and expiration could be adjusted separately. This pressure difference would improve ventilation. Mehta et al. took the next step and designed and executed a study comparing CPAP and BiPAP in acute CHF patients (9). The study concluded that patients receiving BiPAP had significant improvements in PaCO2, pH, heart rate, breathing frequency, and dyspnea score within 30 min of treatment whereas the CPAP group showed significant improvement only in breathing frequency. The present study does not see the improvement in the above cardiopulmonary parameters, nor significant PaO2 improvement. Also in the Mehta et al. study, only one of 13 patients (7.7%) in the CPAP group and one of 14 patients (7.1%) in the BiPAP group required intubation. The 95% CI for the BiPAP group can be calculated as – 6.4 –20.6%. This BiPAP intubation rate is different from that found in the present study, which was 23.8% (95% CI ⫽ 5.6 – 42.0%). Because there is an overlap of the 95% CIs between the Mehta et al. study and the present study, the intubation rates for the BiPAP groups may or may not be the same. The Mehta et al. study was stopped prematurely because of a disproportionate occurrence of AMIs. The study found an AMI rate of 71% (n ⫽ 10) in the BiPAP group and 31% (n ⫽ 4) in the CPAP group. The 95% CI for the BiPAP group can be calculated as 46.0 –94.0%. The AMI rate in the BiPAP group for the present study was 19% (95% CI ⫽ 2.2–35.8%). It is interesting to note that not only was there a much lower prevalence of AMI with BiPAP found in the present study, but there was no overlap, by 10%, with the 95% CIs of these two studies. This implies that the AMI rate with BiPAP in the present study was less than that of the Mehta et al. study by at least 10%. It is beyond the scope of this present study to determine the reason for the difference. There is speculation that the high percentage of chest pain at 71% in the BiPAP group in the Mehta et al. study may be a reason for the high AMI rate (12). For one of the prior CPAP studies that did look at AMI rates, a similar occurrence to the present study was found between groups: 20% (N ⫽ 10) in the CPAP group and 22% (N ⫽ 11) in the control group (8). A limitation of this study is its power. The study was stopped four patients short of the projected sample size. However, the intubation rate findings in this study differed substantially from those in the Bersten et al. study used to determine sample size (6). That study found a need for intubation in seven (35%) of the patients who received oxygen alone and in none of the patients receiving oxygen plus CPAP. Based on the intubation rate findings in the present study, a sample size estimate of 90 patients was calculated per group to detect a 20% difference in intubation rates with an expected intubation rate in the BiPAP group of 25% at a two-tailed ␣ ⫽ .05 and
M. A. Levitt
 ⫽ .2. It was decided that it would not be possible to achieve a sample size this large at this institution. Therefore, the study was not restarted. The decision was made to present the findings of the present study in an attempt to develop a multicenter clinical trial. It was recognized that with the concern of an increase in AMI rate with the use of BiPAP from the Mehta et al. study, it would be difficult to justify any further such research. However, the presentation of the findings of the present study may change that opinion. A second limitation of the study was lack of standardization of adjunct treatment guidelines. The use of pharmacologic treatments was left to the discretion of the senior treating physician. This may have affected clinical outcome differences between groups. However, as was reported, no significant differences were detected between study groups with the different medications used. A third limitation is that the study patients were not consecutive. Enrolling patients only when both a member of the research team and a respiratory therapist were present does not present an obvious bias, but could affect the results. As previously mentioned, it was necessary to have personnel who were knowledgeable about the study available to assure proper data collection and to obtain consent. The decision to intubate was not influenced by the study personnel or investigator. Various ED senior physicians were present over the study period managing the patients.
CONCLUSIONS No true difference was found with the use of BiPAP vs. MASK in patients with severe acute CHF seen in the ED. Importantly, no true difference or trend toward an increase in AMI rates was found with the use of BiPAP. This study provides support for a larger clinical trial to provide a sample size with sufficient power to determine if these findings hold true.
Acknowledgments—I acknowledge Robert Feldman for his help with initiating this study and the Respiratory Therapy Department (specifically LaVerne Grant), Carolyn Celi, Matthew Waxman, and Irene Hsu for their help in performing this study. Thanks also to Darlene Jones for her assistance with manuscript preparation.
