“Hemodynamic effects of nasal intermittent positive pressure ventilation versus nasal continuous positive airway pressure in preterm neonates following extubation: a randomized controlled trial”

Abstract

Background

Nasal continuous positive airway pressure (NCPAP) and nasal intermittent positive pressure ventilation (NIPPV) are two commonly used modalities of non-invasive respiratory support in neonatal intensive care units. Although both are effective, NIPPV is increasingly preferred because of its additional clinical advantages. However, the comparative hemodynamic effects of these strategies remain uncertain. This randomized controlled trial aimed to compare the hemodynamic effects of NIPPV and NCPAP in preterm infants following extubation.

Methods

In this single-center randomized controlled trial, preterm neonates (26–36^+ 6 weeks), previously intubated and treated with surfactant, were randomized to NIPPV or NCPAP following extubation. Functional echocardiography (FnECHO) was performed at 4–6 h and 18–24 h to assess right and left ventricular output (RVO, LVO) and indices of systolic and diastolic function.

Results

Sixty preterm neonates were included in the final analysis (30 in each group). Baseline RVO and LVO on invasive mechanical ventilation were comparable between the groups (p = 0.33, 0.53). Post-extubation, infants on NIPPV demonstrated significantly lower cardiac outputs than those on NCPAP at 4–6 h (RVO, p = 0.006; LVO, p = 0.001) and at 18–24 h (RVO, p < 0.001; LVO, p < 0.001). After adjustment for gestational age and birth weight, these differences remained significant. Sensitivity analysis adjusting for baseline oxygen saturation index attenuated the early differences, whereas those at 18–24 h remained significant (RVO p = 0.030; LVO p = 0.022). Despite statistical significance, all values remained within gestational age-appropriate physiological limits. Mean airway pressure was significantly higher in the NIPPV group at 4–6 h (10.35 ± 1.0 vs. 6.16 ± 1.2 cm H₂O) and 18–24 h (9.72 ± 1.3 vs. 5.71 ± 1.6 cm H₂O; both p < 0.001) and showed a significant inverse correlation with cardiac outputs.

Conclusion

Neonates supported with NIPPV in the post-extubation period demonstrated lower cardiac outputs compared with those on NCPAP, and this was associated with higher mean airway pressures. Although within physiological limits, these changes warrant hemodynamic monitoring and further evaluation of their clinical impact.

Clinical trial registration

This study was prospectively registered with the Clinical Trials Registry of India (CTRI/2024/04/065278) on 05 April 2024. Trial details are publicly accessible at https://ctri.nic.in/Clinicaltrials/pmaindet2.php?trialid=78,344.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12887-026-07122-9.

Introduction

Non-invasive respiratory support is frequently used in neonatal intensive care units (NICUs) as a primary and post-extubation mode of ventilation [1]. Nasal continuous positive airway pressure (NCPAP) and nasal intermittent positive pressure ventilation (NIPPV) are the most widely used modalities of non-invasive ventilation (NIV) [2]. Nasal CPAP provides a continuous distending pressure, while NIPPV superimposes intermittent peak pressures over the end-expiratory pressure [3].

While extensive evidence supports the efficacy of both NCPAP and NIPPV for preterm infants, various studies and meta-analyses have demonstrated advantages of NIPPV over NCPAP, including lower rates of respiratory failure, a reduced need for intubation, and a lower incidence of post-extubation failure, potentially driven by the higher mean airway pressures achieved with NIPPV [3–7]. Consequently, NIPPV is increasingly utilized as a preferred modality of NIV in many NICUs worldwide [8]. Both of these modalities can increase alveolar volume and intrathoracic pressure, leading to decreased venous return to the heart, subsequently affecting cardiac contractility [9]. Although the pressures utilized in NIPPV vary, the mean airway pressures delivered are typically higher than those in traditional NCPAP [10].

Despite the growing use of NIV in NICUs, there remains a paucity of data evaluating its cardiovascular implications. Studies on the hemodynamic effects of NCPAP have shown that even low levels of CPAP (5 cm H₂O) or incremental increases in continuous airway pressure levels can significantly reduce right ventricular output (RVO) and superior vena cava (SVC) flow [11, 12]. Lack of synchronization with the patient’s respiratory efforts on NIPPV may contribute to fluctuations in intrathoracic pressure and influence systemic venous return and pulmonary blood flow [13]. However, data exploring the hemodynamic effects of NIPPV in neonates are sparse, and comparative studies of NCPAP versus NIPPV in this regard remain limited by small sample sizes and short duration of exposure in controlled settings [10, 14]. In this randomized controlled trial (RCT), we aimed to compare the hemodynamic effects of NIPPV and NCPAP in preterm infants who were extubated to NIV following surfactant therapy and mechanical ventilation.

