Hemodynamic Effects of Noninvasive Positive-Pressure Ventilation Assessed Using Transthoracic Echocardiography.

AIMS: The aim of this study is to measure the effect of positive-pressure ventilation on heart chamber dimensions, left ventricular (LV) systolic function, LV diastolic function, right ventricular (RV) systolic function, and RV pressure using transthoracic echocardiography. SETTINGS AND
DESIGN: This is a prospective study in a single secondary health-care center.
MATERIALS AND METHODS: A total of 107 patients with obstructive sleep apnea on continuous positive airway pressure (CPAP) therapy were recruited as participants between April and September 2016. Transthoracic echocardiography was performed twice on each participant, before and 15 min after, they used their own CPAP machines, and the echocardiography parameters of both scans were compared. STATISTICAL ANALYSIS USED: The parametric paired t-test was used to compare heart chamber dimensions, left heart diastolic function, left heart systolic function, right heart systolic function, and right heart pressure effect, without and with CPAP. These data were further examined among several subgroups defined by CPAP when the cutoff point was set at 8 cmH2O and 10 cmH2O. The level of significance was set at 0.05. Statistical analyses were performed using IBM SPSS version 22 (IBM, Armonk, NY, USA).
RESULTS: There were statistically significant reductions, after the application of CPAP, in the heart dimensions, and LV and RV systolic function. There were no significant changes in diastolic function. Concerning right heart pressure, with CPAP, there was a significant increase in the inferior vena cava (IVC) diameter and there was also a significant decrease in IVC variability from 44.56% ± 14.86% to 36.12% ± 11.42%. The maximum velocity of tricuspid regurgitation (TR) decreased significantly from 180.66 ± 6.95 cm/s to 142.30 ± 52.73 cm/s. Such changes were observed in both low and high CPAP subgroups.
CONCLUSIONS: When placed on positive pressure, the clinically significant change in IVC diameter and variability and change in trans-TR velocity mean that it would be inaccurate to predict right heart chamber pressure through echocardiogram. Alternative methods for predicting right heart pressure are recommended.

INTRODUCTION

Positive-pressure ventilation (PPV) has wide clinical use. Its application ranges from acute situations such as acute pulmonary edema to more chronic situations such as end-stage chronic obstructive pulmonary disease. It can be applied invasively through endotracheal intubation or noninvasively through face masks. It can also be used in ambulatory settings, like in persons with obstructive sleep apnea (OSA).

While several studies have reported the good clinical outcomes of PPV in various diseases, the exact immediate hemodynamic effect of PPV has not been well reported. Most relevant studies involved invasive mechanical ventilation with controlled tidal volume, rather than noninvasive ventilation. Moreover, echocardiography has been more commonly used in managing patients on PPV. It seems necessary to ascertain whether echocardiography parameters can be accurately used to assess true hemodynamic changes in patients placed on PPV. Furthermore, blood pressure decrease is not uncommon with the initiation of PPV, especially in individuals with marginal cardiac function. Although multifactorial, these changes must at least be partly influenced by PPV. Studying the magnitude of such hemodynamic effects would ensure safer application of PPV.

The expected physiological hemodynamic effects of PPV are as follows:[12] the preload of the heart decreases because of the decrease in venous return while the afterload depends on the transmural pressure, which is the difference between the intraventricular and pleural pressure. The pleural pressure is increased by the PPV, so the transmural pressure is reduced and the afterload decreases.[34]

Concerning the hemodynamics of the right ventricle, the afterload of the right ventricle increases with an increase in pulmonary vascular resistance and this is due to the compression of pulmonary vessels by positive alveolar pressure.[5] This, coupled with the decreased preload of the right ventricle and thus a smaller right ventricular (RV) end-diastolic volume (RVEDV), should result in a decrease in RV stroke volume (SV).

