Nasal CPAP complications in very low birth weight preterm infants

1 Introduction

For more than four decades, a rise in the use of non-invasive continuous positive airway pressure (CPAP) has been widely recorded, becoming the gold standard for noninvasive respiratory support in very low birth weight (VLBW) infants [1, 2]. The main goal of this ventilatory method is to provide continuous support during both inspiratory and expiratory phases of the respiratory cycle, maintaining a positive pressure and ultimately improving gas exchanges at the alveoli level. This is explained by an increase in alveolar recruitment, functional residual capacity (FRC) and tidal volume [2], with concomitant reduction of the respiratory workload, the need for supplementary oxygen, the number of apnea episodes and the area of atelectasis [3].

Over the last few years, many other non-invasive ventilatory modalities have been developed and increasingly used in order to reduce the need for invasive mechanical ventilation (IMV) in the newborn care setting. Nasal High-Frequency Ventilation (nHFV) [4] and bi-level nasal CPAP, similar to CPAP but relying on not one but two positive end-expiratory pressure (PEEP) levels to improve oxygenation [5], deserve particular emphasis.

Nasal CPAP (nCPAP) can be administered utilizing a variety of patient interfaces, the most frequent being binasal prongs and face mask [6]. Nasal prongs are the most common and generally accepted method since they are a less invasive way of supplying CPAP. These are available in different sizes and are usually made of light flexible material. Despite its advantages, some complications due to its use may include nasal obstruction as a result of increased secretions or incorrect application of nasal prongs, gastric distension due to excessive air swallowing, increased incidence of necrotizing enterocolitis, pneumothorax, pneumomediastinum, pneumopericardium [2, 4] or septal injury [7]. Because of skin and mucous membranes immaturity of preterm neonates, the tension required to apply to the cannulas to maintain a constant pressure in the airways and effectively deliver nCPAP may easily injure their noses (sometimes even with short-term applications) when opting for improperly sized and/or inappropriately positioned cannulas [1].

Nasal trauma is, consequently, one of the most common complications of nCPAP use. Reported incidence in the literature ranges from 15–20% to 60% in the neonatal population submitted to this ventilatory mode – higher percentages can be recorded with younger gestational ages [8] –, occurring primarily in the medial aspect of the nostrils and in the columella, where nasal prongs or transverse cannula exert pressure on the nasal septum [9]. Therefore, besides the misalignment and improper fixation of the prongs, other described risk factors include duration of nCPAP longer than 5 days, gestational age lesser than 32 weeks, birth weight below 1500 g and inappropriate monitorization of the skin and surrounding tissue to the nose [4]. Fischer et al. [8] suggested a classification in three stages for the severity of nasal trauma, according to the standardized classification of decubitus lesions from the US National Pressure Ulcer Advisory Panel (NPUAP): mild (stage I), with persistent hyperemia on an intact septal skin; moderate (stage II), with superficial ulcer or erosion, with partial loss of skin thickness; severe (stage III), with necrosis and complete loss of skin thickness [8]. Without proper care, these lesions may progress to permanent deformity (nares or columella asymmetry, nasal tip deviation or collapse, absence of nasal tip projection and airway obstruction), with a possible need for surgical correction, nosocomial infections and more severely, sepsis [1].

The aims of this study were to evaluate the prevalence, to identify risk factors and to describe the clinical characteristics, management and outcome of VLBW preterm infants with nasal lesions due to nCPAP use, as well as to identify the preventive measures that are imperative for a better prognosis.

2 Material and methods

This retrospective analytical study included neonates hospitalized at our local Neonatal Intensive Care Unit (NICU, a level III unit with 17 beds and about 400 admissions per year), between 1st January 2012 and 31st December 2017, with less than 1500 g of birth weight and less than 32 weeks of gestational age, who had been submitted to nCPAP for more than 12 hours. We excluded from this analysis all patients with major malformations (3 patients), chromosomal abnormalities (1 patient), deceased before completing 7 days of life in the local NICU (3 patients) and those who were outborn and admitted after the first 72 hours of life (10 patients). All infants who were transferred to the center of their residence area before completing 36 weeks of gestational age were in spontaneous breathing and their clinical evolution was assessed until time of discharge from our NICU.