Methods

Study design and setting

This was a prospective, single-center, open-label, randomized controlled trial conducted at a tertiary care neonatal unit of a university hospital between April 2024 and March 2025. The study was approved by the Institutional Ethics Committee (BVDUMC/IEC/150), and written informed consent was obtained from one or both parents before enrolment. The trial was prospectively registered with the Clinical Trials Registry of India (CTRI) on 05 April 2024 (CTRI/2024/04/065278).

Participants

Eligible participants included preterm neonates with a gestational age between 26 and 36 + 6 weeks, who required invasive mechanical ventilation (IMV) and received at least one dose of surfactant therapy. Infants were enrolled at the time of planned extubation from mechanical ventilation to non-invasive respiratory support. Pre-specified exclusion criteria were major congenital malformations (including structural heart disease), clinical or culture-proven sepsis, hemodynamic instability or inotropic support within the preceding 24 h, perinatal asphyxia, hemodynamically significant patent ductus arteriosus (HsPDA), moderate to severe pulmonary hypertension (pulmonary artery systolic pressure ≥ 45 mmHg).

Outcomes

The primary outcome was the comparison of right and left ventricular outputs between the NCPAP and NIPPV groups, measured using functional echocardiography at 4–6 h and 18–24 h following extubation. Secondary outcomes included comparison of additional echocardiographic parameters of systolic and diastolic function - tricuspid annular plane systolic excursion (TAPSE), fractional area change (FAC), fractional shortening (FS), ejection fraction (EF), early to late diastolic flow velocity (E/A) of mitral and tricuspid valves, the ratio of early diastolic mitral velocity to early diastolic mitral annular tissue velocity (LV E/e’), ratio of early diastolic tricuspid velocity to early diastolic tricuspid annular tissue velocity (RV E/e’) and myocardial performance index (MPI), as well as clinical parameters including mean airway pressure, blood pressure, heart rate, and short-term clinical outcomes such as hemodynamic instability- need for inotropes, fluid boluses, or development of pulmonary hypertension post-extubation.

Sample size estimation

The sample size was calculated to detect a between-group difference of 55 mL/kg/min in RVO between the NCPAP and NIPPV groups, based on the magnitude and variability of RVO reported in a randomized crossover study by Chang et al., evaluating the hemodynamic effects of CPAP and NIPPV in preterm infants. Using the pooled standard deviation of 71.03 mL/kg/min reported in that study, and assuming a two-sided alpha of 0.05 with 80% power, the estimated sample size required was 26 infants per group.

Randomization and allocation

Participants were randomized to the NIPPV or NCPAP group following extubation, using a computer-generated sequence. Allocation concealment was ensured using sequentially numbered, opaque, sealed envelopes, opened at the time of extubation.

Intervention protocol

Respiratory support was delivered using the Fabian-light ventilator (ACUTRONIC Medical Systems AG, Switzerland; software version 5.2.2) with a RAM cannula (Neotech, Valencia, CA, USA) as the interface. The NCPAP was initiated at 6 cm H₂O. The NIPPV group received ventilation with peak inflation pressure (PIP) of 18–20 cm H₂O and PEEP of 6 cm H₂O, an inspiratory time of 0.4 s, and a ventilator rate of 40 breaths per minute, with synchronization achieved via the ventilator’s inbuilt pressure trigger. Inspired oxygen was adjusted in both groups to maintain preductal oxygen saturation (SpO2) between 90 and 95%. The mean airway pressure (MAP) at the time of each echocardiographic assessment was recorded as displayed on the ventilator. Ventilation settings were weaned according to the infant’s clinical status by the treating physician; however, the assigned intervention was continued for a minimum of 24 h post-extubation, unless the neonate’s clinical condition warranted escalation of respiratory support. Neonates requiring reintubation or escalation of support within 24 h were not included in the analysis, as continuation on the allocated modality was necessary for valid outcome assessment.

Echocardiographic assessment

All neonates underwent a screening echocardiogram before enrollment to exclude structural cardiac anomalies, HsPDA, or significant pulmonary hypertension, and baseline cardiac outputs were recorded. Functional echocardiography (FnECHO) was performed at two predefined time points—at 4–6 h and again at 18–24 h post-extubation—using a Philips Affiniti 50 ultrasound machine equipped with a 12–4 MHz sector array probe. Echocardiograms were conducted at the bedside by neonatal physicians trained in functional echocardiography. Each measurement was averaged over three consecutive cardiac cycles to improve accuracy. Echocardiographic measurements were conducted according to the previously published guidelines by the American Society of Echocardiography (ASE) for neonatal hemodynamics [15] by two trained investigators (PC and KK) using standardized acquisition and analysis protocols. Inter- and intra-observer variability was assessed in a subset of thirteen neonates using intraclass correlation coefficients and Bland–Altman analysis, demonstrating good to excellent reproducibility. Owing to the nature of the respiratory interventions, treating clinicians and echocardiographic investigators could not be blinded. Echocardiograms with technically inadequate image quality due to poor acoustic windows were excluded to ensure reliable measurements. Maternal and neonatal baseline characteristics and other relevant clinical data were recorded as per a preset proforma.

