Impact of Delayed Initiation of Non-Invasive Ventilation for Respiratory Distress in Outborn Late-Preterm and Term Neonates


 The frequency of non-invasive ventilation (NIV) use increased in neonates of all gestational ages with respiratory distress (RD). However, the impact of delayed initiation of NIV support in outborn neonates remains poor understood. The aim of the present study was to identify the impact of delayed initiation of NIV in outborn late-preterm and term neonates. The medical records of 277 infants (gestational age of ≥ 35 weeks) who were received NIV as primary respiratory therapy <24 h of age between 2016 and 2020 were retrospectively reviewed. Among the 190 outborn neonates, the factors associated with respiratory adverse outcomes were investigated. Infants with RD divided into two groups, mild (FiO2 ≤0.3) and moderate-to-severe RD group (FiO2 >0.3), depending on initial oxygen requirement under NIV support. Median time to start of NIV support at tertiary center was 3.5 (2.2–5.0) h. Male sex (odds ratio [OR], 2.9; 95% CI, 1.1–7.7), high oxygen requirement (FiO2 >0.3) (OR, 4.8; 95% CI, 1.5–15.3), and respiratory distress syndrome (OR, 10.4; 95% CI, 3.9–27.8) were the significant factors associated with adverse outcomes. Subgroup analysis revealed that in the moderate-to-severe RD group, delayed initiation of NIV (≥3 h) was significantly associated with pulmonary air leakage (p=0.033).Conclusions: Our study shows that outborn neonates with moderate-to-severe RD who were treated with delayed NIV were associated with an increased likelihood of pulmonary air leakage. Additional prospective studies are needed to establish the optimal timing and methods of NIV support for outborn late-preterm and term infants.


Introduction
Respiratory distress (RD) is one of the most frequent causes of admission to the neonatal intensive care unit (NICU) in term and preterm infants [1,2]. Among infants admitted to the NICU, 15% of term infants and 29% of late-preterm infants develop signi cant respiratory morbidity. The most common respiratory morbidities are transient tachypnea of the newborn (TTN) and respiratory distress syndrome (RDS) [3].
Respiratory signs and/or requirement of oxygen typically present within the rst hours of life, often immediately after delivery [3,4].
Current neonatal resuscitation program (NRP) guidelines in the delivery room (DR) recommend continuous positive airway pressure (CPAP) as a corrective measure for all spontaneously breathing infants who present with labored breathing or persistent cyanosis [5,6]. Hence, the frequency of CPAP use increased in late-preterm and term infants with RD within the rst hours of life [7]. Various modes of non-invasive ventilation (NIV) (e.g., humidi ed high ow nasal cannula [HHFNC], nasal intermittent positive pressure ventilation [NIPPV], biphasic positive airway pressure [BiPAP]) are used as primary respiratory support together with CPAP, depending on the severity of manifestation [8,9] Despite this practical change, most infants who would have bene tted from NIV are born in community hospitals that do not offer this therapy. Alternatively, these infants receive supplemental oxygen via oxygen hood or nasal cannula while waiting for clinical improvement [10,11]. Consequently, they receive high concentrations of inspired oxygen without su cient distending pressure for a signi cant length of time. There is some evidence that early application of continuous distending pressure has a clinical bene t over delayed application in very preterm infants [12][13][14]. Unfortunately, data on the impact of delayed NIV in outborn late-preterm and term infants is scarce. Furthermore, its early use in these infants with mild RD is controversial due to concern for air leakage (e.g., pneumothorax) [15][16][17].
We conducted this study to assess the factors associated with adverse respiratory outcomes in outborn neonates born at ≥35 +0 weeks of gestation, who received NIV as primary respiratory therapy. We hypothesized that delayed initiation of NIV leads to adverse outcomes that vary according to the severity of the underlying disease.

Methods
We retrospectively reviewed the medical records of 1,709 neonates at ≥35 +0 weeks of gestation who were hospitalized in the tertiary referral NICU of Korea University Ansan Hospital between January 2016 and December 2020. Our center is a regional hospital with level III NICU staff and nursing staff experienced in NIV use. During the 5-year period, there were 526 inborn neonates and 1,183 outborn neonates admitted to the NICU. This study was approved by the Institutional Research Ethics Committee of the Korea University Ansan Hospital (2021AS0193). Patient consent was waived due to retrospective study design.

Inclusion and exclusion criteria
The enrollment criteria were as follows: (1) neonates with RD but breathing spontaneously and received respiratory support, and (2) gestational age ≥35 +0 weeks. We excluded patients with major congenital/chromosomal anomalies, suspected birth asphyxia, meconium-stained amniotic uid, spontaneous pneumothorax, suspected pneumonia or sepsis, idiopathic persistent pulmonary Page 4/19 hypertension (PPHN), non-pulmonary issues, those intubated in the DR and immediately placed on mechanical ventilation, and those who received primary respiratory therapy at ≥24 h of age.