All data were collected from hospital electronic clinical database and patients’ medical records. These data include information regarding pregnancy and delivery (antenatal corticosteroids, gestational age, multiple gestation, type of delivery), gender, birth weight, Apgar score at 1st and 5th minutes, need for resuscitation with endotracheal tube, mother’s health characterization (age, chronic hypertension, gestational hypertension, pre-eclampsia, obesity, diabetes, chronic illness, infections, alcohol, smoking habits and drugs usage) and characterization of the placenta (anatomopathological diagnosis of chorioamnionitis, funisitis or chorionic vasculitis). Data regarding neonatal morbimortality and respiratory support were also obtained, the latter specifically focusing on the intermittent mandatory ventilation (IMV), nCPAP, higher fraction of inspired oxygen (FiO₂) and duration of oxygen therapy. Moreover, nCPAP complications data were also collected, according to Fischer’s classification [8] and considering the best equivalent stage to the described lesion. First recorded stage of nasal lesion, number of days after nCPAP administration until the onset of the injury and evolution to an upper stage were collected from electronic nursing registry.

Assessment of gestational age was made by post-menstrual age, ultrasound examination or the New Ballard Score (in the absence of obstetrical indexes) [10, 11]. Small for gestational age (SGA) was defined as a birth weight below the 10th centile of Fenton’s growth charts [12–14]. We considered a full cycle of antenatal corticosteroids when at least 12 hours had elapsed after the last intramuscular dose of dexamethasone or betamethasone (four doses of 6 mg given 12 h apart or two doses of 12 mg given 24 h apart, respectively) administered to the mother. Such protocol is in line with the National Institutes of Health (NIH) Consensus Development Panel on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes [15].

Chronic hypertension was defined as described by the European Society of Cardiology [16]. We considered the diagnosis of chronic hypertension as blood pressure (BP) values over 140/90 mmHg, either identified before conception or detected before 20 weeks of gestation. Gestational hypertension and pre-eclampsia (PE) were defined in consonance with the American College of Obstetricians and Gynecologists [17]. PE was diagnosed if BP values over 140/90 mmHg were present, with proteinuria higher than 300 mg in 24 h, after 20 weeks of gestation in a formerly normotensive woman. We defined the diagnosis of gestational hypertension as a new-onset elevation of BP after 20 weeks of gestation, in the absence of concomitant proteinuria. Obesity was considered if the body mass index (BMI) value was over 30 or waist circumference >88 cm in women [16]. Gestational diabetes was determined as a primary diagnosis of diabetes in the second or third trimester of pregnancy that is not clearly either pre-existing type 1 or type 2 diabetes [18].

Histological chorioamnionitis and chorionic vasculitis were both described as the infiltration of polymorphonuclear leukocytes: the former in fetal membranes and chorionic plate and the latter in chorionic vessel walls [19]. Funisitis, in turn, was defined as the presence of these cells in umbilical vessel walls or Wharton jelly [19].

Respiratory distress syndrome (RDS) was defined and treated based on the European Consensus on RDS of 2013 [20] and bronchopulmonary dysplasia according to the NIH Consensus definition [21, 22]. Our NICU has a protocol for positive pressure ventilation that includes different ventilatory strategies according to different lung diseases and favors the use of permissive hypercapnia. For very low birth weight infants, nCPAP right after birth is the preferable mode of ventilation and it is done using Infant Flow (CareFusion, Yorba Linda, CA, USA) with prongs. Nurse:patient ratio is 1 : 2. Caffeine citrate was used since day one of life to all infants. Oxygen was administered to maintain saturations between 90–95%, assessed by pulse oximetry (SpO₂).

