Abstract
BACKGROUND: In the period immediately after birth, preterm infants are highly susceptible to lung injury. Ventilator-induced lung injury has been recognized as a major contributing factor for bronchopulmonary dysplasia (BPD) in preterm infants. Noninvasive respiratory support (NIRS) could decrease lung injury, and early respiratory support management might affect pulmonary outcomes. We conducted a study to evaluate the changes in early respiratory support management and their impact on respiratory outcome and complications of preterm infants in 3 different time periods over the last 13 years.
METHODS: This study was a retrospective, single-center cohort study. We retrospectively reviewed the medical records of preterm infants < 32 weeks of gestational age born in our hospital from 2007–2020. The study period was divided into three 3-y discrete periods: 2007–2009 (period A), 2013–2015 (period B), and 2018–2020 (period C). Changes in early respiratory support management were assessed in the 3 periods. The outcomes measured included mortality, BPD, other major neonatal complications, initial respiratory support, and duration of mechanical ventilation.
RESULTS: In all, 1,880 clinical records were assessed in our study, with 358 in period A, 825 in period B, and 697 in period C. The use of antenatal corticosteroids increased over time (56.1% in period A, 56.7% in period B, and 74.0% in period C (P < .001). The need for surfactant decreased from 65.6% in period A to 40.7% in period B and 45.9% in period C. Increased utilization of NIRS was associated with decreased invasive mechanical ventilation within 24 h after birth. NIRS only during the hospital stay increased from 22.9% in period A to 36.8% and 45.1% in the latter 2 periods (P < .001). Oxygen therapy duration decreased from 24.3 d in period A to 14.4 d in period B and 17.2 d in period C (P < .001). The overall incidence of BPD was 32.4% in the first period, 23.9% in the second period, and 25.4% in the third period (P < .001). The moderate-to-severe forms of BPD decreased from 12.8% in period A to 7.9% in period B and 7.6% in period C (P = .009). Other neonatal complications, such as pneumothorax, pulmonary hemorrhage, persistent pulmonary hypertension of the newborn, surgical necrotizing enterocolitis, intraventricular hemorrhage grade III/IV, and periventricular leukomalacia, were unchanged among the 3 periods.
CONCLUSIONS: From 2007–2020, respiratory management was characterized by a marked reduction in invasive mechanical ventilation and an increase in the use of NIRS. Changes in early respiratory support management resulted in improved respiratory outcomes with a decrease in the overall incidence of BPD. It is likely that our aim to reduce lung injury by improving our respiratory management has contributed to a favorable outcome.
- early respiratory support management
- preterm infants
- bronchopulmonary dysplasia
- respiratory outcome
- complications
Introduction
Improved survival of very preterm infants might have resulted in increased numbers of patients bronchopulmonary dysplasia (BPD).1,2 The period immediately after birth is a very vulnerable phase, and unnecessary intubation and ventilation can damage the upper airway and cause iatrogenic respiratory distress. The intubation procedure and multiple intubation attempts after birth could also cause increased incidences of significant hypoxemia, bradycardia, and intraventricular hemorrhage (IVH) in the smallest and most immature infants.3,4 To reduce the incidence of BPD and other complications, focus has shifted toward a gentler respiratory approach. Some studies have suggested that the best protection against ventilator-induced lung injury is to avoid intubation by employing noninvasive strategies.5,6 Avoiding intubation and using noninvasive respiratory support (NIRS) modes in preterm infants minimize the risk for lung injury and optimize neonatal outcomes.7 Minimizing the duration of endotracheal intubation or avoiding it completely has been a goal of neonatal intensive care.8
Measures to reduce the level of ventilator-induced lung injury include avoiding mechanical ventilation, for example, through a single course of prenatal corticosteroid therapy, the early administration of exogenous surfactant, the use of new methods for administering surfactant through less invasive surfactant administration or minimally invasive surfactant therapy (MIST), the prophylactic use of NIRS, the use of caffeine therapy as a respiratory stimulant, or the use of ventilatory strategies to mitigate lung injury in those patients who require intubation and mechanical ventilation.9-11 Although there is a growing body of evidence to guide decision-making, there is not consensus on the best treatment approach for early respiratory support management.
Based on the available and advancing insights, respiratory support management in our delivery room (DR) and neonatal ICU (NICU) has changed over the past 13 years. Thus, we hypothesized that the changes in respiratory support strategies would lead to improved lung outcomes and other major complications for preterm infants.
QUICK LOOK
Current Knowledge
Improved survival of very preterm infants might have resulted in increased numbers of patients bronchopulmonary dysplasia (BPD). To reduce the incidence of BPD, focus has shifted toward a gentler respiratory approach. Recognition of the link between prolonged invasive mechanical ventilation and increased respiratory morbidity has led to reduced exposure to invasive mechanical ventilation. Avoiding intubation and using noninvasive respiratory support (NIRS) modes in preterm infants minimize the risk for lung injury and optimize neonatal outcomes.
What This Paper Contributes to Our Knowledge
The changes in early respiratory management of preterm infants introduced at our center over the last decade have resulted in an improvement in respiratory outcomes. With early CPAP and minimally invasive surfactant therapy and the increased use of NIRS, there were decreases in endotracheal intubation, surfactant therapy, and oxygen therapy along with reductions in BPD and mortality.
Methods
Population and Data Collection
This study was a retrospective, single-center study of 3 cohorts of preterm infants who were born in Guangdong Women and Children Hospital, China, from 2007–2020. Our center, a tertiary NICU, is a referral center for neonatal patients in Guangdong Province, with an average of 7,700 admissions per year, including approximately 1,500 referral infants per year. Our center is a national neonatal specialty unit and has 45 neonatologists and 150 nurses. This study reviewed the medical records of very preterm infants of < 32 weeks of gestational age (GA) (hereafter referred to as preterm infants) who were admitted to our NICU by dividing the study into 3 discrete 3-y periods (2007–2009, 2013–2015, 2018–2020). Infants affected by major congenital anomalies, a TORCH infection, hydrops fetalis, and chromosomal anomalies were excluded.
