Abstract
Objective To test the hypothesis that non-invasive high frequency oscillatory ventilation (NHFOV) is more efficacious than nasal continuous positive airway pressure (NCPAP) in reducing invasive mechanical ventilation as primary respiratory support for extremely preterm infants with respiratory distress syndrome.
Design A multicentre, randomised controlled trial.
Setting Twenty tertiary neonatal intensive care units in China.
Participants 342 extremely preterm infants (gestational age between 24 weeks +0 day and 28 weeks +6 days) with respiratory distress syndrome were enrolled in the study between August 2022 and August 2024.
Interventions Participants were randomly allocated to receive NCPAP or NHFOV as primary respiratory support for respiratory distress syndrome.
Main outcome measures The primary outcome was treatment failure, defined as the need for invasive mechanical ventilation within 72 hours after birth.
Results Treatment failure within 72 hours occurred in 27 of` 170 infants (15.9%) in the NHFOV group and 48 of 172 infants (27.9%) in the NCPAP group (risk difference −12.0 percentage points, 95% confidence interval −20.7 to −3.4; P=0.007). Treatment failure within seven days was also lower in the NHFOV group (−12.5 percentage points, 95% confidence interval −21.9 to −3.2; P=0.008) compared with the NCPAP group. All observed associations remained significant after sensitivity analysis including study sites and antenatal steroid use. No significant differences were found in any other secondary outcomes between the two groups.
Conclusions NHFOV appeared superior to NCPAP in reducing the need for intubation when used as a primary respiratory support strategy in extremely preterm infants. Both techniques did not show significant differences in neonatal adverse events.
Introduction
Respiratory distress syndrome is a leading cause of respiratory failure in extremely preterm infants (≤28 weeks of gestational age).1 Tracheal intubation and invasive mechanical ventilation have significantly improved the survival rates of these infants.2 However, prolonged invasive mechanical ventilation is associated with an increased risk of ventilator induced lung injury, which can lead to bronchopulmonary dysplasia.3 As a result, European consensus guidelines recommend nasal continuous positive airway pressure (NCPAP) as the preferred first line respiratory support strategy for respiratory distress syndrome.4 However, cumulative evidence has suggested that up to 40% of extremely preterm infants are prone to a high risk of NCPAP failure because of chest wall collapse and poor diaphragmatic strength.56
Non-invasive high frequency oscillatory ventilation (NHFOV) applies a continuous distending pressure with active oscillations superimposed on spontaneous tidal breathing, enabling carbon dioxide removal and alveolar recruitment without requiring synchronisation.7 Although surfactant and NCPAP remain first line treatments for respiratory distress syndrome, emerging evidence shows that early hypercapnia occurs in a significant minority of these infants, particularly those with chest wall immaturity or delayed surfactant administration.8 NHFOV may provide dual physiological advantages over NCPAP by maintaining lung recruitment while actively clearing carbon dioxide through oscillatory gas mixing—a mechanism particularly beneficial for non-uniformly diseased lungs.9 Because of these potential advantages, NHFOV has seen growing adoption, especially in certain European countries.10 When used as support after extubation, NHFOV reduces reintubation rates compared with NCPAP.11 However, its efficacy as primary respiratory support for respiratory distress syndrome in preterm infants remains unproven.
Several randomised controlled trials comparing NHFOV and NCPAP have reported inconsistent findings about their effectiveness in managing respiratory distress syndrome.12131415 To address this gap, we conducted a multicentre randomised controlled trial to compare the effectiveness of NHFOV and NCPAP as primary respiratory support in extremely preterm infants with respiratory distress syndrome. We hypothesised that NHFOV would be more effective than NCPAP in reducing treatment failure within 72 hours after birth in a homogeneous population of Chinese extremely preterm infants.