The central venous pressure (CVP) increases as 25% of the positive end-expiratory pressure (PEEP) is transmitted to the central veins,[6] and the CVP increases linearly with PEEP.[7] While the transtricuspid valve pressure gradient is commonly used to predict the RV systolic pressure (RVSP), its change during PPV is difficult to predict. The blood flow through the right heart decreases, and the tricuspid regurgitation (TR) flow or pressure, if any, should decrease. However, the transtricuspid gradient depends on the pressure difference between the right atrium (RA) and right ventricle. When the TR (if any) decreases, the RA pressure decreases, and thus, the transtricuspid gradient should increase. The effect of PPV on RV systolic function would also be difficult to predict, and previous studies have shown inconsistent results.[89101112]

Concerning the hemodynamics of the left ventricle, the compliance of the left ventricle decreases as the high-volume lung compresses the heart. The diastolic function of the left ventricle should worsen. Johnson et al., however, reported no change in diastolic function and filling pressures by continuous positive airway pressure (CPAP) in a group of patients with OSA and heart failure.[13] The effect on the left ventricular (LV) end-diastolic volume (LVEDV) is difficult to predict, and it may not parallel the change in RVEDV because of the effect of ventricular interdependence. The effect of LV SV and other LV systolic functions, which depend on the degree of the effect of PPV on LV preload, afterload, and compliance, remains difficult to predict. Results from previous publications have shown inconclusive results.[1415161718]

As the hemodynamic effect of PPV is very complex, whether these predicted changes would have significant clinical or at least echocardiography (echo)-measurable effects that affect patient management need to be studied. Moreover, with these possible hemodynamic changes, it needs to be ascertained whether the interpretation of echo parameters should be different for patients undergoing PPV. This study aimed to clarify these issues by measuring the effect of PPV on the heart chamber dimensions, LV systolic function, LV diastolic function, RV systolic function, and RV pressure through transthoracic echo (TTE).

MATERIALS AND METHODS

Design and subjects

This study was approved by the Joint CUHK-NTEC Clinical Research Ethics Committee (reference no.: 2016.107). Informed consent was obtained from all participants.

This was a prospective cross-sectional single-center study performed in Alice Ho Miu Ling Nethersole Hospital, Hong Kong SAR, between April and September 2016. Using the hospital registry, 528 patients with the diagnosis of sleep apnea were first identified. Those not receiving CPAP therapy, with central sleep apnea rather than OSA, and those with moderate-to-severe respiratory or cardiac conditions (including congenital heart disease, moderate-to-severe valvular pathologies, and mechanical heart valves) that could potentially affect echo parameters were excluded from the study. The remaining 253 patients were identified and contacted; 118 patients consented to participate in the study. Eleven were further excluded after the initial echo assessment as their results could affect the generalizability of the outcome: 2 with mechanical heart valves and 9 with poor echo views. Finally, 107 patients with OSA using long-term CPAP were recruited.

These patients came with their own CPAP machines. They were assessed twice through TTE: without the CPAP machines, i.e., under atmospheric pressure and with their CPAP machines on, using their usual CPAP levels. The echo machine was a Philips IE33 (serial number 039 × 32) using the standard 3.5 Hz probe. Standard views and measurements were obtained according to guidelines from the American Society of Echocardiography.[19] Echo parameters, both with and without CPAP, were collected for data analysis. All echo scans were performed by a single operator to avoid interobserver variability. All the echo outcomes were reviewed and validated by an echocardiologist.

Data collection

For each participant, the following data were collected once without CPAP and repeated on the same participant with CPAP:

Baseline demographics included age, sex, body weight, height, body mass index (BMI), presence of atrial fibrillation (AF) or LV hypertrophy (LVH), apnea–hypopnea index (AHI), and CPAP pressure.

Heart chamber dimensions included LV end-diastolic diameter (LVEDD), LV end-systolic diameter (LVESD), LVEDV, LV end-systolic volume (LVESV), left atrium volume (LAvol), RA area (RAarea) obtained from a four-chamber view at end diastole, inferior vena cava (IVC) diameter, and IVC variability.

