(Associate Editor: Amanda Piper).
Acute haemodynamic effects of continuous positive airway pressure in awake patients with heart failure
Version of Record online: 23 DEC 2013
© 2013 The Authors. Respirology © 2013 Asian Pacific Society of Respirology
Volume 19, Issue 1, pages 47–52, January 2014
How to Cite
Schroll, S., Sériès, F., Lewis, K., Benjamin, A., Escourrou, P., Luigart, R., Pfeifer, M. and Arzt, M. (2014), Acute haemodynamic effects of continuous positive airway pressure in awake patients with heart failure. Respirology, 19: 47–52. doi: 10.1111/resp.12193
Conflict of interest statement: M.A. has previously received lecture fees from Astra Zeneca, Covidien, Philips Home Healthcare Solutions (Murrysville, PA, USA) and Resmed (Martinsried, Germany), and received grant support from Philips Home Healthcare Solutions. He is also the holder of an endowed professorship from the Free State of Bavaria at the University of Regensburg, donated by Philips Home Healthcare Solutions and Resmed. K.L. has received grant support from Pzier Inc., and lecture fees and reimbursement for attending conferences from Pfizer, Astra Zeneca and Glaxo-Smith Kline.
- Issue online: 23 DEC 2013
- Version of Record online: 23 DEC 2013
- Accepted manuscript online: 19 SEP 2013 08:47PM EST
- Manuscript Accepted: 1 AUG 2013
- Manuscript Received: 23 JUL 2013
- Manuscript Revised: 23 JUL 2013
- Manuscript Revised: 18 JUL 2013
- Manuscript Revised: 8 JUL 2013
- Manuscript Revised: 7 JUL 2013
- blood pressure;
- heart failure;
- heart rate;
- positive airway pressure;
- sleep apnoea
Background and objective
Continuous positive airway pressure (CPAP) has been used to treat patients with chronic heart failure (CHF) and sleep-disordered breathing (SDB). CPAP treatment in severe CHF with concomitant SDB and atrial fibrillation has been linked to impairment of cardiac output (CO) as a potential cause for adverse outcome. The aim of the present study was to test whether incremental CPAP application in awake CHF patients with SDB, with and without atrial fibrillation, induces acute alterations of blood pressure (BP), heart rate (HR) and CO.
During daytime, we applied incremental CPAP (4–10 cmH2O) in 37 stable patients with CHF and SDB. BP and HR were assessed after each 1 cmH2O CPAP increase in 5-min intervals in the entire sample, and CO was assessed at one centre (n = 11).
Neither mean BP, HR nor CO changed significantly with incremental CPAP (at 0 and 10 cmH2O: 85 ± 2 and 84 ± 2 mm Hg, P = 1.0, 63 ± 1 to 61 ± 2 b.p.m., P = 0.88 and 2.03 ± 0.5 and 2.35 ± 0.8 L/min/m2, P = 0.92, respectively). No significant differences in maximum BP drop or HR drop between patients with sinus rhythm and atrial fibrillation were found. In 1 of 37 patients, a prespecified event of haemodynamic compromise (drop of mean BP >15 mm Hg) without clinical signs occurred.
These results contribute to the evidence that CPAP does not cause haemodynamic compromise in the vast majority of normotensive CHF patients with SDB.
chronic heart failure
continuous positive airway pressure
positive airway pressure
In patients with chronic heart failure (CHF), sleep-disordered breathing (SDB) is highly prevalent (51–71%).[1-3] CHF patients are commonly affected by both types of SDB; obstructive sleep apnoea caused by upper airway obstruction during sleep and central sleep apnoea, which is often associated with a Cheyne–Stokes breathing pattern and caused by instability of the respiratory control system. The clinical significance of SDB, both central sleep apnoea and obstructive sleep apnoea, in CHF patients is twofold. First, apnoeas and hypopnoeas lead to repetitive arousals and impaired sleep quality. Treatment of SDB in CHF patients with SDB-related daytime symptoms such as hypersomnolence can lead to reversal of such symptoms and improvement of quality of life.[4, 5] Second, by exposing the failing heart to intermittent hypoxia, increased cardiac preload and after load leading to sympathetic nervous system activation, SDB may contribute to the progression of heart failure, and the presence of SDB independently confers an increased risk for mortality.[6, 7] SDB especially obstructive sleep apnoea in CHF patients can be treated by continuous positive airway pressure (CPAP). Thus, current guidelines for the management of CHF recommend to consider treatment of obstructive sleep apnoea, and awareness to diagnose and treat SDB in patients with CHF is rising.