Methods
Study design
This multicentre, randomised controlled superiority trial was conducted across 20 tertiary neonatal intensive care units in 13 provinces, autonomous regions, and municipalities in China from August 2022 to August 2024 (supplementary eTable 1). The study was approved by the ethics committee of the Children’s Hospital of Chongqing Medical University (No 2019.161) and was registered at ClinicalTrials.gov (NCT05141435). The trial protocol was previously published, providing full details of the methods.16 All authors reviewed the protocol and ensured adherence throughout the trial. Informed consent was obtained from parents or guardians antenatally or upon admission to the neonatal intensive care unit, and data were anonymised in accordance with relevant local regulations. The trial was conducted in compliance with CONSORT (consolidated standards of reporting trials) guidelines.17 Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.
Study participants
Newborns were enrolled if they met all of the following criteria: gestational age between 24 weeks +0 day and 28 weeks +6 days; diagnosis of respiratory distress syndrome and prerandomisation support with NCPAP set at 6 cm H2O of positive end expiratory pressure (after delivery room stabilisation at 6-8 cm H2O per local guidelines) with a fraction of inspired oxygen (FiO2) >0.25 to maintain a target peripheral oxygen saturation (SpO2) of 89-94%; less than two hours after birth at the time of enrolment; and informed parental consent obtained before randomisation. Infants were excluded in the presence of any of the following: intubation at birth for any reason; major congenital malformations or known complex congenital heart disease; and transfer out of the neonatal intensive care unit before randomisation.
Randomisation and masking
Eligible newborns were randomly assigned to receive NCPAP or NHFOV upon meeting all inclusion criteria. Simple randomisation was performed using a computer generated random number sequence, which was securely posted on a dedicated, password protected website accessible 24/7. Randomisation was centrally managed by the website, ensuring allocation concealment from investigators. Infants were stratified according to centre. Crossover between study arms was not permitted throughout the study period to maintain protocol integrity.
Owing to the nature of the intervention with the distinct characteristics of the ventilation modalities, masking of care givers was not feasible, and masking of patients was not applicable. However, outcome assessors were masked to the treatment allocation to minimise ascertainment bias. Outcome data were recorded by investigators who were not part of the neonatal intensive care unit staff and who reviewed patient files with masked treatment assignments. A dedicated assessor was designated for each recruiting neonatal intensive care unit to ensure unbiased evaluation.
Study procedure
In the delivery room, all eligible infants who were spontaneously breathing received NCPAP stabilisation at 6–8 cm H2O and were transferred to the neonatal intensive care unit with continuous NCPAP support. After admission to the neonatal intensive care unit, all enrolled infants were maintained on standardised NCPAP at 6 cm H2O for a minimum of 30 minutes before randomisation to establish uniform baseline respiratory support conditions.
Continuous flow devices (Comen NV8, China; Mindray NB350, China; supplementary eTable 2) were used to provide NCPAP. Newborns randomised to the NCPAP group were started on a positive end expiratory pressure of 6 cm H2O, with adjustments up to a maximum of 10 cm H2O according to oxygenation (in increments of 1 cm H2O). The FiO2 was started at 25%, with adjustments up to a maximum of 40% (in increments of 5%) to maintain a preductal SpO2 target of 89-94%.
NHFOV was delivered using piston or membrane oscillators capable of active expiration, including the Fabian-III (Acutronic, Switzerland), SLE 5000 (SLE (UK)) and Leoni+ (Löwenstein Medical, Germany; supplementary eTable 2). Newborns randomised to NHFOV were started with the following parameters: mean airway pressure 6 cm H2O (adjustable in 1 cm H2O increments within a range of 6–10 cm H2O according to oxygenation); frequency 10 Hz (modifiable in 1 Hz increments within a range of 8–12 Hz); inspiratory time 50% (1:1 ratio)18; amplitude 15 cm H2O (modifiable in 5 cm H2O increments within a range of 15–30 cm H2O).