Sepsis was considered in the presence of a positive blood culture, combined with clinical and laboratory parameters [23]. Intraventricular hemorrhage was defined according to Volpe [24]. Retinopathy of prematurity was diagnosed and classified as stated in the revised International Classification of Retinopathy of Prematurity [25]. Hemodynamically significant patent ductus arteriosus was screened and diagnosed based on echocardiographic findings, in agreement with the national guidelines [26]. Necrotizing enterocolitis was defined by clinical findings and radiological features, according to the modified Bell’s criteria [27]. Periventricular leukomalacia was diagnosed by ultrasound and classified according to de Vries et al. [24]. Pneumothorax was assessed by chest radiograph [28].

The study protocol was approved by local Ethics Committee.

3 Results

A total of 135 patients were included in the study, of them 73 males and 62 females. Mean gestational age was 28 (26–30) weeks and mean birth weight was 1072 (833–1311) grams. All included infants received nCPAP and 30 patients were transferred to the center of their residence area before completing 36 weeks of gestational age in spontaneous breathing. Prevalence of nasal lesion due to noninvasive ventilation with nCPAP in VLBW preterm infants at our NICU was 65% (88 patients). Of these, 83 patients (94%) had an initial diagnosis of stage I nasal lesion, according to the criteria considered by Fischer et al. [8], and 5 of them (6%) had an initial diagnosis of stage II nasal lesion, when first perceived by the nursing team. Median number of days after nCPAP administration until the onset of a clinically identifiable nasal lesion was 4 (1–14). Despite being given adequate nursing care measures, 18 patients (20%) evolved into an upper stage (17 (94%) from a stage I into a stage II and 1 (6%) from a stage II into a stage III) during hospitalization. Median time of evolution to an upper stage was 4 days (1–24). Summarily, 47 (35%) patients who were submitted to nCPAP ventilation did not develop nasal injury, 66 (49%) patients developed a stage I nasal lesion, 21 (16%) developed a stage II nasal lesion and 1 (1%) patient developed a stage III nasal lesion. The analyzis of sociodemographic data of both groups reported in Table 1 showed statistically significant differences in gestational age (p < 0.001) and birth weight (p = 0.006).

Table 2 describes the maternal, gestational, prenatal, perinatal and placenta characteristics among infants who were ventilated with nCPAP for more than 12 hours at our NICU and Table 3 reports the neonatal morbimortality of this population, both according to the onset of nasal lesion. Statistically significant differences were found between the groups with and without nasal lesion, concerning respiratory distress syndrome (p = 0.032), bronchopulmonary dysplasia (p = 0.001) and ductus arteriosus patency (p = 0.004). The average number of days of hospitalization at NICU was 58 (15–111) in the lesioned group and 36 (5–100) in the group without nasal lesion; the average weight on discharge of these groups was 2163 (1306–3020) and 1935 (1023–2915) grams, respectively. Both groups presented statistical significance (p < 0.001 and p = 0.010, respectively). No statistically significant differences were reported when comparing the two groups concerning death rates.

Comparison between the lesioned group and the group without nasal lesion regarding ventilatory support is presented in Table 4. Infants who developed nasal lesion were ventilated in average 31 (11–54) days with nCPAP, while in the group without lesion the average number of nCPAP ventilation days was 7 (1–46) (p < 0.001). Of 135 patients supplemented with nCPAP, 58 (43%) infants were ventilated solely with this noninvasive method (p = 0.046). Ventilation with bi-level nCPAP (57 patients, 42%) and IMV (77 patients, 57%) were also associated with the development of nasal injury (p = 0.012 for bi-level nCPAP and p < 0.034 for IMV). Use of oxygen therapy (p = 0.004) and a longer duration of its use (p = 0.003) also led to a significant increase in nasal lesion cases. In the lesion group, the fraction of inspired oxygen was, on average, 0.35 (0.21–0.80), whereas in the group without lesion was 0.30 (0.21–0.80), with a p-value of 0.035. As reported in Table 5, the multivariate logistic regression revealed that patients with a longer duration of nCPAP ventilation were more likely to develop nasal lesion (OR = 1.098, 95% CI: 1.055–1.142), as well as those who were submitted to oxygen therapy (OR = 3.174, 95% CI: 1.014–9.929).