Left heart systolic function included SV, ejection fraction (EF) by Simpson biplane, and velocity time integral of the LV outflow tract (LVOTVTI).

Left heart diastolic function included the ratio of peak early filling velocity (E) through a transmitral Doppler image obtained from an apical four-chamber view to the corresponding peak atrial velocity (A) (E/A ratio) and the ratio of E to the early diastolic medial mitral annular velocity (E'), as measured by tissue Doppler imaging at the septal mitral annulus (E/E').

Right heart systolic function included tricuspid annular peak systolic excursion (TAPSE) and the velocity time integral over the RV outflow tract (RVOTVTI).

Right heart pressure effect included the maximum velocity of TR (TRVmax).

Statistical analysis

Continuous data, including baseline characteristics and echo parameters, are expressed as a mean ± standard deviation.

The parametric paired t-test was used to compare heart chamber dimensions (LVEDD, LVESD, LVEDV, LVESV, LAvol, RAarea, maximum inferior vena cava diameter (IVCmax), minimum diameter of inferior vena cava (IVCmin), and IVC variability), left heart diastolic function (E/A ratio and E/e' ratio), left heart systolic function (SV, EF, and LVOTVTI), right heart systolic function (TAPSE and RVOTVTI), and right heart pressure effect (TRVmax), without and with CPAP. These data were further examined among several subgroups defined by CPAP when the cutoff point was set at 8 cmH2O and 10 cmH2O.

The level of significance was set at 0.05. Statistical analyses were performed using IBM SPSS version 22 (IBM, Armonk, NY, USA).

RESULTS

Baseline demographics

We included 107 patients for data analysis [Table 1] (mean age: 59.66 ± 9.79 years, 85% male). Patients' BMI was 29.79 ± 4.58 kg/m2. The AHI was 13.91 ± 22.01; 75% of the patients had minimal-to-mild OSA severity according to their AHI. The baseline EF was 58.97% ±7.37%. The CPAP used was 7.84 ± 3.37 cmH2O. Almost 42.1% of the patients had LVH and 6.5% had AF.

Table 1

Baseline demographics of the patients (n=107)

Parameter (unit)	Results
Age (years)	59.66±9.79 (22-89)
Sex (n)
Male	91
Female	16
AHI	13.91±22.01 (0.2-84.7)
Body weight (kg)	81.81±14.21 (54.5-134.8)
Body height (m)	1.66±0.08 (1.41-1.81)
BMI (kg/m2)	29.79±4.58 (21.72-47.76)
Baseline EF (%)	58.97±7.37 (40.0-72.0)
CPAP (cmH2O)	7.84±3.37 (3.7-20.0)

Continuous data are expressed as a mean±SD (range). AHI=Apnea-hypopnea index, BMI=Body mass index, EF=Ejection fraction, CPAP=Continuous positive airway pressure, SD=Standard deviation

Heart chamber dimensions [Table S1]

There was a statistically significant reduction in LVEDV by 4.15 ± 2.51 mL (P = 0.001) after CPAP [Table 2]. Other LV dimensions, including LVESD, LVEDD, and LVESV, did not demonstrate significant changes. Both RAarea and LAvol decreased. RAarea decreased by 1.05 ± 2.81 cm2 (P < 0.001) and LAvol decreased by 3.76 ± 13.88 mL (P = 0.006) after CPAP application.

Table 2

Differences in the heart chamber dimension measurements of the patients before and after continuous positive airway pressure therapy (n=107)

Parameter (unit)	Paired differences (baseline minus CPAP)				P
	Mean±SD	SEM	95% CI
			Lower	Upper
LVEDD (cm)	0.065±0.646	0.062	−0.059	0.189	0.299
LVESD (cm)	−0.065±0.606	0.059	−0.181	0.051	0.271
LVEDV (mL)	4.15±12.508	1.209	1.757	6.551	0.001
LVESV (mL)	0.57±6.423	0.621	−0.666	1.796	0.365
LAvol (mL)	3.76±13.878	1.342	1.098	6.419	0.006
RAarea (cm2)	1.05±2.812	0.272	0.516	1.594	<0.001
IVCmax (cm)	−0.44±0.388	0.038	−0.519	−0.371	<0.001
IVCmin (cm)	−0.42±0.334	0.0322	−0.485	−0.358	<0.001
IVC variability (%)	8.44±17.662	1.707	5.053	11.823	<0.001