Perinatal factors and de nitions
Perinatal data were collected from the discharge summary report and con rmed by a detailed chart review. The collected information included maternal complications, gestational age, birth weight, size for gestational age, sex, mode of delivery, Apgar scores, the need for resuscitation or surfactant replacement therapy, presence of air leakage requiring thoracentesis, ventilatory support (presence, duration, modes, and type; invasive or noninvasive), duration of oxygen therapy, and days of NICU stay.
Data on symptom onset time, method of oxygen administration, and time to initiation of NIV therapy, which was de ned as the time (in hours) between symptom onset and the initiation of NIV at the referred hospital, were collected from the neonatal transport summary.
TTN was diagnosed in the presence of RD with onset within 6 h after birth, de ned as respiratory rate >60 breaths/min, grunting, nasal aring, and intercostal, subcostal, and/or suprasternal retractions. Radiographic features of TTN included bilateral alveolar and interstitial edema, prominent pulmonary vascular pattern with increased perihilar markings, hyperin ated lungs, and/or uid in the interlobar ssure. RDS was diagnosed based on the presence of at least two of the following symptoms: the need for supplemental oxygen, tachypnea, intercostal retraction, and grunting, with exclusion of other causes of RD. Radiographic features of RDS included low lung volumes and air bronchograms with diffuse, bilateral, and ground glass elds prior to surfactant treatment.
Neonates with RD were divided into two groups depending on their initial oxygen requirement with any NIV modalities; neonates in the mild RD group received a fraction of inspired oxygen (FiO 2 ) ≤30, while infants in the moderate-to-severe RD group received a FiO 2 >30.

Respiratory support techniques
Once admitted to the NICU, the neonates were provided primary treatment in the form of four possible NIV devices: nasal CPAP, synchronized nasal intermittent positive pressure ventilation (sNIPPV), BiPAP, and HHFNC. The choice of whether and when to use NIV was at the discretion of the medical team, with no speci c unit guideline on the choice of one strategy over another. HHFNC support was delivered using the Opti ow System (Fisher & Paykel Opti ow System, Healthcare, Auckland, New Zealand), with short binasal prongs. Infants received gas ow of 5 L/min initially, which was adjusted to 3-7 L/min, according to their condition.

Outcomes
Adverse outcomes were de ned as mechanical ventilation therapy or the incidence of pulmonary air leakage. The latter was de ned as pneumothorax or pneumomediastinum with RD that occurred during hospitalization, as diagnosed with chest radiography.

Statistical analysis
All data analyses were performed using SPSS 20.0 for Windows (SPSS Inc., Chicago, IL, USA). Continuous variables were analysed using the t-test and Mann-Whitney U test for normal and skewed distributions, respectively. Proportions were tested using chi-squared and Fisher's exact tests; statistical signi cance was set at p <0.05. The signi cant variables identi ed in the univariate analysis were further assessed using a multivariable logistic regression analysis. Data are presented as mean ± standard deviation (SD), median and range, or rate.

Results
Out of the 1,709 neonates at ≥35 +0 weeks of gestation who underwent hospitalization between January 2016 and December 2020, 502 late-preterm and term neonates with RD required respiratory support. Once the differential diagnoses were excluded, 277 neonates remained; this was further divided into inborn (n = 87) and outborn (n = 190) patients, with the latter serving as the subjects of this study ( Figure 1).
The demographic and clinical data of the patients are presented in Table 1. The mean gestational age and birth weight of the patients were 37.9 ± 1.5 weeks and 3,228 ± 486 grams, respectively. Seventy-two percent (n = 135) of the neonates showed symptoms of RD shortly after birth and 65.3% (n = 124) required oxygen supplementation at DR. None of the outborn neonates received CPAP in the DR. The median time to start of NIV at our hospital was 3.5 (2.2-5.0) h, and the initial median SpO 2 and FiO 2 were 90% and 0.3, respectively. More cases of TTN (n = 116; 61.1%) were diagnosed than of RDS (n = 74; 38.9%). Outborn neonates had a greater history of mechanical ventilation (24.7% vs. 4.6%, p <0.001) and air leakage (17.4% vs. 1.1%, p <0.001) than inborn patients (Supplemental Table 1).
The following factors were signi cantly associated with adverse outcomes during hospitalization (  Table 3). Table 4 shows the clinical data for the mild and moderate-to-severe RD groups, further divided according to the time of NIV initiation (with 3 h as the cut-off). This allowed us to clarify the effects of underlying disease severity. In infants with moderate-to-severe RD, a signi cantly high incidence of air leakage occurred when the initiation of NIV therapy was delayed by >3 h, as compared to when NIV therapy was initiated earlier (35.7% vs. 5.9%, p = 0.033) (Figure 2). The median time to identify the pulmonary air leakage from initiation of rst NIV was not signi cantly different between two groups (15.  Table 5). There was no signi cant difference in treatment of pulmonary air leakage (p = 1.000), 63.6% of neonates with mild RD and 54.5% of neonates with moderate-to-severe RD were treated with conservative managements such as observation or oxygen therapy (Supplemental Table 2).