In our study, only one patient developed a stage III lesion, with necrosis, complete loss of skin thickness and septal defect. After a topical treatment with an antibiotic ointment, a hydrocolloid solution and a close follow-up with regular appointments, plastic surgery was not required, and the patient had a full recovery with a minimal aesthetic sequelae. Of those patients with nasal injury, 21 (24%) also developed pressure-derived lesions in other facial regions and 17 (19%) gastric distention at the time of nCPAP use; nonetheless, other complications such as pneumothorax, pneumopericardium, pneumediastinum or pulmonary interstitial emphysema weren’t recorded in any of the 135 infants covered by this study.

In this study, we aimed to compare the clinical characteristics between VLBW preterm infants with and without nasal lesions due to nCPAP use, with the intention of identifying risk factors for pressure injuries in this fragility area and predicting the outcomes, for a more critical and indispensable approach respecting the protective measures and, consequently, a better prognosis.

Our results show that nasal injury remains, in fact, a frequent problem in very low birth weight preterm infants receiving noninvasive ventilation with nCPAP. Its importance in clinical practice relies on the discomfort caused to the infant, the requirement for a change in the mode of respiratory support and the risks of sepsis and long-term nasal sequelae [29]. Most of the reported cases of nasal injury in our study are mild (stage I) and, as anticipated, occurred in neonates with younger gestational age and lower birth weight, which was also stated by Fischer et al., Casey et al. and Badr et al. [2, 8], who defend the implication of an immature skin in the pathogenesis of such trauma.

Besides pressure-derived lesions in other facial regions and gastric distention, no other complication (i.e. pneumothorax, pneumopericardium, pneumediastinum or pulmonary interstitial emphysema) was recorded. This finding contributes to emphasizing the safety of nCPAP as the gold standard for non-invasive respiratory support.

Nasal CPAP is currently the preferred approach to respiratory distress in most neonates. In fact, nCPAP therapy is usually indicated in cases of RDS, transient tachypnea of the newborn, pulmonary edema, meconium aspiration syndrome; obstructive airways diseases such as bronchopulmonary dysplasia and bronchiolitis; apnea and bradycardia of prematurity [3]; weaning from mechanical ventilation [30]; tracheomalacia and diaphragmatic paralysis [3]. This explains why pathologies such as respiratory distress syndrome and bronchopulmonary dysplasia are significantly associated with a higher frequency of nasal injury in our study.

In the same line of thought, invasive mechanical ventilation, bi-level nCPAP use, oxygen therapy and its duration, and higher values of FiO₂ all showed statistically significant differences between the two groups of our study, as a result of the treatment of the underlying respiratory pathologies.

The onset of nasal lesion was significantly related to the time duration of nCPAP and infants who were ventilated for a longer period with nCPAP were more likely to develop nasal lesion. This was also found by several other researchers, including Badr et al., Fischer et al., Casey et al. and Yong et al. [2, 31]. Intriguingly, oxygen therapy was also associated with a higher likelihood of nasal injury, which has not been yet described in the literature. This is presumably explained by a weakening of the columella and nasal skin through the pressure induced by the prongs used to supplement the neonates, which has a combined effect on the regional fragility.