However, cardiovascular effects of CPAP therapy and the safety of such treatments are still under some debate. A large randomized controlled trial demonstrated that CPAP modestly improved left ventricular ejection fraction and catecholamine levels but did not improve transplant-free survival. In particular, within the first 18 months after enrolment transplant-free survival was significantly better in the control group compared with the CPAP group. This finding raised concerns about whether CPAP treatment of patients with CHF and SDB is safe. One possibility could be that positive airway pressure (PAP) can cause acute haemodynamic compromise: a previous study showed that short-term application of CPAP in awake CHF patients with atrial fibrillation reduces cardiac output (CO). In addition, there are conflicting results of CPAP on the acute effects on CO in awake CHF patients. In CHF patients with increased pulmonary capillary wedge pressures, both an increase and a decrease in CO by CPAP application were observed.[11-14] In CHF patients with normal pulmonary wedge pressure, application of CPAP consistently causes a drop of CO.[12-14] In a study of 61 CHF patients with SDB, a significant drop in blood pressure (BP) without change of heart rate (HR) was observed when fixed PAP (CPAP, bi-level PAP or adaptive servoventilation) was applied during the daytime. Previous testing of the acute effects of CPAP on haemodynamics in CHF patients with SDB is limited by either low sample size[10, 11, 13, 16, 17] or an uncontrolled study design using one fixed PAP level.[11, 12, 15, 16] The aim of the present study was to test the acute effects of incremental CPAP on BP, HR, CO and clinical signs of haemodynamic compromise in stable awake CHF patients with SDB.
Candidates for participation in the trial included men and women aged 18–80 years, with CHF (New York Heart Association Class II or III) due to ischaemic, non-ischaemic or hypertensive cardiomyopathy, a left ventricular ejection fraction ≤40% calculated according to a modification of the Simpson's method (end-diastolic minus end-systolic volume divided by end-diastolic volume), stable clinical status and stable optimal medical therapy according to the guidelines of the European Society of Cardiology with no medication changes for at least 4 weeks, and an apnoea–hypopnoea index ≥20/h of sleep assessed by in-laboratory polysomnography. Patients were excluded if they had unstable angina, myocardial infarction, heart failure due to valvular heart disease or rhythm-related heart failure, cardiac surgery or hospital admissions within the previous 3 months, if they had contraindications for PAP therapy, were using oxygen therapy, had severe restrictive (forced vital capacity of <50% predicted) or obstructive airways disease (forced expiratory volume in 1 s of <40% predicted), or daytime symptoms of SDB that required immediate treatment (e.g. sleepiness while driving). Subjects gave written informed consent and the protocol was approved by the local ethics boards of each participating centre. The present study was part of the standardized safety procedures before initiation of PAP therapy in the registered trial (http://www.controlled-trials.com; ISRCTN04353156).
Polysomnography was performed using the core equipment available at each centre. Patients lay in the supine position with one pillow. Airflow and thoraco-abdominal effort and airflow were recorded quantitatively by respiratory inductance plethysmography and nasal pressure cannula. Arterial oxyhaemoglobin saturation was assessed by pulse oximetry. A blinded analysis of each sleep study was centralized and performed by two experienced sleep technicians. The absence of airflow ≥10 s (measured reduction of airflow to less than 10% peak ‘nominal’ airflow) were classified as apnoeas. Hypopnoeas were defined as a ≥50% reduction in airflow from baseline for 10 s, or a discernable reduction in airflow if it was in association with a 4% oxygen desaturation or an arousal. Apnoeas or hypopnoeas were obstructive classified if out of phase thoraco-abdominal motion or airflow limitations were present. Mixed apnoeas were classified as central throughout the study. The apnoea–hypopnoea index was calculated as the number of apnoeas and hypopnoeas per hour of sleep.
Protocol and intervention: CPAP test
After diagnostic polysomnography, baseline BP and HR were measured every 5 min for 15 min starting early afternoon (Fig. 1). After baseline analysis, patients were titrated on CPAP starting from 4 cmH2O. Every 5 min, BP and HR reading was taken, and CPAP was increased by 1 cmH2O up to the maximum CPAP of 10 cmH2O. The CPAP titration was stopped if mean BP was <60 mm Hg or a drop of >15 mm Hg occurred or the patient did not tolerate CPAP pressure. Mainly nasal masks were used (76%).
BP and HR measurement
BP was measured automatically according to current guidelines when patients were lying supine. Dependent on patient's upper arm circumference, the appropriate cuff size was chosen. The patients were instructed to keep the recording arm at rest during each cuff inflation. HR was recorded by a five lead electrocardiogram.
Non-invasive measurement of CO
In a subset of 11 patients (all patients from one centre), CO was measured non-invasively by impedance cardiography (Task Force Monitor, CNSystems, Graz, Austria) at baseline and every time CPAP was increased by 1 cmH2O starting from 4 cmH2O up to the maximum CPAP of 10 cmH2O.