NCPAP and NHFOV were administered using the same short binasal prongs because these represent the standard of care in the participating neonatal intensive care units and offer the lowest resistance and minimal leakage.1920 Prong size was selected based on the diameter of the nares, ensuring the best fit—the largest prongs that fit the nares without causing blanching of the surrounding tissues. To minimise air leaks, special attention was given to supportive measures, including the use of pacifiers and optimal positioning. Ventilatory parameters were managed and weaned as outlined in the trial protocol.16 Suggested initial and maximum settings for respiratory support were provided to the study sites. To ensure protocol adherence, a dedicated training programme was deployed to disseminate this ventilatory management policy to all participating neonatal intensive care units.
Caffeine administration
Prophylactic caffeine treatment (caffeine citrate injection, Chiesi Pharmaceuticals, Parma, Italy) was administered within the first 24 hours after birth.4 The initial loading dose was 20 mg/kg, followed by a maintenance dose of 5 mg/kg per day. Caffeine treatment was discontinued once apnea resolved or when the infant reached 36 weeks of corrected gestational age.
Surfactant treatment
Surfactant treatment (Curosurf, Chiesi Pharmaceuticals) was administered at a dose of 200 mg/kg if FiO2 >30% on positive end expiratory pressure or mean airway pressure≥6 cm H2O was required to maintain a target SpO2 of 89-94% using the less invasive surfactant administration technique.4 A repeat dose of surfactant of 100 mg/kg was administered if necessary. Surfactant doses were given at 6-12 hour intervals, with a maximum of four doses permitted.
Other treatments
Antibiotics are frequently started in newborns with respiratory distress syndrome until sepsis is excluded. Fluid treatment starts at approximately 70-80 mL/kg/day, with subsequent adjustments individualised based on fluid balance, weight changes, and serum electrolyte levels. Cardiac ultrasound should be performed within the first three days of life to evaluate cardiac morphology, pulmonary pressures, and patent ductus arteriosus. If a haemodynamically significant patent ductus arteriosus is identified, it is treated with ibuprofen: an initial dose of 10 mg/kg followed by two 5 mg/kg doses at 24 hour intervals, typically administered orally. Should the haemodynamically significant patent ductus arteriosus fail to close after two courses of ibuprofen and significantly compromise cardiopulmonary function, surgical ligation is considered.
Primary outcome
The primary outcome was respiratory support failure, defined by the need for invasive mechanical ventilation within 72 hours after birth. Respiratory support failure was considered present if any of the following criteria were met: severe respiratory acidosis, defined as partial pressure of CO2 >60 mm Hg with pH <7.2 for at least one hour; hypoxia refractory to study intervention, defined as SpO2 <90% despite FiO2 ≥0.4 and maximum allowable pressures within the assigned study arm, persisting for at least one hour after surfactant administration; severe apnea, defined as recurrent apnea (more than three episodes per hour) associated with a heart rate <100/min or a single apnea episode requiring bag mask ventilation; and urgent intubation as determined by the attending physician.
Secondary outcomes
Prespecified secondary outcomes included reasons for treatment failure, the need for invasive mechanical ventilation within seven days after birth, other complications like complications of prematurity such as surfactant treatment, duration of supplemental oxygen, bronchopulmonary dysplasia defined in 2019,21 intraventricular haemorrhage,22 necrotising enterocolitis,23 retinopathy of prematurity,24 and haemodynamically significant patent ductus arteriosus.25
Safety outcomes
Serious adverse events were defined as pneumothorax or other air leaks occurring during the assigned treatment and death in hospital. Additionally, thick secretions causing an airway obstruction and grade III-IV nasal injury.26
Statistical analysis
Based on the results of a recent multicentre study, the estimated risk of treatment failure with NCPAP in extremely preterm infants with respiratory distress syndrome was 30%, with a 20% reduction expected for infants receiving NHFOV.12 To ensure adequate statistical power with a relatively conservative estimate, we aimed to detect a 15% absolute difference in the rate of respiratory support failure between groups. With a significance level (α) of 0.05 and 90% power, a sample size of 170 newborns per group was required, leading to a total enrolment target of at least 340.