Statistical analysis

Continuous variables were expressed as mean ± SD or median (IQR), and categorical variables as frequencies. Between-group comparisons were performed using independent t-tests for normally distributed continuous variables and chi-square or Fisher’s exact tests for categorical variables. Paired t-tests were used to analyze within-group changes in echocardiographic parameters between the 4–6 and 18–24-hour assessments. To account for multiple comparisons across secondary outcomes, false discovery rate (FDR) adjustment was performed using the Benjamini–Hochberg procedure. Pearson correlation analysis was used to assess the association between mean airway pressure (MAP) and key hemodynamic parameters. Piecewise linear regression models were applied to identify MAP thresholds associated with significant shifts in cardiac function as assessed by echocardiography. Subgroup analysis across gestational age bands was performed using Welch ANOVA to account for heteroscedasticity, followed by Games–Howell post-hoc testing for pairwise comparisons. A linear mixed-effects model (LMM) with a random intercept for each infant was used to account for repeated measures. Gestational age and birth weight were included as covariates to adjust for baseline differences in maturity and size, ensuring that the estimated group effect reflected the independent contribution of ventilation mode. A sensitivity analysis was performed using multivariable linear regression adjusting for baseline oxygen saturation index. A p-value < 0.05 was considered statistically significant for all analyses. Data was analyzed using IBM SPSS Statistics version 29.0 (IBM Corp., Armonk, NY, USA).

Results

The consort flow chart of the study is illustrated in Fig. 1.

Fig. 1.

Consort flow diagram of study patients. (HsPDA, hemodynamically significant patent ductus arteriosus; *technically inadequate echocardiograms due to poor acoustic windows)

Six hundred and twenty-three preterm infants were screened during the study period, and 60 were randomized to the NIPPV or NCPAP group (30 neonates in each arm). Baseline maternal and neonatal characteristics were comparable between the CPAP and NIPPV groups, as shown in Table 1. In subgroup analysis, the number of infants with gestational age < 29 weeks was higher in the NIPPV group (13 vs. 5 in the NCPAP group).

Table 1.

Maternal and neonatal characteristics

Parameter	CPAP (n = 30)	NIPPV (n = 30)
Maternal characteristics n (%)
Primigravida	10 (33.3%)	13 (43.3%)
Antenatal Steroids	16 (53.3%)	18 (60.0%)
Caesarean section	23 (76.7%)	28 (93.3%)
PPROM > 18 h	6 (20.0%)	4 (13.3%)
Multi-fetal Gestation	2 (6.7%)	4 (13.3%)
IVF Conception	1 (3.3%)	2 (6.7%)
Pre-eclampsia	8 (26.7%)	10 (33.3%)
GDM	3 (10.0%)	1 (3.3%)
Neonatal characteristics
Birth Weight (g) a	1655.3 ± 528.2	1545.3 ± 631.8
Gender – Male n (%)	23 (76.7%)	18 (60.0%)
Gestational Age (weeks) a	31.60 ± 2.58	30.60 ± 3.44
APGAR 5 min b	8 [7.75–9.0]	8 [6.25–8.75]
Number of Doses of Surfactant a	1.4 ± 0.6	1.6 ± 0.6
Postnatal Day at Extubation b	3 (2–4)	4 (3–7)
Duration of Mechanical Ventilation (days) b	2 (2–2)	4 (2–7.5)

Comparisons used: t-test, Mann–Whitney U, or chi-square/Fisher’s exact test as appropriate

PPROM Preterm premature rupture of membranes, IVF in vitro fertilization, GDM Gestational diabetes mellitus, NIV Non-invasive ventilation, BPD Bronchopulmonary dysplasia

^aMean ± standard deviation (SD)

^bmedian (interquartile range, IQR)

Primary outcome

While RVO and LVO were comparable in both groups at baseline prior to extubation (p = 0.33 and p = 0.53, respectively), the neonates managed with NIPPV demonstrated lower cardiac outputs compared to those on NCPAP at 4–6 h post-extubation (RVO, p = 0.006; LVO, p = 0.001), with the difference becoming more pronounced at 18–24 h (RVO, p < 0.001; LVO, p < 0.001) (Fig. 2).

Fig. 2.