The shaded boxes indicate those with statistically significant results (P<0.05). LVEDD=Left ventricular end-diastolic diameter, LVESD=Left ventricular end-systolic diameter, LVEDV=Left ventricular end-diastolic volume, LVESV=Left ventricular end-systolic volume, LAvol=Left atrial volume, RAarea=Right atrial area, IVCmax=Maximum inferior vena cava diameter, IVCmin=Minimum diameter of inferior vena cava, IVC variability=Inferior vena cava variability, SD=Standard deviation, SEM=Standard error of the mean, CI=Confidence interval, CPAP=Continuous positive airway pressure

There was a statistically significant increase in IVC diameter and decrease in IVC variability with CPAP use [Table 2]. IVCmax increased by 0.44 ± 0.39 cm (P < 0.001) and IVCmin increased by 0.42 ± 0.33 cm (P < 0.001) after CPAP application. IVC variability decreased by 8.44% ± 17.66% (P < 0.001) after CPAP use.

Left heart systolic function [Table S2]

There was a statistically significant decrease in LV systolic function [Table 3]. SV decreased by 5.44 ± 13.22 mL (P < 0.001), EF decreased by 1.94% ±7.48% (P = 0.008), and LVOTVTI decreased by 1.64 ± 4.14 cm (P < 0.001) after CPAP use.

Table 3

Differences in the left heart systolic function measurements of the patients before and after continuous positive airway pressure therapy

Parameter (unit)	Paired differences (baseline minus CPAP)				P
	Mean±SD	SEM	95% CI
			Lower	Upper
SV (mL)	5.44±13.218	1.284	2.890	7.981	<0.001
EF (%)	1.94±7.475	0.723	0.507	3.372	0.008
LVOTVTI (cm)	1.64±4.141	0.402	0.840	2.435	<0.001

The shaded boxes indicate those with statistically significant results (P<0.05). SV=Stroke volume, EF=Ejection fraction, LVOTVTI=Velocity time integral of the left ventricular outflow tract, SD=Standard deviation, SEM=Standard error of the mean, CI=Confidence interval, CPAP=Continuous positive airway pressure

Left heart diastolic function [Table S3]

There was no statistically significant change in the LV diastolic function after CPAP use [Table S4].

Table S4

Differences in the left heart diastolic function measurements of the patients before and after continuous positive airway pressure therapy

Parameter	Paired differences(baseline minus CPAP)				P
	Mean±SD	SEM	95% CI
			Lower	Upper
E/A ratio	0.26±0.296	0.030	−0.033	0.085	0.391
E/E’ ratio	0.44±3.976	0.388	−0.333	1.206	0.263

E/A ratio=Ratio of the peak early filling velocity to the corresponding peak atrial velocity, E/E' ratio=Ratio of the peak early filling velocity to the early diastolic medial mitral annular velocity, SD=Standard deviation, SEM=Standard error of the mean, CPAP=Continuous positive airway pressure, CI=Confidence interval

Right heart systolic function [Table S5]

There was a statistically significant decrease in right heart systolic function after CPAP use [Table 4]. TAPSE decreased by 0.17 ± 0.48 cm (P = 0.001) and RVOTVTI decreased by 1.92 ± 3.30 cm (P < 0.001) after CPAP use.