Outcomes were analysed on an intention-to-treat basis. For dichotomous outcomes, the risk difference in percentage points with a 95% confidence interval was calculated between study arms. For continuous outcomes, the mean difference with a 95% confidence interval or the Hodges-Lehmann median difference with interquartile range was reported, as appropriate. Comparisons were performed using Student’s t test for parametric continuous variables, Mann-Whitney U test for non-parametric continuous variables, and χ2 test for dichotomous variables, respectively. A two tailed P value <0.05 was considered statistically significant. All statistical analyses were conducted in R software (version 4.4.2).
Patient and public involvement
No parents were involved in setting the research question or the outcome measures, nor were they involved in developing plans for recruitment, design, or implementation of the study. No patients were asked to advise on interpretation or writing up of results. It was not the policy of the involved institutions to include parents or members of the public in planning or decision making processes at the time when the study was planned, submitted to ethical committees and funding agencies, and started.
Results
Study population
From August 2022 to August 2024, 20 sites screened 684 infants. Of these, 360 met the eligibility criteria and consent was provided by parents or guardians. This consent was withdrawn before randomisation for 18 newborns, leaving 342 preterm infants to proceed to randomisation. These infants received routine clinical care, and their trial data were destroyed in accordance with ethical guidelines. Consequently, data from 342 newborns were included in the final analysis (fig 1). Baseline characteristics were well balanced between the two groups except for clinical risk index for babies II score (P=0.05) and PaO2 before enrolment (P=0.02), which is significantly higher in the NHFOV group (table 1). The interim analysis did not reveal any safety concerns, and the trial was completed without major adverse events, discontinuation, or protocol modifications.
Baseline characteristics in NHFOV and NCPAP groups*
Primary outcome
Treatment failure within 72 hours after birth occurred in 27 of 170 infants (15.9%) in the NHFOV group and 48 of 172 (27.9%) in the NCPAP group (risk difference −12.0 percentage points, 95% confidence interval −20.7 to −3.4; P=0.007; table 2). The reasons for treatment failure within 72 hours after delivery did not differ significantly between the NHFOV and NCPAP groups, with hypoxia reported as the most common reason (70.4%, 19/27 v 66.7%, 32/48), followed by apnea (14.8%, 4/27 v 18.8%, 9/48), respiratory acidosis (7.4%, 2/27 v 8.3%, 4/48), and urgent need for intubation (7.4%, 2/27 v 6.2%, 3/48; fig 2), respectively. However, those with reported hypoxia seemed to have borderline reduction in risk in the NHFOV group compared with the NCPAP group (−7.4 percentage points, 95% confidence interval −14.9 to 0.1; P=0.05; table 2). Additionally, NHFOV provided better oxygenation at 24 hours after intervention (supplementary eTable 3). To account for baseline difference between the two groups, we also performed a sensitivity analysis by additionally adjusting for clinical risk index for babies II score, PaO2, study sites, and use of antenatal steroid individually in all examined associations, and the significance was not attenuated (supplementary eTables 4 and 5).
Comparison of risk difference between NHFOV and NCPAP in treatment failure within 72 hours after birth among extremely preterm infants
Pie chart displaying reasons for treatment failure within first 72 hours in NHFOV (n=27) and NCPAP (n=48) groups. NCPAP=nasal continuous positive airway pressure; NHFOV=non-invasive high frequency oscillatory ventilation
Secondary outcomes
Treatment failure within seven days after birth was reported in 36 of 170 infants (21.2%) in the NHFOV group and 58 of 172 (33.7%) in the NCPAP group (risk difference −12.5 percentage points, 95% confidence interval −21.9 to −3.2; P=0.008; table 3). Additionally, no significant differences were observed between groups in surfactant administration, bronchopulmonary dysplasia incidence, bronchopulmonary dysplasia and mortality composite, duration of study intervention, duration of supplemental oxygen, and extrapulmonary outcomes (intraventricular haemorrhage, necrotising enterocolitis, retinopathy of prematurity, and haemodynamically significant patent ductus arteriosus; table 3).