Temporal trends in mean (±SD) right ventricular output (RVO) and left ventricular output (LVO) at baseline on invasive mechanical ventilation (IMV), at 4–6 h and 18–24 h post-extubation in the CPAP and NIPPV groups. P-values indicate between-group comparisons at each time point

Given a higher percentage of extremely preterm infants in the NIPPV group, we repeated the analysis after adjusting for gestational age and birth weight. Both linear mixed-effects models (LMM) and analysis of covariance (ANCOVA) confirmed that, even after adjustment, NIPPV was associated with significantly lower cardiac outputs post-extubation compared with NCPAP. Adjusted mean differences were approximately − 23 to − 25 mL/kg/min at 4–6 h (p ≈ 0.006–0.008) and − 38 to − 40 mL/kg/min at 18–24 h (p < 0.001) for both RVO and LVO, indicating an effect independent of baseline maturity and size (Supplementary Table 1). Stratified ANCOVA by gestation and birth weight further showed numerically greater adjusted differences in the 29–32 week and 1000–1500 g strata (Supplementary Tables 2 and 3). Given the higher proportion of extremely preterm infants in the NIPPV group, a sensitivity analysis of the primary outcome was performed in the 29–32 weeks subgroup. In this gestational age–stratified analysis with birth weight as a covariate, the findings were directionally consistent with the primary analysis, with lower RVO and LVO observed in the NIPPV group at 4–6 h (RVO: p = 0.002; LVO: p = 0.001) and 18–24 h (RVO: p < 0.001; LVO: p < 0.001) post-extubation. Detailed adjusted estimates are provided in Supplementary Table 4.

Due to differences in oxygenation parameters between the two groups at baseline and post-extubation (Supplementary Table 5), a post-hoc sensitivity analysis using multivariable linear regression adjusting for baseline oxygenation saturation index was performed (Table 2). After adjustment, between-group differences in cardiac outputs at 4–6 h were attenuated and no longer statistically significant (RVO p = 0.35; LVO p = 0.30). At 18–24 h, cardiac outputs remained significantly lower in the NIPPV group (RVO p = 0.030; LVO p = 0.022).

Table 2.

Post-hoc sensitivity analysis of cardiac output adjusted for baseline oxygenation saturation index

Outcome	Adjusted Mean Difference (NIPPV − CPAP)	95% Confidence Interval	p value
RVO 4–6 h	−9.71	−30.28 to 10.87	0.35
LVO 4–6 h	−9.90	−28.86 to 9.07	0.30
RVO 18–24 h	−22.39	−42.52 to − 2.25	0.030
LVO 18–24 h	−22.19	−41.06 to − 3.32	0.022

Values represent adjusted mean differences derived from post hoc multivariable ordinary least squares linear regression models assessing the association between ventilatory mode and cardiac output after adjustment for baseline oxygen saturation index. Negative values indicate lower cardiac output in the NIPPV group

Secondary outcomes

Table 3 summarizes the secondary outcomes. At both 4–6 hours and 18–24 hours post-extubation, measures of ventricular systolic and diastolic function, including TAPSE, FAC, FS, EF, and E/A of mitral and tricuspid valves, were significantly lower in the NIPPV group. The LV E/e’ was higher in the NIPPV group, while the RV E/e’ remained comparable. The mean airway pressures were significantly higher in the NIPPV group. The mean blood pressure was significantly lower in the NIPPV group, although remaining within normative limits for gestational age.

Table 3.

Cardiorespiratory and echocardiographic parameters in NCPAP vs. NIPPV groups at 4–6 h and 18–24 h post-extubation

Parameter	4–6 h			18–24 h
	CPAP (n = 30)	NIPPV (n = 30)	p-value	CPAP (n = 30)	NIPPV (n = 30)	p-value
Clinical parameters (Mean ± SD)
BP Mean (mmHg)	48.57 ± 3.4	45.37 ± 3.8	0.001	47.07 ± 2.9	42.8 ± 3.9	< 0.001
Heart Rate (beats/min)	149.77 ± 3.6	149.13 ± 5.1	0.580	151.80 ± 3.7	152.50 ± 6.5	0.610
MAP (cm H₂O)	6.16 ± 1.2	10.35 ± 1.0	< 0.001	5.71 ± 1.6	9.72 ± 1.3	< 0.001
Secondary outcome (Other echocardiographic parameters) (Mean ± SD)
RV MPI	0.32 ± 0.05	0.34 ± 0.06	0.120	0.33 ± 0.05	0.37 ± 0.06	0.002
LV MPI	0.33 ± 0.06	0.35 ± 0.05	0.277	0.34 ± 0.06	0.36 ± 0.05	0.157
TAPSE (mm)	0.72 ± 0.12	0.63 ± 0.14	0.010	0.70 ± 0.11	0.59 ± 0.14	0.001
FAC (%)	37.79 ± 5.4	33.56 ± 5.2	0.003	36.66 ± 5.19	31.32 ± 5.39	< 0.001
FS (%)	33.72 ± 3.19	31.83 ± 3.24	0.030	32.50 ± 2.96	29.72 ± 3.53	0.002
EF (%)	66.56 ± 6.3	62.05 ± 6.0	0.006	64.65 ± 5.85	58.44 ± 6.15	< 0.001
LV E: A Ratio	0.88 ± 0.06	0.84 ± 0.06	0.002	0.87 ± 0.05	0.80 ± 0.06	< 0.001
RV E: A Ratio	0.83 ± 0.06	0.79 ± 0.05	0.002	0.82 ± 0.05	0.75 ± 0.06	< 0.001
RV E/e′	7.46 ± 0.75	7.58 ± 0.92	0.580	7.60 ± 0.73	8.02 ± 0.97	0.060
LV E/e′	8.00 ± 0.45	8.46 ± 0.43	< 0.001	8.18 ± 0.45	8.91 ± 0.57	< 0.001