REFERENCES 1. Pingleton SK. Complications of acute respiratory failure. Am Rev Respir Dis 1988;137:1463–93. 2. Staufer JL, Olson DE, Petty TL. Complications and consequences
BiPAP In Congestive Heart Failure
3. 4. 5. 6. 7.
of endotracheal intubation and tracheotomy: a prospective study of 150 critically ill adult patients. Am J Med 1981;70:65–76. Fagon JY, Chastre J, Hance AJ, et al. Nosocomial pneumonia in ventilated patients: a cohort study evaluating attributable mortality and hospital stay. Am J Med 1993;94:281– 8. Poulton PE. Left-sided heart failure with pulmonary edema: its treatment with the “pulmonary plus pressure machine.” Lancet 1936;2:981–3. Rasanen J, Heikkila J, Downs J, et al. Continuous positive airway pressure by face mask in acute cardiogenic pulmonary edema. Am J Cardiol 1985;55:296 –300. Bersten AD, Holt AW, Vedig AE, et al. Treatment of severe cardiogenic pulmonary edema with continuous positive airway pressure delivered by face mask. N Engl J Med 1991;325:1825–30. Lin M, Chiang HT. The efficacy of early continuous positive airway pressure therapy in patients with acute cardiogenic pulmonary edema. J Formosan Med Assoc 1991;90:736 – 43.
369 8. Lin M, Yang YF, Chiang HT. Reappraisal of continuous positive airway pressure therapy in acute cardiogenic pulmonary edema. Chest 1995;107:1379 – 86. 9. Mehta S, Jay GD, Woolard, RH, et al. Randomized, prospective trial of bi-level versus continuous positive airway pressure in acute pulmonary edema. Crit Care Med 1997;25:620 – 8. 10. Wilson RC, Jones PW. A comparison of the visual analogue scale and modified Borg scale for the measurement of dyspnea during exercise. Clin Sci 1989;76:277– 82. 11. Keenan SP, Kernerman PD, Cook DJ, et al. Effect of noninvasive positive pressure ventilation on mortality in patients admitted with acute respiratory failure: a meta-analysis. Crit Care Med 1997;25: 1685–92. 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. Chest 1998;114:1185–92.
PII S0736-4679(01)00385-7
Original Contributions
A PROSPECTIVE, RANDOMIZED TRIAL OF BiPAP IN SEVERE ACUTE CONGESTIVE HEART FAILURE M. Andrew Levitt,
DO
Department of Emergency Medicine, Alameda County Medical Center–Highland Campus, Oakland, California, USA Reprint Address: M. Andrew Levitt, DO, Department of Emergency Medicine, Alameda County Medical Center–Highland Campus, 1411 East 31st Street, Oakland, CA 94602-1018
e Abstract—Noninvasive positive pressure ventilation has been found to be efficacious in the setting of acute respiratory failure, specifically in chronic obstructive pulmonary disease exacerbations. Its use in congestive heart failure (CHF) is less well established. Additionally, it has been reported that there is an increase in acute myocardial infarction (AMI) rate with the use of bilevel positive pressure ventilation (BiPAP) in CHF patients. This study examined whether BiPAP decreases the intubation rate or improves cardiopulmonary parameters in severe CHF patients compared to high flow O2 by mask (MASK), and whether there is an increase in AMI rate with the use of BiPAP. A prospective, randomized clinical trial at a county hospital teaching Emergency Department was conducted by enrolling 38 patients who were in severe CHF. Patients were randomized to receive either BiPAP or MASK in addition to adjunct therapy. Age and gender were not different between the groups. Heart rate, systolic blood pressure, diastolic blood pressure, respiratory rate, and pulse oximetry all showed no significant difference in change over time between groups, but there was a significant change over time within groups. Arterial pH, pCO2, and pO2 also showed no significant difference in change over time between groups, but there was a significant change over time within groups. The intubation rate for BiPAP was 23.8% (5) vs. MASK at 41.2% (7). The AMI rate was 19% (4) in the BiPAP group and 29.4% (5) in the MASK group. No true differences were detected between groups for increased oxygenation or a reduction in intuba-
tion rate. An increase in AMI rate with BiPAP was not found in this study as previously reported. This study provides support for a larger clinical trial assessing the safety and efficacy of BiPAP in acute CHF. © 2001 Elsevier Science Inc. e Keywords—noninvasive positive pressure ventilation; BiPAP; congestive heart failure; randomized trial
INTRODUCTION Acute cardiogenic pulmonary edema is a frequent cause of respiratory failure among patients presenting to the Emergency Department (ED). Initially, most of these patients can be treated with diuretics, oxygen, and nitrates; some require endotracheal intubation and mechanical ventilation to survive. These patients are generally in extremis, appearing tachypneic and diaphoretic, or are lethargic and too tired from rapid breathing to oxygenate themselves. Although a lifesaving procedure, artificial airways may lead to infectious complications and injury to the trachea (1–3). Additionally, endotracheal intubation and mechanical ventilation may prolong intensive care unit and hospital stays because additional time may be necessary for weaning or for managing the complications of intubation. Noninvasive positive pressure ventilation (NPPV), a possible alternative to endotracheal intubation, was first described more than 50 years ago (4). Today, there are
Presented at the 1999 Society for Academic Emergency Medicine Meeting, Boston, Massachusetts.