Hence the nasal columella breakdown risk associated with nCPAP use, a written and clear nursing protocol on how to apply the best preventive and therapeutic measures to preterm infants must be implemented in NICUs to minimize or revert this complication. Through a sequence of evidence-based steps, helpful to the homogenization of the procedures in the unit and to the time of intervention, beneficial effects on the incidence of nasal skin trauma could be achieved. As management strategies, we suggest a restrict use of lubricants for the initial positioning of the prongs (besides a few drops of saline), due to their friction effect which impairs skin integrity [32], as well as skin protection with hydrocolloid or silicone barriers, maintenance of a dry area and assessments every 4 to 6 hours with repositioning of the neonate for pressure points relief and gentle massages of such points [5]. As previously stated, short binasal prongs are considered the first-line interface to use in preterm neonates with respiratory distress, because of the reduced resistance they provide to gas flow and diminished work of breathing, and for being less invasive than nasopharyngeal tubes. However, one of the most imperative protective procedures to avoid skin breakdown is the alternation between nasal prongs and mask once the first signs of nasal trauma are recognized, for it shifts the pressure points on the nasal columella [5]. As detailed by several other and previously mentioned studies, such as Fischer et al. and Yong et al. [8, 31], this is one of the most clinically efficient steps to avoid iatrogenic injury in pressure areas, which should be of primary importance in the nursing care setup. In the context of this study, the comparison between both interfaces on the nasal injury incidence was not aimed, because the alternation was pre-emptively adopted by the nursing team. For further investigation, we would suggest the conduction of an ethically-approved prospective study in order to assess if there is, in the Portuguese neonatal population, any clinical and/or statistical difference on the incidence of nasal complications between a group whose ventilatory interface is nasal prongs, a second group ventilated through solely nasal mask and a third group whose interface is alternated if any signs of skin compromising due to pressure is detected by the nursing team.

Two other proposed alternatives which could help reduce the incidence of nasal complications due to non-invasive respiratory support in the neonatal intensive care unit would be: 1) Hood CPAP system, based on its high tolerability and good pressure transmission without the risk on inappropriate positioning of nasal prongs thereby avoiding nasal trauma [33]; and 2) Nasal High-Frequency Ventilation (nHFV), an increasingly popular mode used as an alternative to nCPAP in preterm infants which requires smaller prongs and a looser fixation into the nares, reducing the rates of nasal injury in this population [1].

Thus far, there are limited reports in the literature concerning nasal injury complications. Previous to the classification suggested by Fischer et al., descriptions and definitions of nasal trauma were highly variable; this adaptation of US National Pressure Ulcer Advisory Panel classification, however, uniformizes the descriptive registry, if implemented in the NICU setting. Once our nursing team is not attuned to this codification, interindividual variability regarding the description of clinically identified nasal injuries cannot be disregarded. In light of this, we consider beneficial to the nasal injury incidence if a formative process to the nursing team is carried out in our local NICU, to acquaint it with the same classification and, therefore, create more homogenous electronic records, helpful for the ensuing assessment and management.

The use of the same nCPAP system during the study strengthens its internal validity, even if it limits its generalizability. Other limitations of this study are to be a retrospective one, to have a small size sample and to represent a single center experience.

5 Conclusion

Nasal injury due to nCPAP use is relatively frequent in very low birth weight preterm infants. Even with skilled nursing care, involving monitoring of the skin and use of protective barriers, it can be difficult to prevent nasal breakdown. Since the duration of nCPAP ventilation and the use of oxygen therapy present as significant risk factors for the onset of trauma, it is strongly advised to limit the time on nasal prongs, in order to prevent this potentially critical situation. While on nCPAP, relieving the pressure is key to healing and prevention and prominence should be given to the correct sizing of binasal prongs and alternation between nasal prongs and mask if any signs of skin lesion are identified.

Disclosure statements

André Guimarães does not have conflicts of interest. Gustavo Rocha does not have conflicts of interest. Hercília Guimarães does not have conflict of interest.

The authors did not receive funding to write this paper from any entity in the public of private domains.

Manuela Rodrigues does not have any conflict of interest.