Values are presented as mean ± standard deviation. Comparisons between groups at each time point were performed using independent sample t-tests

BP Blood pressure, MAP Mean airway pressure, RVO Right ventricular output, LVO Left ventricular output, MPI Myocardial performance index, TAPSE Tricuspid annular plane systolic excursion, FAC Fractional area change, FS Fractional shortening, EF Ejection fraction, LV Left ventricle, RV Right ventricle, E A ratio, early to late diastolic flow velocity of mitral and tricuspid valves, LV E/e′ ratio of early diastolic mitral velocity to early diastolic mitral annular tissue velocity, RV E/e′ ratio of early diastolic tricuspid velocity to early diastolic tricuspid annular tissue velocity

To account for multiple comparisons across secondary outcomes, a false discovery rate (FDR) adjustment using the Benjamini–Hochberg procedure was performed (Supplementary Table 6). After adjustment, most echocardiographic parameters—including systolic (TAPSE, FAC, EF) and diastolic indices (mitral and tricuspid E/A ratios, LV E/e′)—differed between groups, particularly at 18–24 h, with directions consistent with the primary analysis. However, the difference in fractional shortening at 4–6 h was no longer statistically significant.

Mean airway pressure across time points

Baseline (pre-extubation, on IMV) mean airway pressure (MAP) was comparable between the NCPAP and NIPPV groups (8.47 ± 1.30 vs. 8.58 ± 1.24 cmH₂O; p = 0.72). Following extubation, MAP increased from baseline in the NIPPV group and was higher compared to the NCPAP group at both 4–6 h and 18–24 h (10.35 ± 1.0 vs. 6.16 ± 1.2 cmH₂O and 9.72 ± 1.3 vs. 5.71 ± 1.6 cmH₂O, respectively; both p < 0.001), remaining elevated compared to baseline at 18–24 h, despite a modest reduction from 4 to 6 h (Supplementary Table 7).

Relationship between mean airway pressure and cardiac outputs in the NIPPV group

Pearson correlation analysis demonstrated no significant association between MAP and cardiac outputs at baseline on invasive ventilation. In contrast, post-extubation, MAP showed a statistically significant inverse correlation with both right RVO and LVO at 4–6 h and 18–24 h (Supplementary Table 8). The strength of this negative correlation increased over time, as illustrated by the scatter plots with regression lines in Fig. 3.

Fig. 3.

Correlation of mean airway pressure (MAP) with cardiac outputs in neonates at baseline on invasive mechanical ventilation (IMV), 4–6 h, and 18–24 h post-extubation on NIPPV. Blue dots – individual neonates; red line – regression slope; shaded area – 95% CI. Narrower CIs at higher MAP indicate stronger negative correlations over time

Mean airway pressure thresholds

Piecewise linear regression with 1,000 bootstrap resamples was applied to estimate MAP breakpoints and the corresponding slopes before and after each cutoff at 4–6 and 18–24 h. Median (IQR) values from the bootstrap distributions are reported in Supplementary Table 9. The median MAP thresholds for both LVO and RVO were clustered between 10.3 and 10.6 cmH₂O, beyond which cardiac outputs declined more steeply, an effect more pronounced at 18–24 h (Supplementary Fig. 1).

Subgroup analysis by gestational age

Cardiac outputs were comparable across gestational age (GA) groups at baseline but differed at 4–6 h and 18–24 h post-extubation, with the lowest RVO and LVO in infants < 29 weeks in both NCPAP and NIPPV groups. The decline from baseline was greater with NIPPV, most pronounced at 18–24 h, and infants 29–32 weeks also had lower outputs than those > 32 weeks (Supplementary Tables 10,11).