Table 4

Differences in the right heart systolic function measurements of the patients before and after continuous positive airway pressure therapy

Parameter (unit)	Paired differences (baseline minus CPAP)				P
	Mean±SD	SEM	95% CI
			Lower	Upper
TAPSE (cm)	0.17±0.476	0.047	0.075	0.260	0.001
RVOTVTI (cm)	1.92±3.304	0.321	1.287	2.560	<0.001

The shaded boxes indicate those with statistically significant results (P<0.05). TAPSE=Tricuspid annular peak systolic excursion, RVOTVTI=Velocity time integral over the right ventricular outflow tract, SD=Standard deviation, SEM=Standard error of the mean, CPAP=Continuous positive airway pressure, CI=Confidence interval

Right heart pressure effect [Table S6]

There was a statistically significant decrease in TRVmax after CPAP use [Table 5].

Table 5

Differences in the measurements reflecting right heart pressure effect in the patients before and after continuous positive airway pressure therapy

Parameter (unit)	Paired differences (baseline minus CPAP)				P
	Mean±SD	SEM	95% CI
			Lower	Upper
TRVmax (cm/s)	38.36±39.059	3.776	30.873	45.846	<0.001

The shaded boxes indicate those with statistically significant results (P<0.05). TRVmax=Maximum velocity of tricuspid regurgitation, SD=Standard deviation, SEM=Standard error of the mean, CPAP=Continuous positive airway pressure, CI=Confidence interval

Subgroup analyses

Subgroup 1: CPAP <8 cmH2O (n = 59) versus CPAP ≥8 cmH2O (n = 48) [Table S7]

Table S7

Right heart pressure effect before and after continuous positive airway pressure therapy in patients with continuous positive airway pressure <8 cmH2O (n=59) and heart2O (n=48)

CPAP (cmH2O)	Parameter (unit)	Timing	Mean±SD	SEM
<8	TRVmax(cm/s)	Baseline	178.96±64.464	8.393
		CPAP	143.73±53.121	6.916
≥8	TRVmax(cm/s)	Baseline	182.74±56.936	8.218
		CPAP	140.54±52.751	7.614

CPAP=Continuous positive airway pressure, TRVmax=Maximum velocity of tricuspid regurgitation, SD=Standard deviation, SEM=Standard error of the mean

There was a statistically significant decrease in TRVmax in both the high (decrease of 42.20 ± 40.07 cm/s, P < 0.001) and low (decrease of 35.24 ± 38.28 cm/s, P < 0.001) CPAP groups [Table S8].

Table S8

Differences in the measurements reflecting right heart pressure effect before and after continuous positive airway pressure therapy in patients with continuous positive airway pressure therapy <8 cmH2O (n=59) and ≥8 cmH2O (n=48)

CPAP (cmH2O)	Parameter (unit)	Paired differences(baseline minus CPAP)				P
		Mean±SD	SEM	95% CI
				Lower	Upper
<8	TRVmax(cm/s)	35.24±38.277	4.983	25.262	45.212	<0.001
≥8	TRVmax(cm/s)	42.20±40.068	5.783	30.564	53.832	<0.001

The shaded boxes indicate those with statistically significant results (P<0.05). CPAP=Continuous positive airway pressure, TRVmax=Maximum velocity of tricuspid regurgitation, SD=Standard deviation, SEM=Standard error of the mean, CI=Confidence interval

Subgroup 2: CPAP <10 cmH2O (n = 73) versus CPAP ≥10 cmH2O (n = 34) [Table S9].

Table S9

Right heart pressure effect before and after continuous positive airway pressure therapy in patients with continuous positive airway pressure <10 cmH2O (n=73) and ≥10 cmH2O (n=34)

CPAP (cmH2O)	Parameter (unit)	Timing	Mean±SD	SEM
<10	TRVmax(cm/s)	Baseline	177.63±62.501	7.315
		CPAP	142.79±51.530	6.031
≥10	TRVmax(cm/s)	Baseline	187.15±57.841	9.920
		CPAP	141.24±55.997	9.604

CPAP=Continuous positive airway pressure, TRVmax=Maximum velocity of tricuspid regurgitation, SD=Standard deviation, SEM=Standard error of the mean

There was a statistically significant decrease in TRVmax in the high (decrease of 45.91 ± 43.04 cm/s, P < 0.001) and low (decrease of 34.84 ± 36.85 cm/s, P < 0.001) CPAP groups [Table S10].