RECEIVED: 22 September 2000; FINAL ACCEPTED: 24 April 2001
SUBMISSION RECEIVED:
2 March 2001;
363
364
M. A. Levitt
basically two types of devices that can be employed for NPPV. The first type is referred to as a continuous positive airway pressure device (CPAP) and the second type, a newer version of CPAP, is referred to as a bi-level positive airway pressure device (BiPAP; Respironics, Inc., Murrysville, PA). Both types of devices have been studied as to their efficacy in acute respiratory failure. When this study was initiated there were three randomized controlled clinical trials (RCCT) of CPAP in acute pulmonary edema (5– 8). All three demonstrated a reduction in intubation rate. At the time this study was initiated, BiPAP had yet to be studied in a RCCT in cardiogenic pulmonary edema. The present study was designed to test the hypothesis that BiPAP could reduce the rate of endotracheal intubation and improve cardiopulmonary parameters of patients presenting to the ED in severe congestive heart failure (CHF). While this present study was ongoing, in April of 1997 a fourth RCCT of CPAP in acute pulmonary edema was published by Mehta et al. Their protocol not only evaluated CPAP, but studied it against BiPAP. The study was stopped early because of a larger prevalence of acute myocardial infarction (AMI) in the BiPAP group (9). The present study was interrupted when the author became aware of the Mehta et al. study, and the prevalence of AMI was compared between groups. Based on the findings of the present study, it was decided that it would be of interest to present these data as preliminary findings. Additionally, the different intubation rates found in this study led to a much higher sample size calculation than could be achieved at one institution. Therefore, the study was not restarted.
MATERIALS AND METHODS The study was approved by the hospital’s Institutional Review Board. Because of their state of extremis, patients at first gave verbal consent, and then written consent was obtained as soon as possible. Subjects were recruited from patients presenting to the ED with severe respiratory distress and suspected CHF. Severe respiratory distress was evidenced by tachypnea (generally 30 breaths/min), diaphoresis, or accessory muscle use. Congestive heart failure was suspected by clinical findings of pulmonary rales, distended neck veins, peripheral edema, or a history of CHF. Patients were enrolled only when either the author or one of the research personnel and one of the respiratory therapists participating in the study were present in the ED. This was to ensure the protocol and data collection were done correctly. The research personnel or the author obtained consent from the patient and collected data for the study.