Data are given as mean ± standard deviation unless otherwise stated. Statistical comparisons among BP, HR and CO on the different CPAP levels were calculated using repeated measures analysis of variance, followed by multiple planned comparisons with a modified Bonferroni's correction based on treatment degrees of freedom. A P-value of <0.05 was considered to be statistical significant. Statistical analyses were performed using IBM SPSS Statistics Version 20 (SPSS, Chicago, IL, USA).
A total of 194 patients were screened, and the majority were excluded due to apnoea–hypopnoea index <20/h of sleep (n = 74) or left ventricular ejection fraction >40% (n = 26) and 22 patients for other reasons (Fig. 2). Of the 72 eligible patients, 37 patients were randomized to PAP therapy (the remainder to continue optimal medical therapy) according to the trial protocol (http://www.controlled-trials.com; ISRCTN04353156). Table 1 describes the 37 CHF patients. They were middle aged, predominantly men with a moderate degree of heart failure and with severe SDB with 40% central events.
|Age (years)||64 ± 10|
|Gender male, n (%)||34 (92)|
|BMI (kg/m2)||28.9 ± 4.3|
|REM sleep (%)||49 ± 22|
|Slow-wave sleep (%)||1 ± 3|
|Total sleep time (min)||341 ± 66|
|Sleep efficiency (%)||73 ± 1|
|AHI (events/h)||49 ± 17|
|Central AHI (events/h)||20 ± 13|
|Central sleep apnoea, n (%)||18 (49%)|
|Mean SaO2 (%)||93.7 ± 2.1|
|Epworth Sleepiness Scale Score||7.5 ± 4.4|
|NYHA-class II, n (%)||28 (76)|
|NYHA-class III, n (%)||9 (24)|
|Cause of heart failure|
|Ischaemic, n (%)||19 (51)|
|Non-ischaemic, n (%)||18 (49)|
|Rhythm and pacing|
|Atrial fibrillation, n (%)||10 (27)|
|Bi-ventricular pacemaker, n (%)||3 (8)|
|Implanted cardiac defibrillator, n (%)||16 (43)|
|Diuretic, n (%)||24 (65)|
|Spironlactone, n (%)||18 (49)|
|ACE-inhibitor, n (%)||27 (73)|
|AT-receptor blocker, n (%)||12 (32)|
|Beta-receptor blocker, n (%)||29 (78)|
When CPAP was stepwise increased from 0 to 10 cmH2O, neither systolic, diastolic or mean BP nor HR changed significantly (P = 0.98, P = 0.99, P = 1.00 and P = 0.87; Table 2 and Fig. 3). Also, CO in a subset of 11 patients did not change significantly with increasing CPAP (P = 0.92; Table 2). Figure 4 shows the individual data of mean BP and maximum BP drop during daytime CPAP titration according to whether they were in sinus rhythm or atrial fibrillation. No significant differences in maximum BP drop, and HR drop between patients with sinus rhythm and atrial fibrillation were found (4.1 ± 4 vs 3.4 ± 3 mm Hg; 4.3 ± 3 vs 4.5 ± 4 b.p.m., P = 0.44 and P = 0.11, respectively).
|Variable||CPAP (cmH2O)||P for trend|
|Systolic BP||117 ± 2 (82, 151)||114 ± 3 (80, 147)||114 ± 2 (83, 141)||114 ± 2 (82, 142)||114 ± 2 (86, 142)||114 ± 2 (86, 153)||114 ± 2 (86, 149)||114 ± 2 (85, 144)||0.988|
|Diastolic BP||70 ± 2 (48, 95)||69 ± 2 (48, 97)||69 ± 2 (48, 94)||70 ± 2 (47, 99)||69 ± 2 (53, 98)||69 ± 2 (51, 102)||71 ± 2 (51, 99)||70 ± 2 (51, 100)||0.999|
|Mean BP||85 ± 2 (64, 110)||85 ± 2 (60, 111)||84 ± 2 (60, 110)||84 ± 2 (60, 112)||84 ± 2 (65, 113)||84 ± 2 (63, 115)||85 ± 2 (63, 114)||84 ± 2 (63, 115)||1.000|
|Heart rate||63 ± 1 (46, 81)||63 ± 1 (47, 77)||62 ± 1 (44, 82)||63 ± 1 (44, 100)||61 ± 1 (44, 83)||61 ± 1 (44, 77)||61 ± 1 (44, 80)||61 ± 2 (45, 84)||0.877|
|Cardiac Indexb||2.0 ± 0.5 (1.2, 2.5)||2.2 ± 0.5 (1.3, 2.9)||2.1 ± 0.5 (1.2, 2.7)||2.2 ± 0.6 (1.2, 3.2)||2.3 ± 0.7 (1.3, 4.0)||2.2 ± 0.7 (1.2, 3.7)||2.3 ± 0.9 (1.2, 4.4)||2.3 ± 0.8 (1.2, 4.0)||0.920|
Thirty-six of the 37 patients underwent the full protocol, achieving a maximum CPAP of 10 cmH2O. In one patient, CPAP titration was stopped at 7 cmH2O due to a drop of mean BP by 16 mm Hg to a mean BP of 76 mm Hg. However, this drop was not accompanied by symptoms of haemodynamic compromise or of discomfort (Fig. 4).