Table S10

Differences in the measurements reflecting right heart pressure effect before and after continuous positive airway pressure in patients with continuous positive airway pressure therapy <10 cmH2O (n=73) and ≥10 cmH2O (n=34)

CPAP (cmH2O)	Parameter (unit)	Paired differences(baseline minus CPAP)				P
		Mean±SD	SEM	95% CI
				Lower	Upper
<10	TRVmax(cm/s)	34.84±36.846	4.313	26.245	43.439	<0.001
≥10	TRVmax(cm/s)	45.91±43.037	7.381	30.896	60.928	<0.001

The shaded boxes indicate those with statistically significant results (P<0.05). CPAP=Continuous positive airway pressure, TRVmax=Maximum velocity of tricuspid regurgitation, SD=Standard deviation, SEM=Standard error of the mean, CI=Confidence interval

DISCUSSION

This study successfully demonstrated measurable hemodynamic effects of PPV on echo.

The significant change in IVC diameter and IVC variability meant that it would be inaccurate to use these echo parameters to predict RA pressure in patients on CPAP, and alternative methods to predict RA pressure are recommended.

The estimated transtricuspid pressure gradient (by Bernoulli equation) in patients on CPAP may be lower than the original pressure gradient without CPAP. With the reduction in TRVmax and the PPV-related inappropriateness of IVC collapsibility in RA pressure assessment, alternative methods such as right heart catheterization are recommended for RV pressure measurement in patients placed on PPV.

Our results showed a significant reduction in LV systolic function, RV systolic function, and heart chamber dimensions. They may seem of little clinical significance, but their cumulative negative effects on heart function through PPV could be crucial, especially for those with marginal heart function. The effect of CPAP was shown to be different in various patient subgroups. Li et al. reported that PEEP decreased cardiac output and cardiac index in patients with poor cardiac function but not in persons with normal cardiac function.[20] However, CPAP is primarily indicated for individuals with acute cardiogenic pulmonary edema, who are at the highest end of the Starling curve. This is because of its beneficial effect on preload reduction, with an improvement in transmural pressure and a consequent reduction in afterload.[21] Acute pulmonary edema in left heart failure decreases lung compliance and therefore required greater respiratory effort, which in turn induces more negative intrathoracic pressure. This causes a further increase in preload and afterload and a greater demand on the already failing LV. CPAP interrupts this cycle. Thus, despite the potentially negative effect on the hemodynamics, CPAP is beneficial in persons with left heart failure causing acute pulmonary edema. Jellema et al. demonstrated the safety of PPV in patients with septic shock,[22] but they usually have a baseline decrease in left heart systolic function and systemic vascular resistance. In practice, it is common to see an initial adverse hemodynamic change, for example, a decrease in blood pressure with CPAP initiation. Furthermore, the CPAPs used for patients with OSA are generally lower than the CPAPs used in acute settings. A high PEEP is commonly applied in the critically ill. Thus, the actual hemodynamic effects in these patients by CPAP might be even greater than those shown in this study.[2324]

Patients with OSA on CPAP were recruited as participants in this study. However, OSA affects heart function. In this study, 42.6% of the patients had LVH, 6.5% had AF, and all patients had a certain degree of TR. This baseline abnormal cardiac function might have affected some echo parameters. To allow better generalizability, patients with moderate-to-severe respiratory or cardiac diseases that could potentially affect the echo parameters were excluded from this study.