Patients were not eligible for study entry if they met the above criteria but required immediate intubation; this was based on a decision by the senior physician that the patient had no respiratory reserve and required immediate mechanical ventilatory support. The senior physician was either an attending or senior level Emergency Medicine resident. Patients who were initially enrolled in the study, but subsequently were found not to have CHF on chest radiograph, were excluded from data analysis. The radiographical findings included enlarged cardiac silhouette, pulmonary venous distension and possible redistribution to the apices, pleural or interlobar effusions, or aveolar edema. When patients met the above entry criteria on initial evaluation on arrival in the ED, they were randomized from a previously computer-generated list to receive either BiPAP or high-flow oxygen by mask (MASK). Patients additionally were treated at the senior physician’s discretion with sublingual, cutaneous, or i.v. nitroglycerin, i.v. lasix, and i.v. morphine. Baseline measurements included blood pressure, heart rate, respiratory rate, Borg dyspnea scale score (10), pulse oximetry, and arterial blood gas (ABG). If the patient was randomized to the BiPAP group, a mask was placed on the patient and the machine engaged by a respiratory technician. The spontaneous/timed mode was employed. This mode allowed the machine to be triggered by the patient’s spontaneous breathing. If the patient failed to breathe, the unit would cycle to the inspiratory phase based on a preset interval. The machine was started with an initial inspiratory positive airway pressure (IPAP) of 8 cm H2O and an initial expiratory positive airway pressure of 3 cm H2O. These initial pressures could be adjusted in 2 cm H2O increments, maintaining a pressure support of 5 cm H2O (IPAP-expiratory positive airway pressure). The mask could be either a face type or a nose type depending on which worked the best for the patient. Because most patients were mouth breathing due to severe respiratory distress, the face mask was more commonly employed. The intention to treat principle was maintained in this study. This meant that in the case where a patient could not tolerate BiPAP and was switched to a simple high flow oxygen mask, the patient still would be analyzed as being in the BiPAP group. The reverse situation (switching from ambient pressure mask to BiPAP) was not permitted. The previously mentioned baseline cardiopulmonary measurements were repeated for study data collection at 30 min, 60 min, and 120 min from patient entry into the study protocol. The exception was the ABG, which was obtained at 60 min and 120 min after the baseline ABG. No further physiologic data were collected after 120 min. If the patient was intubated before the 2 h end of the study, the study was considered concluded for that pa-
BiPAP In Congestive Heart Failure
tient. It was then up to the managing physician to decide how to further treat the patient. The decision to endotracheally intubate the patient was made by the senior physician treating the patient based on clinical or ABG findings. The decision was made without input from the research personnel present. These criteria included severe unrelenting respiratory distress, a deterioration in mental status, or further deterioration in vital signs with increasing heart rate or respiratory rate or a decrease in blood pressure. A decrease in pO2 below 60 or a rise in pCO2 above 50 along with the above signs of clinical deterioration warranted intubation. The primary outcome measurement was intubation rate within the first 24 h of study entry. Secondary outcome measurements included blood pressure, heart rate, respiratory rate, pulse oximetry, Borg dyspnea scale score, arterial pH, arterial oxygen, and arterial CO2. As previously stated, the prevalence rates of AMI for each group were compared post hoc. The presence of an AMI was determined from review of the medical records and based on the World Health Organization (WHO) criteria. Acute myocardial infarction was considered present if it occurred during the first 24 h of hospitalization.
Statistical Analysis Group comparisons of AMI rates, intubation rates, gender distributions, and mortality rates were analyzed using 2 testing to detect proportional differences between groups. Comparisons of group intubation rates and AMI rates are also reported as the mean difference in proportions with a 95% confidence interval (95% CI). Age and hospital length of study between groups were analyzed by using an unpaired Student’s t test. Secondary outcome measurements were all analyzed by using repeated measures ANOVA testing for between group and within group differences. A sample size calculation was performed. Based on a previously published trial of CPAP in severe cardiogenic pulmonary edema, a sample size of 21 patients per group was calculated to detect a 30% difference in intubation rates at a two-tailed alpha level of .05 and a beta level of .20 (6).
365
Figure 1. The systolic and diastolic blood pressure changes for the BiPAP and MASK groups.
were randomized to the BiPAP group and 17 patients to the MASK group. Age was 67.4 ⫾ 15 years for the BiPAP group and 68.5 ⫾ 15 years for the MASK group, p ⫽ 0.8173. The gender distribution for the BiPAP group was 52.4% women and for the MASK group was 82.4% women, p ⫽ 0.0528. Because no true difference was detected in gender distribution between groups, no further analysis was done. Four patients did not tolerate the BiPAP treatment and were switched to an ambient pressure mask. There was no statistical difference between groups in the use of nitroglycerin (p ⫽ 0.2064), morphine (p ⫽ 0.4194), or lasix (p ⫽ 0.7775). These data included prehospital and in hospital medication administration. Specifically, 33.3% of the patients in both groups received i.v. nitroglycerin. No further medication analysis was done. Systolic blood pressure, p ⫽ 0.3327 (Figure 1); diastolic blood pressure, p ⫽ 0.0826 (Figure 1); heart rate, p ⫽ 0.671 (Figure 2); respiratory rate, p ⫽ 0.4495 (Figure 3); Borg dyspnea scale score (Figure 4), p ⫽
RESULTS The study period commenced in December 1995 and finished in June 1997. During this period of time, four patients who did not subsequently demonstrate CHF on chest radiograph were initially entered into the study and were removed from data analysis, resulting in a convenience sample of 38 patients satisfying study criteria and being included in the data analysis. Twenty-one patients
Figure 2. Heart rate changes for the BiPAP and MASK groups.