Discussion
To our knowledge, this is the rst study to directly investigate the association between delayed initiation of NIV and pulmonary air leakage in outborn late-preterm and term neonates with signi cant oxygen requirement.
Non-invasive respiratory support (e.g., CPAP utilization) seems to be increasing, particularly in later gestational age categories, following the release of the NRP guidelines in 2011. Moreover, recent studies have shown that NIV, as compared to free ow O 2 , may decrease the duration and severity of RD in latepreterm and term infants with TTN or RDS [9,10,18,19]. In the present study, we found that out of the total number of neonates who were in RD and breathing spontaneously, 83% received CPAP or noninvasive positive pressure ventilation (NPPV) as their primary respiratory support, whereas 17% received HHFNC after NICU arrival.
In agreement with previous studies, a low GA, male sex, birth by cesarean section, high oxygen requirement, and RDS were signi cant factors associated with adverse outcomes in our cohort [3,20]. Additionally, our results showed that the frequency of air leakage was particularly high when there was a delay in respiratory support in cases of moderate-to-severe RD; while oxygen supplementation was initiated in all referred hospitals prior to NICU arrival, continuous pressure support was delayed due to the lack of appropriate devices. In a few small studies in the pre-surfactant era, delayed initiation of CPAP in infants with RDS was associated with signi cant development of pneumothorax [12,21,22]. The resulting difference in outcomes can be attributed to the alteration of the course of RDS among infants treated with continuous distending pressure early [21,22]. Smithhart et al. showed in a large birth cohort study that the incidence of pneumothorax decreased with oxygen requirement among neonates who received early CPAP in the DR, suggesting that the risk of CPAP-induced pneumothorax is lower in lungs with underlying disease [15]. Similarly, a recent study by Hirata et al. reported that adverse prognosis of severe TTN was strongly associated with a longer time between birth and transport to a level III NICU [23]. This is consistent with the results of our study; we classi ed the severity of RD, allowing us to further clarify the effects of the primary disease.
The impact of outborn status on the greater risk of severe respiratory morbidities is unclear, given the inconsistencies in diagnosis and management [24][25][26][27]. In the present study, the incidence of air leakage in outborn neonates (17.4%) was higher than that in inborn neonates (1.1%). Additionally, RDS was the underlying primary lung disease in a large number of outborn neonates. Previous studies that utilized the national epidemiologic data of Korea reported that among term infants with RDS, the incidence of pneumothorax notably increased from 6.9% to 11.7% over 5 years. In contrast, there was no change in the incidence of pneumothorax in preterm infants with RDS [28]. The clinical characteristics of outborn and inborn neonates were also different, suggesting a different underlying mechanism of pneumothorax and/or resuscitation quality in the DR and during medical transport [26][27][28][29]. It is possible that the initiation of delayed CPAP in these neonates augment the levels of lung distending pressure towards the threshold for lung rupture; this is due to the fact that neonates born at term or moderate-to-late-preterm are capable of generating auto-positive end-expiratory pressure through grunting and struggling [25].
Careful monitoring and precise adjustment of modes of respiratory support for air leakage are needed in outborn neonates with moderate-to-severe RD, including those diagnosed with RDS. In addition, based on the ndings of our study, we believe that early initiation of continuous distending pressure (e.g., CPAP) might lead to a marked reduction in adverse outcomes. Immediate application of non-invasive respiratory support, instead of oxygen supplementation, should be considered in late-preterm and term neonates who show apparent respiratory symptoms and signi cant oxygen requirement. Additional prospective studies are needed to establish the optimal timing and methods of primary respiratory support.
However, this study has several limitations-one of which is its retrospective design. The restriction of the study to neonates who were admitted to the NICU resulted in a small sample size. In the current study, antenatal steroids were not administered to the mothers of all outborn neonates; this is not in concordance with the by the American College of Obstetricians and Gynecologists (ACOG) guidelines in 2020 which recommend the administration of betamethasone for pregnant women who are at risk of having a late-preterm birth [30]. The indication for each NIV device was at the discretion of the attending neonatologist, which may have in uenced the outcomes. Notably, there were no signi cant differences in the type of primary respiratory support between the inborn and outborn neonates, with or without adverse outcomes. Since the accessibility of outborn neonates with RD varies widely by center, our results may not be generalizable to other centers where the characteristics of patients differ from ours.
In conclusion, this study con rmed that outborn infants with moderate-to-severe RD who were treated with delayed non-invasive respiratory support were associated with an increased likelihood of pulmonary air leakage.

Declarations
Funding This research received no external funding.

Con icts of interest
The authors declare no con ict of interest Availability of data and material The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Flow chart of the study population