RVSP is commonly estimated using the Bernoulli formula, 4 × TRVmax2 + estimated RA pressure. RA pressure can be estimated by the IVC diameter and IVC respiratory variability. This study confirmed that in patients under PPV, TRVmax, IVCmax diameter, and IVC variability were all significantly altered. The maximum IVC diameter increased by 4.4 mm and IVC variability decreased by 8.44%. This implies that the IVC parameters were no longer good predictors of RA pressure. It is also suggested in the American Society of Echocardiography guideline that IVC variability in patients on PPV should not be used to assess RA pressure.[19] In patients on PPV, the positive pressure transmitted to the RA reduces venous return. This results in an inversion of the cyclic changes in IVC diameter and in a respiratory variation in IVC diameter exceeding 12% that should only be observed in volume-depleted patients. Actually, this 12% IVC variability index in persons on PPV is used to predict fluid responsiveness but is limited to individuals receiving mandatory ventilator breathing with at least 8 mL/kg of tidal volume.[19] This result, however, suggested that the inaccuracy of using IVC variability to predict RA pressure was not limited to the group on invasive mechanical ventilation but also applied to those with spontaneous breaths on CPAP alone without controlled tidal volume. TRVmax decreased by 38.36 cm/s with CPAP. These effects were observed across the low and high CPAP groups. Thus, caution was needed in estimating the transtricuspid pressure gradient. Alternative methods for estimating RVSP are thus recommended.

This study has several strengths. It was prospective, and paired data before and after CPAP use were obtained from the same participants to avoid confounding factors. Measures were also taken to enhance the generalizability of the outcomes, such as excluding patients with moderate-to-severe cardiac or respiratory diseases. Furthermore, all echo data were collected by a single operator to ensure consistency, and all echo images and recordings were subsequently reviewed by an echocardiologist to guarantee accuracy.

There are also several limitations to this study. First, this was a single-center study and all participants were Chinese. Moreover, 85% of these were men. Most were overweight or obese. All these factors limit the generalizability of the results. Second, this was an unblinded study. The unblinded echo operator could have been biased in interpreting echo parameters. Third, measurement errors were possible, especially with several difficult echo views in this group of patients with a high BMI. Potentially inaccurate measurements were recognized as missing data and nine participants were excluded due to poor echo views. Fourth, only a few parameters were selected to assess a particular heart function. Extra caution is especially needed when interpreting the LV diastolic function using E/A and E/E' ratio alone. These two parameters were selected for the easy acquisition and consistency of data. The change in E/A and change in diastolic function do not follow a linear relationship while the change in E/E', representing the change in filling pressure, is largely dependent on the preload. A more comprehensive assessment of the various parameters of diastolic function would be needed before it can be concluded that the diastolic function remained unchanged with the application of CPAP. Fifth, the echo data related to the use of CPAP were taken immediately after CPAP was applied. It was uncertain how much time was needed to develop maximal physiological changes secondary to acute CPAP application. Finally, the dose–response relationship could not be demonstrated by subgroup analyses with such a study design. Paired echo parameters with the application of both low and high CPAP on the same participant would better reflect the dose–response relationship.

CONCLUSIONS

There were echo-measurable hemodynamic effects after the application of CPAP, and these were largely consistent with physiological predictions. Despite statistical significance, the magnitude of change in the echo parameters with the application of CPAP seemed too small to yield important clinical implications. Yet, this could suggest the safety of PPV in relatively healthier participants. Further studies, however, may be required to evaluate whether these effects would be exaggerated in persons with marginal baseline heart function.

When placed on positive pressure, the clinically significant increase in IVC diameter and the decrease in its variability meant that it would be inaccurate to predict right atrial pressure through echocardiography. The change in transtricuspid regurgitation velocity meant that the estimation of the RVSP by Bernoulli's equation would be inaccurate. Alternative methods for predicting right heart pressure are recommended. Moreover, this effect was first demonstrated to be present in participants put on noninvasive position ventilation, and this is not limited to those put on invasive mechanical ventilation with controlled tidal volume.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Table S1

Heart chamber dimensions of the patients before and after continuous positive airway pressure therapy (n=107)