366
M. A. Levitt
Figure 3. Changes in respiratory rate for the BiPAP and MASK groups. Figure 5. The pulse oximetry measurements the BiPAP and MASK groups.
0.7247; and pulse oximetry, p ⫽ 0.7427 (Figure 5) all showed no difference in change over time between groups, but did show significant change over time within groups. Arterial pH, p ⫽ 0.1954 (Figure 6); arterial CO2, p ⫽ 0.2192 (Figure 7); and arterial oxygen, p ⫽ 0.4876 (Figure 8) also showed no significant difference in change over time between groups, but did show a significant change over time within groups. Three patients died in the BiPAP group (15.0%), and three patients died in the MASK group (21.4%). No patients died during the study period. The time to death ranged from 1 day to 38 days with a median of 10 days. The mean length of hospital stay was 8.1 ⫾ 6.4 days for the MASK group and 7.3 ⫾ 8.0 days for the BiPAP group, p ⫽ 0.7446. The intubation rate was 5 (23.8%, 95% CI ⫽ 5.6 – 42.0%) in the BiPAP group vs. 7 (41.2%, 95% CI ⫽ 17.8 – 64.6%) in the MASK group. All intubations took place during the 2 h study period. The mean difference in proportions between intubation rates was 17.4% (95% CI ⫽ –12.3 to 47.0%). The AMI prevalence was four patients (19%, 95% CI ⫽ 2.2–35.8%) in the BiPAP group and five patients (29.4%, 95% CI ⫽ 7.7–51.1%) in the MASK group. The mean difference in proportions
Figure 4. Measurements in Borg Dyspnea Scale Scores for the BiPAP and MASK groups.
between groups rates was –10.4% (95% CI ⫽ –37.8 to 17%). No patients had an electrocardiogram, which was consistent with AMI on presentation. Cardiac markers were not analyzed other than to be used to determine the presence of AMI as a discharge diagnosis.
DISCUSSION This present study demonstrated no true difference in intubation rate with the addition of BiPAP in patients presenting in acute severe CHF. An improvement in arterial oxygenation and intubation rate would be consistent with previous randomized controlled trials of CPAP in cardiogenic pulmonary edema (5– 8). Perhaps of greatest interest is that there was not an increase in AMI rate found with the use of BiPAP. Since the first report of the use of what has generally been termed NPPV, the majority of research has been with the CPAP version. This device delivers a constant
Figure 6. The arterial pH changes over time for the BiPAP and MASK groups.
BiPAP In Congestive Heart Failure
Figure 7. The arterial CO2 changes for the BiPAP and MASK groups.
positive airway pressure that can be adjusted to the desired level. The chosen pressure then would be maintained during the entire respiratory cycle. Continuous positive airway pressure has been studied mostly in the treatment of sleep apnea and chronic obstructive pulmonary disease (COPD). To date there are four RCCT’s of CPAP in acute cardiogenic pulmonary edema. The first, by Rasanen et al., was published in 1985 (5). This study of 40 patients found during the first 10 min of treatment a slight increase in arterial blood oxygen partial pressure and a slight decrease in respiratory rate, systolic arterial pressure, and heart rate in the CPAP group. The ambient airway pressure group showed only a slight change in respiratory rate. The improvement in arterial blood oxygenation persisted throughout the investigation. Thirteen patients (65%) in the control group and seven patients (35%) in the CPAP group met the authors’ criteria for treatment failure during the study (p ⫽ 0.068). In 1991, Berstein et al. reported on a study of 39 patients who were randomized to receive CPAP or oxygen alone. At 30 min of treatment, the CPAP group showed greater improvements in respiratory rate, arterial CO2 tension, arterial pH, and the ratio of arterial oxygen
Figure 8. Changes in arterial oxygenation over time in the BiPAP and MASK groups.