Within-group changes in cardiac output

Paired t-test analyses demonstrated distinct temporal patterns of cardiac outputs between NCPAP and NIPPV. In the NCPAP group, RVO and LVO were higher at 4–6 h compared with baseline and remained elevated at 18–24 h. In the NIPPV group, RVO and LVO were unchanged between baseline and 4–6 h, but both declined significantly by 18–24 h. (Supplementary Table 12)

Short-term clinical outcomes

None of the neonates in either group developed immediate post-extubation hemodynamic instability, including hypotension requiring inotropes or fluid boluses, or persistent pulmonary hypertension within 24 h. Infants in the NIPPV group required a longer duration of non-invasive ventilation compared to the NCPAP group (13.5 [6.8–35.3] vs. 5 [3–9] days; p = 0.004).

Discussion

In this single-center randomized controlled trial, preterm neonates extubated after surfactant therapy for RDS were randomized to nasal CPAP or NIPPV. Right and left ventricular outputs were significantly lower in the NIPPV group at 4–6 h, with a further decline at 18–24 h, and correlated inversely with mean airway pressure. These differences persisted after adjustment for gestational age and birth weight. Adjustment for baseline oxygen saturation index attenuated early differences, whereas lower cardiac outputs in the NIPPV group remained significant at 18–24 h, suggesting a potential contribution of baseline respiratory status to early findings. Mean airway pressures were consistently higher with NIPPV. Although these differences reached statistical significance, the cardiac outputs remained within gestational age-specific physiological limits and were not accompanied by clinical signs of hemodynamic instability, suggesting subclinical yet measurable effects of sustained higher MAP with NIPPV.

While optimal lung recruitment is a desirable goal of any form of respiratory support, the right balance must be achieved between ventilation and its cardiovascular effects. Compared with conventional NCPAP, NIPPV typically delivers higher mean airway pressures [9]. Preterm infants are particularly vulnerable to the hemodynamic impacts of standard respiratory care interventions [9]. An increase in alveolar volume and high intrathoracic pressure on NIV can potentially impede venous return to the heart and increase pulmonary vascular resistance, in turn affecting both right and left ventricular function [9, 13].

Several studies have assessed the hemodynamic effects of PEEP in preterm infants on NCPAP with variable results. Beker et al. reported no significant changes in cardiac output or systemic blood flow with 15-minute exposures of CPAP levels of 4–8 cm H₂O in stable preterm infants [16]. In contrast, Abdel-Hady et al. observed that even 5 cm H₂O of NCPAP reduced RVO and SVC flow in preterm infants recovering from RDS, without affecting blood pressure or vital signs [11]. Mukerji et al., in a dose-escalation study of NCPAP up to 13 cm H₂O, found no change in cardiac output, although each CPAP level was applied for only ~ 10 min, reflecting very short duration exposures [17]. De Waal et al. examined the effects of PEEP increments in ventilated neonates and found a significant reduction in RVO with an increase in PEEP from 5 to 8 cmH₂O, without short-term effects on systemic venous return [12].

We found two studies that compared the hemodynamic effects of NIPPV and NCPAP. Chang and colleagues, in a randomized crossover study of 20 stable preterm infants, compared NCPAP with NIPPV over 30-minute intervals and found no significant differences in RVO, LVO, SVC flow, or cerebral hemodynamics between the two groups. ¹⁴ However, their methodology and patient population differed substantially from ours. While Chang et al. studied older, stable preterm infants (median postnatal age ~ 20 days), our study focused on preterm neonates (median age 3.5 days) recovering from RDS and recently extubated. The post-extubation phase was selected in our study, as it reflects RDS recovery, when respiratory support requirements are more stable than immediately after birth. Thus, while Chang et al. performed a short-term physiological comparison in stable neonates under experimental conditions, our randomized parallel-group RCT evaluated uninterrupted use of each modality in a routine clinical setting, incorporating a broader range of echocardiographic assessment beyond cardiac output and SVC flow.

In a similar physiological crossover study, Mukerji et al. compared the hemodynamic effects of high CPAP (≥ 9 cm H₂O) and NIPPV at equivalent MAP in 15 stable preterm infants and found no difference in the cardiac outputs [10]. The duration of exposure to each modality in their study was only 30 min, as opposed to a more sustained exposure in our study. Their study population consisted of neonates more than 2 weeks of postnatal age and more than 28 weeks of postmenstrual age, while we included a wider range of preterm neonates recovering from RDS in their first week of life. Another major difference in methodology was the use of equivalent MAP in both groups, whereas in routine practice, the mean airway pressures are often substantially higher with NIPPV, as demonstrated in our study. Thus, in contrast to brief, experimental comparisons in stable neonates, our study evaluated the cardiopulmonary effects of NIV in a high-risk, post-extubation population over a longer and clinically relevant timeframe.