This study shows that in awake, stable CHF patients with severe SDB, CPAP is well tolerated and the application of incremental CPAP up to 10 cmH2O does not lead to clinically significant changes in systolic, diastolic and mean BP, HR and CO. The vast majority of patients (97%) had some decrease of mean BP; however, these changes were not accompanied by symptoms of haemodynamic compromise or subjective discomfort and were not considered clinically significant. Only one patient had an asymptomatic drop of BP of 16 mm Hg with a minimum mean BP in the normal range.
Different effects of CPAP treatment on haemodynamics were observed in previous studies. Two small studies describe a decrease of systolic BP,[10, 17] one study an increase of systolic BP or no effect on systolic BP,[16, 20, 21] whereas diastolic BP and mean BP were not altered during CPAP treatment.[10, 11, 16, 17] In previous studies, HR was either not altered[10, 11, 16] or a reduction of HR during CPAP treatment was observed.[17, 20, 22]
These inconsistent observations might be because the acute haemodynamic response to CPAP in CHF patients may depend on cardiac rhythm and cardiac preload and after load: One study found a significant fall of CO and systolic BP in CHF patients with atrial fibrillation compared with CHF patients with sinus rhythm during CPAP treatment. In our study, effects on BP, HR and CO by CPAP treatment were similar in CHF patients with and without atrial fibrillation. However, the study of Kiely et al. and the current study are likely underpowered to determine whether cardiac rhythm influences the effect of CPAP on haemodynamics.
Short-term application of CPAP in CHF patients with normal pulmonary capillary wedge pressures generally provokes a decrease of CO, whereas CPAP in CHF patients with high left ventricular filling pressures augments CO.[12, 13, 20] This is probably why CPAP has been used successfully in acute left ventricular failure where filling pressures from pulmonary oedema are high. In the present study, we cannot provide data on cardiac-filling pressures, but in a subset of patients, we could show that non-invasively measured CO was not significantly changed with increasing CPAP. Other explanations for the diverse haemodynamic response could be differences in the severity of heart failure according to New York Heart Association classification and variations in baseline BP.
In a recent study of 61 CHF patients with SDB, a drop of mean BP to <70 mm Hg at the beginning of PAP therapy was observed in 10% of the study population. In this study, PAP therapy was initiated using estimated and fixed values for pressure support and not incremental values to evaluate dose–response relationships. The incremental titration of CPAP to achieve 10 cmH2O in our study was higher than the fixed CPAP or expiratory PAP (5.8 ± 0.9 cmH2O) in the study from Oldenburg and colleagues. Different to our study, Oldenburg et al. defined haemodynamic adverse events by a sustained systolic BP <100 mm Hg, a decrease in systolic BP by ≥40 mm Hg or a mean BP ≤70 mm Hg. Due to the different cut-off values used and the different modalities of PAP application, the comparison of our results with those from Oldenburg et al. is limited. Our study complements previous work by simultaneously asking about symptoms during any change in BP or HR.
One limitation of our study is that we did not systemically rule out subacute pulmonary oedema by measurement of cardiac-filling pressures or chest X-rays. Furthermore, our study represents the acute effects of CPAP on BP and HR during wakefulness, and findings should be extrapolated to more long-term effects during sleep in such patients with caution.[22, 23] Because of the low numbers of patients and the absence of a control group in our study, larger controlled studies are necessary to confirm our results.
In conclusion, daytime application of incremental CPAP did not lead to clinically significant changes in systolic, diastolic or mean BP, HR or CO in the majority of CHF patients with SDB. The presence or absence of atrial fibrillation does not alter the effect of CPAP on BP, HR or CO in CHF patients. These results contribute to the evidence that CPAP does not cause haemodynamic compromise in the vast majority of normotensive CHF patients with SDB. The results cannot be extrapolated to CHF patients with hypotension. Nevertheless, awareness of haemodynamic compromise is appropriate, when CPAP therapy is initiated in such patients.
- 8ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 2008: the Task Force for the diagnosis and treatment of acute and chronic heart failure 2008 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association of the ESC (HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM). Eur. J. Heart Fail. 2008; 10: 933–989., , et al.