367
tension to the fraction of inspired oxygen (6). After 24 h, however, there were no significant differences between the two treatment groups in any of these respiratory indexes. Seven (35%) of the patients who received oxygen alone, but none who received oxygen plus CPAP, required intubation and mechanical ventilation (p ⫽ 0.005). However, no significant difference was found in in-hospital mortality (oxygen alone, four of 20 patients; oxygen plus CPAP, two of 19; p ⫽ 0.36). Berstein et al. concluded that CPAP in severe cardiogenic pulmonary edema can result in early physiologic improvement and reduce the need for intubation and mechanical ventilation. They further concluded that their short-term study could not establish whether CPAP has any long-term benefit or whether a larger study would have shown a difference in mortality between the treatment groups. Lin and Chiang in 1991 and 1995 reported in two papers on an RCCT of CPAP vs. face-mask oxygen therapy in acute cardiogenic pulmonary edema (7,8). The first report included 80 patients, and the second report included a final count of 100 patients. The authors reported that at 3 h, PaO2 in the CPAP group showed a significant increase, whereas the intrapulmonary shunt and alveolar-arterial oxygen tension gradient was significantly reduced (p ⬍ 0.005). The CPAP group had significantly lower rate-pressure product and higher stroke volume index compared with the control group. The therapeutic failure rate over 6 h was 24% in the CPAP group and 50% in the control group (p ⬍ 0.01). The CPAP group had a lower incidence of tracheal intubation, at eight of 50 patients (16%) vs. the control group, 18 of 50 patients (36%), p ⬍ 0.01. There was no significant difference in short-term mortality and hospital stay between the study groups. The authors concluded that CPAP resulted in physiologic cardiovascular and pulmonary function improvement and significantly reduced the need for intubation. Mortality was not significantly decreased, but this may have been because of a small sample size. Finally, Keenan et al. reported a meta-analysis in 1997 on the effect of NPPV on patients in acute respiratory failure (11). The study found that NPPV was associated with a decreased mortality (odds ratio ⫽ 0.29; 95% CI ⫽ 0.15– 0.59) and a decreased need for endotracheal intubation (odds ratio ⫽ 0.20; 95% CI ⫽ 0.11 to 0.36). However, the study concluded that these effects were restricted to patients whose cause of acute respiratory failure was an exacerbation of COPD. The study further stated that future research is warranted to determine whether NPPV confers benefit in patients without COPD who have acute respiratory failure. More recently, a modification of the CPAP apparatus was developed and named BiPAP. The device was so named because, unlike CPAP, the pressure during inspi-
368
ration and expiration could be adjusted separately. This pressure difference would improve ventilation. Mehta et al. took the next step and designed and executed a study comparing CPAP and BiPAP in acute CHF patients (9). The study concluded that patients receiving BiPAP had significant improvements in PaCO2, pH, heart rate, breathing frequency, and dyspnea score within 30 min of treatment whereas the CPAP group showed significant improvement only in breathing frequency. The present study does not see the improvement in the above cardiopulmonary parameters, nor significant PaO2 improvement. Also in the Mehta et al. study, only one of 13 patients (7.7%) in the CPAP group and one of 14 patients (7.1%) in the BiPAP group required intubation. The 95% CI for the BiPAP group can be calculated as – 6.4 –20.6%. This BiPAP intubation rate is different from that found in the present study, which was 23.8% (95% CI ⫽ 5.6 – 42.0%). Because there is an overlap of the 95% CIs between the Mehta et al. study and the present study, the intubation rates for the BiPAP groups may or may not be the same. The Mehta et al. study was stopped prematurely because of a disproportionate occurrence of AMIs. The study found an AMI rate of 71% (n ⫽ 10) in the BiPAP group and 31% (n ⫽ 4) in the CPAP group. The 95% CI for the BiPAP group can be calculated as 46.0 –94.0%. The AMI rate in the BiPAP group for the present study was 19% (95% CI ⫽ 2.2–35.8%). It is interesting to note that not only was there a much lower prevalence of AMI with BiPAP found in the present study, but there was no overlap, by 10%, with the 95% CIs of these two studies. This implies that the AMI rate with BiPAP in the present study was less than that of the Mehta et al. study by at least 10%. It is beyond the scope of this present study to determine the reason for the difference. There is speculation that the high percentage of chest pain at 71% in the BiPAP group in the Mehta et al. study may be a reason for the high AMI rate (12). For one of the prior CPAP studies that did look at AMI rates, a similar occurrence to