Parameter(unit)	Timing	Mean±SD	SEM
LVEDD(cm)	Baseline	4.43±0.642	0.062
	CPAP	4.36±0.751	0.073
LVESD(cm)	Baseline	2.75±0.757	0.073
	CPAP	2.82±0.779	0.075
LVEDV(mL)	Baseline	60.60±16.811	1.625
	CPAP	56.44±17.788	1.720
LVESV(mL)	Baseline	25.03±9.423	0.911
	CPAP	24.47±10.181	0.984
LAvol(mL)	Baseline	47.45±23.656	2.287
	CPAP	43.69±22.330	2.159
RAarea(cm2)	Baseline	14.53±3.995	0.386
	CPAP	13.48±3.723	0.360
IVCmax(cm)	Baseline	1.46±0.409	0.039
	CPAP	1.91±0.431	0.042
IVCmin(cm)	Baseline	0.80±0.283	0.027
	CPAP	1.22±0.371	0.036
IVC variability(%)	Baseline	44.56±14.849	1.436
	CPAP	36.12±11.423	1.104

LVEDD=Left ventricular end-diastolic diameter, LVESD=Left ventricular end-systolic diameter, LVEDV=Left ventricular end-diastolic volume, LVESV=Left ventricular end-systolic volume, LAvol=Left atrial volume, RAarea=Right atrial area, IVCmax=Maximum inferior vena cava diameter, IVCmin=Minimum diameter of inferior vena cava, IVC variability=Inferior vena cava variability, SD=Standard deviation, SEM=Standard error of the mean

Table S2

Left heart systolic function of the patients before and after continuous positive airway pressure therapy

Parameter (unit)	Timing	Number of patients	Mean±SD	SEM
SV(mL)	Baseline	106	71.54±16.689	1.621
	CPAP	106	66.10±14.131	1.372
EF(%)	Baseline	107	58.97±7.371	0.713
	CPAP	107	57.03±7.806	0.755
LVOTVTI(cm)	Baseline	106	21.91±4.758	0.462
	CPAP	106	20.28±3.869	0.376

SV=Stroke volume, EF=Ejection fraction, LVOTVTI=Velocity time integral of the left ventricular outflow tract, SD=Standard deviation, SEM=Standard error of the mean, CPAP=Continuous positive airway pressure

Table S3

Left heart diastolic function of the patients before and after continuous positive airway pressure therapy

Parameter	Timing	Number of patients	Mean±SD	SEM
E/A ratio	Baseline	99	0.99±0.294	0.030
	CPAP	99	0.96±0.283	0.028
E/E’ ratio	Baseline	105	12.36±3.541	0.346
	CPAP	105	11.92±4.250	0.415

E/A ratio=Ratio of the peak early filling velocity to the corresponding peak atrial velocity, E/E' ratio=Ratio of the peak early filling velocity to the early diastolic medial mitral annular velocity, SD=Standard deviation, SEM=Standard error of the mean, CPAP=Continuous positive airway pressure

Table S5

Right heart systolic function of the patients before and after continuous positive airway pressure therapy

Parameter (unit)	Timing	Number of patients	Mean±SD	SEM
TAPSE(cm)	Baseline	104	2.43±0.488	0.048
	CPAP	104	2.26±0.419	0.041
RVOTVTI(cm)	Baseline	106	18.46±4.220	0.410
	CPAP	106	16.53±4.190	0.407

TAPSE=Tricuspid annular peak systolic excursion, RVOTVTI=Velocity time integral over the right ventricular outflow tract, SD=Standard deviation, SEM=Standard error of the mean, CPAP=Continuous positive airway pressure

Table S6

Right heart pressure effect in the patients before and after continuous positive airway pressure therapy (n=107)

Parameter (unit)	Timing	Mean±SD	SEM
TRVmax(cm/s)	Baseline	180.66±60.949	5.892
	CPAP	142.30±52.729	5.098

TRVmax=Maximum velocity of tricuspid regurgitation, SD=Standard deviation, SEM=Standard error of the mean, CPAP=Continuous positive airway pressure