Secondary outcomes and safety outcomes
Safety outcomes
No significant differences were observed between groups for serious adverse events, including rates of grade III-IV nasal injury, air leaks, death in hospital, and thick secretions causing an airway obstruction (table 3).
Discussion
Principal findings
In this multicentre randomised controlled trial, we found that NHFOV significantly reduced the need for invasive mechanical ventilation within 72 hours and within seven days after birth compared with NCPAP when applied as the primary approach to respiratory distress syndrome in extremely preterm infants. Additionally, there was no difference between the groups in other respiratory and extra-respiratory secondary outcomes and adverse events.
Comparison with other studies
Our findings should be contextualised within previous randomised controlled trials comparing NHFOV with NCPAP as primary treatment for respiratory distress syndrome. Although Malakian and colleagues13 reported no significant reduction in intubation rates with NHFOV versus NCPAP, their study population differed critically from ours: it included more mature infants and had a smaller sample size. Iranpour and colleagues14 showed much lower intubation rates with NHFOV versus NCPAP in infants of comparable gestational age to Malakian et al’s cohort. However, their study population encompassed infants with respiratory distress syndrome of varying severity, including those with severe respiratory failure requiring FiO2 >0.6 at enrolment. Zhu and colleagues12 likewise reported that NHFOV significantly reduced intubation rates in preterm infants with moderate to severe respiratory distress syndrome compared with NCPAP. In a small randomised controlled trial of infants weighing <1250 g after continuous positive airway pressure failure, Mukerji and colleagues27 compared NHFOV with biphasic continuous positive airway pressure and identified a non-significant trend towards fewer intubations with NHFOV. Further evidence supporting NHFOV’s efficacy comes from Rüegger and colleagues,28 who found that NHFOV significantly reduced desaturation episodes and bradycardic events in very preterm infants compared with NCPAP. In contrast, Klotz and colleagues,29 in a small crossover trial of a comparable population, detected no significant improvement in carbon dioxide clearance with NHFOV; however, the study was underpowered to evaluate clinical outcomes such as intubation rate. In a large multicentre, randomized controlled trial, Zhu and colleagues15 showed that NHFOV significantly reduced treatment failure rates compared with NCPAP in preterm infants with gestational age <30 weeks and those with very low birth weight. Taken together, these data suggest that NHFOV may mitigate the need for invasive mechanical ventilation, particularly in infants of lower gestational age or with more severe respiratory failure.
Although NHFOV significantly reduced intubation rates compared with NCPAP, we found no between group difference in bronchopulmonary dysplasia incidence. This null finding requires nuanced interpretation for three reasons. Firstly, bronchopulmonary dysplasia pathogenesis involves multifactorial processes, including mechanisms independent of ventilator induced injury.30 Secondly, our trial was underpowered for bronchopulmonary dysplasia as a secondary outcome. Thirdly, we observed a clinically meaningful, though non-significant, reduction in bronchopulmonary dysplasia with NHFOV, suggesting potential lung protective effects that warrant evaluation in larger trials.
In prespecified safety analyses, NHFOV and NCPAP showed comparable adverse event rates, reinforcing the clinical feasibility of NHFOV as a respiratory support modality for extremely preterm infants. This finding is particularly noteworthy given that the high amplitude oscillations inherent to NHFOV could theoretically increase the risk of complications such as air leaks, secretion induced airway obstruction, or severe (grade III-IV) nasal trauma. Our results align with the safety profile reported in the NASONE trial,9 providing further evidence that NHFOV does not significantly increase barotrauma risk—even at higher mean airway pressure—when applied under rigorous monitoring and appropriate patient selection. Nevertheless, we emphasise the need for individualised mean airway pressure titration, as optimal pressure settings may vary depending on the underlying pathophysiology (eg, respiratory distress syndrome versus meconium aspiration syndrome). These observations suggest that earlier concerns about NHFOV associated barotrauma may have been overstated, provided clinicians adhere to standardised protocols for real time adjustment and complication prevention.