Our study demonstrated a negative correlation between MAP and ventricular output. In piglet models, Mirro et al. showed that stepwise increases in MAP (5–20 cm H₂O) resulted in a progressive fall in cardiac output and organ perfusion. The arterial pressures, however, fell only at 20 cm H₂O, indicating that positive pressure ventilation can impair cardiovascular function before hypotension develops [18]. When contextualized with prior research, our findings contribute new insight into how sustained MAP levels may influence neonatal hemodynamics.

We examined mean airway pressure trends. Mean airway pressure increased in the NIPPV group following extubation and despite a modest decline over time, remained above the pre-extubation baseline at 18–24 h, suggesting that sustained exposure to higher airway pressures may have contributed to the observed reduction in cardiac output. At higher mean airway pressures, NIV may reduce right ventricular preload, increase pulmonary vascular resistance, and decrease right ventricular output, thereby reducing pulmonary blood flow and left ventricular preload, with additional influence from ventricular interdependence. Non-invasive ventilation provides positive end-expiratory pressure and increased mean airway pressure while allowing spontaneous breathing [13]. Although mean airway pressure appears to be a key contributor, other mechanisms may also play a role. As lung compliance improves in preterm infants recovering from respiratory distress syndrome, a greater proportion of airway pressure may be transmitted to the mediastinum, potentially further reducing cardiac output [19].

We also found differences in the echocardiographic measures of systolic and diastolic function between the two groups. Systolic functions (EF, FS, FAC, and TAPSE) and diastolic parameters (E: A) were lower in the NIPPV group, while the elevated LV E/e′ suggested increased filling pressures. Abdel-Hady et al. similarly reported reduced left atrial diameter and venous return with CPAP [11]. These differences became more pronounced at 18–24 h and, although within normative limits, may reflect subtle shifts in ventricular compliance from sustained elevated MAP and higher intrathoracic pressure [13]. After correction for multiple comparisons, most parameters still showed between-group differences at this time point, although differences in fractional shortening at 4–6 h was no longer significant. However, these findings should be interpreted cautiously as secondary and exploratory.

Subgroup analyses across gestational strata demonstrated a maturational gradient, with the lowest cardiac outputs in infants < 29 weeks, consistent with their physiologic immaturity. Despite their higher representation in the NIPPV arm, baseline cardiac outputs were comparable, and adjusted analyses controlling for gestational age and birth weight showed lower cardiac outputs with NIPPV. A sensitivity analysis restricted to infants 29–32 weeks showed directionally similar findings. However, the subgroup analysis was limited by reduced sample size, and these findings should be interpreted with caution as exploratory. While moderately preterm infants appeared to show a more pronounced decline in cardiac outputs with NIPPV, differences in the most immature infants were less distinct, possibly reflecting a floor effect related to their inherently low cardiac outputs.

The elevated MAP in the NIPPV group was associated with modestly lower systemic blood pressures persisting at 18–24 h, though within gestational norms, likely reflecting reduced preload from increased intrathoracic pressure. Prior studies similarly showed reduced RVO and systemic flow with higher PEEP, without hypotension, ^11,12 while others found no short-term BP changes with high CPAP and NIPPV [10]. The absence of immediate post-extubation hemodynamic instability or early pulmonary hypertension in either group suggests comparable early tolerance of NIPPV and CPAP despite differences in airway pressure delivery. The longer duration of non-invasive ventilation in the NIPPV group may primarily reflect a higher proportion of extremely preterm infants, which may have contributed to a more gradual weaning approach in this cohort; however, this cannot be definitively established from the present data.

Strengths and limitations

To our knowledge, this is the first parallel-group RCT to have comprehensively evaluated and compared the hemodynamic effects of NIPPV and nasal CPAP in preterm infants. Unlike previous crossover designs limited to brief (30-minute) exposures to NIV in older, more stable neonates, our study utilized serial functional echocardiography over 24 h post-extubation in preterm infants recovering from RDS, allowing assessment of evolving hemodynamic responses to sustained NIV exposure in a clinically relevant setting. The correlation of mean airway pressure with echocardiographic parameters and identification of MAP thresholds in our study provides novel insights into the influence of mean airway pressure on hemodynamic trends.