the present study was found between groups: 20% (N ⫽ 10) in the CPAP group and 22% (N ⫽ 11) in the control group (8). A limitation of this study is its power. The study was stopped four patients short of the projected sample size. However, the intubation rate findings in this study differed substantially from those in the Bersten et al. study used to determine sample size (6). That study found a need for intubation in seven (35%) of the patients who received oxygen alone and in none of the patients receiving oxygen plus CPAP. Based on the intubation rate findings in the present study, a sample size estimate of 90 patients was calculated per group to detect a 20% difference in intubation rates with an expected intubation rate in the BiPAP group of 25% at a two-tailed ␣ ⫽ .05 and
M. A. Levitt
 ⫽ .2. It was decided that it would not be possible to achieve a sample size this large at this institution. Therefore, the study was not restarted. The decision was made to present the findings of the present study in an attempt to develop a multicenter clinical trial. It was recognized that with the concern of an increase in AMI rate with the use of BiPAP from the Mehta et al. study, it would be difficult to justify any further such research. However, the presentation of the findings of the present study may change that opinion. A second limitation of the study was lack of standardization of adjunct treatment guidelines. The use of pharmacologic treatments was left to the discretion of the senior treating physician. This may have affected clinical outcome differences between groups. However, as was reported, no significant differences were detected between study groups with the different medications used. A third limitation is that the study patients were not consecutive. Enrolling patients only when both a member of the research team and a respiratory therapist were present does not present an obvious bias, but could affect the results. As previously mentioned, it was necessary to have personnel who were knowledgeable about the study available to assure proper data collection and to obtain consent. The decision to intubate was not influenced by the study personnel or investigator. Various ED senior physicians were present over the study period managing the patients.
CONCLUSIONS No true difference was found with the use of BiPAP vs. MASK in patients with severe acute CHF seen in the ED. Importantly, no true difference or trend toward an increase in AMI rates was found with the use of BiPAP. This study provides support for a larger clinical trial to provide a sample size with sufficient power to determine if these findings hold true.
Acknowledgments—I acknowledge Robert Feldman for his help with initiating this study and the Respiratory Therapy Department (specifically LaVerne Grant), Carolyn Celi, Matthew Waxman, and Irene Hsu for their help in performing this study. Thanks also to Darlene Jones for her assistance with manuscript preparation.
REFERENCES 1. Pingleton SK. Complications of acute respiratory failure. Am Rev Respir Dis 1988;137:1463–93. 2. Staufer JL, Olson DE, Petty TL. Complications and consequences
BiPAP In Congestive Heart Failure
3. 4. 5. 6. 7.
of endotracheal intubation and tracheotomy: a prospective study of 150 critically ill adult patients. Am J Med 1981;70:65–76. Fagon JY, Chastre J, Hance AJ, et al. Nosocomial pneumonia in ventilated patients: a cohort study evaluating attributable mortality and hospital stay. Am J Med 1993;94:281– 8. Poulton PE. Left-sided heart failure with pulmonary edema: its treatment with the “pulmonary plus pressure machine.” Lancet 1936;2:981–3. Rasanen J, Heikkila J, Downs J, et al. Continuous positive airway pressure by face mask in acute cardiogenic pulmonary edema. Am J Cardiol 1985;55:296 –300. Bersten AD, Holt AW, Vedig AE, et al. Treatment of severe cardiogenic pulmonary edema with continuous positive airway pressure delivered by face mask. N Engl J Med 1991;325:1825–30. Lin M, Chiang HT. The efficacy of early continuous positive airway pressure therapy in patients with acute cardiogenic pulmonary edema. J Formosan Med Assoc 1991;90:736 – 43.
369 8. Lin M, Yang YF, Chiang HT. Reappraisal of continuous positive airway pressure therapy in acute cardiogenic pulmonary edema. Chest 1995;107:1379 – 86. 9. Mehta S, Jay GD, Woolard, RH, et al. Randomized, prospective trial of bi-level versus continuous positive airway pressure in acute pulmonary edema. Crit Care Med 1997;25:620 – 8. 10. Wilson RC, Jones PW. A comparison of the visual analogue scale and modified Borg scale for the measurement of dyspnea during exercise. Clin Sci 1989;76:277– 82. 11. Keenan SP, Kernerman PD, Cook DJ, et al. Effect of noninvasive positive pressure ventilation on mortality in patients admitted with acute respiratory failure: a meta-analysis. Crit Care Med 1997;25: 1685–92. 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. Chest 1998;114:1185–92.