Clinical implications
This multicentre randomised controlled trial is a large scale study comparing NHFOV with NCPAP as primary respiratory support for respiratory distress syndrome in extremely preterm infants. Our findings corroborate and extend the conclusions of two recent meta-analyses,3132 and provide new insights into NHFOV’s potential use in managing respiratory distress syndrome in this vulnerable population.
The results advance our current understanding by showing that NHFOV is feasible and potentially advantageous as a primary respiratory treatment for extremely preterm infants—a population notably underrepresented in earlier trials. This aligns with, and strengthens, previous evidence from retrospective studies.333435 Importantly, our study design addressed a key historical limitation in comparing NHFOV and NCPAP: variability in applied airway pressures. By maintaining comparable pressure ranges between modalities, we enabled a more direct and clinically meaningful comparison.
Strengths and limitations
This study possesses several key strengths: enrolment of a large cohort of extremely preterm infants; implementation of a rigorous, physiology targeted protocol with standardised ventilatory management; and execution across an established neonatal intensive care unit network where all clinicians underwent protocol specific training. Participating neonatologists included leading academic experts from the Chinese Medical Doctor Association and the Chinese Society of Perinatal Medicine, ensuring technical proficiency and strict adherence to the study protocol. These features bolster the internal validity and clinical applicability of our findings for comparable populations.
Nevertheless, several limitations should be acknowledged. Clinician masking was unfeasible owing to the inherently distinct interventions, which may have introduced selection, performance, or detection bias. To address this, we used masked outcome assessments and adhered to predefined objective criteria. Although the less invasive surfactant administration technique reduced intubation risks compared with INSURE (intubate-surfactant-extubate), most treatment failures stemmed from procedural complications (eg, derecruitment and uneven surfactant distribution36) rather than respiratory modality efficacy. Additionally, although our protocol allowed clinician directed surfactant redosing at 6-12 hour intervals based on respiratory parameters, recent pharmacokinetic evidence suggests this approach may require refinement.37
Our study focused solely on the continuous flow continuous positive airway pressure device and did not include the bubble continuous positive airway pressure system. Although bubble continuous positive airway pressure is widely available and generally less expensive globally, its exclusion may limit the generalisability and applicability of our findings to settings where it remains the primary system. Additionally, the use of binasal prongs (to ensure consistency across centers) precluded evaluation of potentially less traumatic mask interfaces—a critical comparison for future studies, particularly regarding prolonged applications. Our cohort included relatively few infants with gestational age <26 weeks (41/342, 12%), leaving the generalisability of results to this subgroup uncertain. Finally, while our findings are robust for Chinese extremely preterm infants, broader external validity necessitates validation in more diverse populations across varying gestational ages and clinical settings.
Conclusions
This randomised clinical trial found that NHFOV significantly reduced the need for invasive mechanical ventilation within 72 hours and within seven days after birth compared with NCPAP when applied as the primary approach to respiratory distress syndrome in extremely preterm infants. Both respiratory support techniques were equally safe.
What is already known on this topic
-
Previous retrospective studies and small scale observational trials suggested that non-invasive high frequency oscillatory ventilation (NHFOV) may be superior to nasal continuous positive airway pressure (NCPAP) as a primary respiratory treatment for preterm infants with respiratory distress syndrome
-
Subsequent randomised controlled trials comparing NHFOV and NCPAP have yielded inconsistent findings for their efficacy to manage respiratory distress syndrome
What this study adds
-
This randomised clinical trial found that NHFOV significantly reduced the need for invasive mechanical ventilation within 72 hours and seven days after birth compared with NCPAP when applied as the primary approach to respiratory distress syndrome in extremely preterm infants
-
Both respiratory support techniques were found to be equally safe