This study has several limitations. As the study population included preterm neonates in the immediate post-extubation period, the findings may not be generalizable to other clinical settings, such as NIV as a primary modality, or its application for durations exceeding 24 h, or its use in term infants. Neonates requiring reintubation or escalation within 24 h were excluded to enable valid serial assessments on the allocated modality; however, this may have introduced selection bias toward clinically stable infants, thereby limiting generalizability. Since all enrolled neonates had been previously mechanically ventilated and received surfactant therapy, the residual impact of these interventions on hemodynamics remains uncertain. There was a higher proportion of extremely preterm infants in the NIPPV group. Higher oxygenation indices in the NIPPV group suggest greater respiratory support requirements, which may have influenced the observed differences in cardiac output. While additional analyses adjusting for baseline oxygenation saturation index yielded consistent findings, such covariate-adjusted models cannot fully eliminate residual confounding and therefore should be considered supportive to the randomized comparison. Blinding of echocardiographic investigators was not feasible given the nature of the interventions, introducing a potential source of observer bias. Although measurements were protocolized, averaged over three cardiac cycles, and inter- and intra-observer variability were assessed to enhance reliability, operator-dependent measurement bias cannot be excluded and remains a potential limitation of such assessments. Due to its ease of application, better tolerance, and reduced nasal trauma, the RAM cannula is used as the routine non-invasive interface in our unit. Emerging randomized controlled studies support its effectiveness for non-invasive respiratory support, including NIPPV, in preterm infants [20–22]. However, the reported MAP values represent ventilator-displayed pressures via RAM cannula and may not accurately reflect actual airway pressures delivered, as effective transmission may be influenced by leak, cannula fit, and airway dynamics. At the same time, the uniform use of the same interface and ventilator across the cohort allows comparison of relative differences in displayed MAP and associated trends in cardiac output within the study context. We did not include additional advanced echocardiographic parameters such as strain imaging or speckle tracking, which offer load-independent assessment of cardiac function and could have provided complementary insights. Lastly, the study was limited to evaluating short-term physiological and clinical effects in the immediate post-extubation period; further studies are needed to assess the impact on longer-term outcomes, including duration of hospitalization and neurodevelopment.

Conclusion

In our study, we compared the hemodynamic effects of NIPPV and NCPAP in the post-extubation period among preterm infants recovering from RDS and found that NIPPV was associated with significantly higher mean airway pressures, accompanied by subtle but measurable alterations in cardiac performance on functional echocardiography. Cardiac outputs were lower with NIPPV and correlated inversely with MAP, with effects becoming more pronounced after sustained exposure; this association persisted after adjustment for gestational age and birth weight. While early differences were attenuated after accounting for baseline oxygenation parameters, lower cardiac outputs were still observed at later assessment. Despite statistical significance, all values remained within gestational age-appropriate physiological norms, without affecting vital signs, suggesting subclinical effects. Nevertheless, our findings highlight the need for hemodynamic monitoring when using higher mean airway pressures on non-invasive ventilation and warrant further studies to assess how these hemodynamic changes influence long-term clinical and neurodevelopmental outcomes.

Abbreviations

ANCOVA

Analysis of covariance

ASE

American Society of Echocardiography

BP

Blood pressure

CI

Confidence interval

CPAP

Continuous positive airway pressure

CTRI

Clinical Trials Registry of India

EF

Ejection fraction

FAC

Fractional area change

FnECHO

Functional echocardiography

FS

Fractional shortening

GA

Gestational age

HsPDA

Hemodynamically significant patent ductus arteriosus

HR

Heart rate

IMV

Invasive mechanical ventilation

IQR

Interquartile range

LMM

Linear mixed—effects model

LVO

Left ventricular output

MAP

Mean airway pressure

MPI

Myocardial performance index

NCPAP

Nasal continuous positive airway pressure

NIPPV

Nasal intermittent positive pressure ventilation

NIV

Non—invasive ventilation

PEEP

Positive end—expiratory pressure

PIP

Peak inspiratory pressure

RDS

Respiratory distress syndrome

RCT

Randomized controlled trial

RV

Right ventricle / right ventricular

RVO

Right ventricular output

SD

Standard deviation

SpO₂

Oxygen saturation measured by pulse oximetry

SVC

Superior vena cava

Authors’ contributions

P.S., S.D., and P.V.C. conceptualized and designed the study. P.V.C. and K.K contributed to data collection and acquisition, and P.V.C drafted the initial manuscript. S.D., S.L., and P.V.C contributed to analysis, interpretation, and verification of data. A.K., and A.S. contributed to investigation and data curation. S.D., P.S., R.N., and S.L. reviewed the manuscript. All authors reviewed and approved the final version and agree to be accountable for all aspects of the work.

Funding

No funding or sponsorship was received for this study.

Data availability

De-identified participant data and statistical code are not publicly available due to ethical and institutional restrictions but may be obtained from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study was initiated after approval from the Institutional Ethics Committee of Bharati Vidyapeeth Deemed University (approval number: BVDUMC/IEC/150) and was prospectively registered with the Clinical Trials Registry of India (CTRI/2024/04/065278). Written informed consent was obtained from the legal guardians of all enrolled participants prior to study inclusion. The study was conducted in accordance with the principles of the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.