Respiratory support management in our study has changed over the past 13 years. Changes in early respiratory support management were initiated in 2011. In addition to synchronized intermittent mandatory ventilation and nasal CPAP, other invasive mechanical ventilation or NIRS modes have been introduced in our hospital and used. Similarly, the second change also occurred from 2016–2017, including early CPAP in the delivery room, MIST technology, and the mobile NICU equipment. Mobile NICU equipment was a combination of a T-piece ventilator (Neopuff infant resuscitator, Fisher and Paykel, Auckland, New Zealand), binasal prongs, air-oxygen mixer, and Giraffe incubator (GE Healthcare, Madison, Wisconsin), which is to refine precise pressure values and to prevent newborns from losing body heat. Children born in 2010–2012 and 2016–2017 were excluded to ensure that the comparison did not include the transition period after the initiation of the new practices.
The Ethics Review Board of Guangdong Women and Children Hospital approved the study protocol. The data were collected without identification.
Variable Collection
Definitions were used consistently over the study period. The following morbidities were included and diagnosed as follows. The criteria of Jobe and Bancalari were used to diagnose BPD and classify its severity.12 IVH was based on the most severe sonography result during the hospital stay and graded according to the previously published scheme of Papile et al.13 Retinopathy of prematurity (ROP) was rated using the international classification published by the committee for the classification of ROP.14 Necrotizing enterocolitis was defined according to Bell criteria.15 Antenatal corticosteroids were defined as any administration of steroids irrespective of completion of therapy. The definition of major complications was determined as moderate or severe BPD, IVH grade III/IV, periventricular leukomalacia, and ROP treatment.
In addition, detailed information regarding the respiratory support management of the infants was collected. The duration of invasive mechanical ventilation was defined as the cumulative duration of conventional mechanical ventilation and high-frequency oscillatory ventilation (HFOV), and the duration of NIRS was defined as the cumulative duration of noninvasive ventilation (NIV), nasal HFOV, CPAP, and high-flow nasal cannula.
Statistical Analysis
Statistical analyses were performed using SPSS statistical software for Windows (version 26.0, IBM, Armonk, New York). Continuous variables with a symmetric distribution are expressed as the mean SD; those with an asymmetric distribution are expressed as the median (minimum-maximum values), and categorical variables are expressed as frequencies and percentages. The chi-square test or Fisher exact test was used for categorial data and Student t test for noncategorical date. The Kruskal-Wallis H test was used to compare 3 independent samples. Binary logistic regression analysis was performed to analyze the impact of the initial respiratory support mode after birth, invasive mechanical ventilation > 7 d, and NIRS only during the hospital stay on BPD. The logistic regression model was also adjusted by the following factors: GA, birthweight, sex, and delivery mode. Adjusted odds ratios were calculated with 95% CI. We used a forward linear regression model. Probability levels (P value) < .05 were considered significant.
Main Changes in Respiratory Support Practices Over the Study Period
In period A, infants were electively intubated in the delivery room. In period B, preterm infants were managed in the delivery room according to their symptoms and umbilical cord blood gas. Different levels of respiratory support modes and HFOV were use in this period. Exogenous surfactant was also administered mainly through the endotracheal tube. MIST was introduced in this period and was not widely used. In period C, our regular practice for the respiratory support of preterm infants < 30 weeks of GA was modified to prevent tracheal intubation by administering prophylactic CPAP in the delivery room to all spontaneously breathing infants. MIST technology was widely used and routinely performed since preterm infants started NIRS with FIO2 > 0.30 to achieve a target oxygen saturation. Changes in respiratory support management practices in the 3 different periods are summarized in Table 1. Intubation and extubation criteria in the 3 different periods are summarized in Table 2.
Changes in Respiratory Support Management Practices in 3 Different Periods
Intubation and Extubation Criteria in 3 Different Periods
Results
Basic Information and Maternal Characteristics
In total, 1,880 infants met the inclusion criteria, and data were collected on all infants until death or discharge to home, with 358 in period A, 825 in period B, and 697 in period C. There were no significant differences in sex, GA, delivery mode, or first- and fifth-minute Apgar scores. The percentage of mothers who received antenatal corticosteroids increased significantly (56.1% in period A, 56.7% in period B, 74.0% in period C, P < .001). The birth details and maternal characteristics of the 3 cohorts of preterm infants are summarized in Table 3.
Demographic Details and Maternal Characteristics of the 3 Cohorts of Preterm Infants in the Study
Comparison of Respiratory Support Practice in 3 Periods
The need for surfactant decreased from 65.6% in period A to 40.7% in period B and 45.9% in period C. The need for intubation within 24 h after birth decreased over time, whereas the use of NIRS increased from 20.7% to 52.8% and 65.6% in periods A, B, and C, respectively, (P < .001). The initial respiratory care practices are shown in Table 4. Among the use of NIRS within 24 h of birth, 74 (100%) infants were supported with CPAP in period A, whereas 385 (88.3%) infants were supported with NIV and only 8 (1.8%) with CPAP in period B. In period C, 87 (19.0%) infants were supported with nasal HFOV and 340 (74.4%) with NIV. The use of theophylline increased from 78.2% (280/358) in period A to 85.6% (706/825) in period B. In period C, caffeine replaced theophylline as a respiratory stimulant. A total of 90.5% (631/697) of preterm infants was treated with caffeine during this period.
Short-Term Respiratory Care Practice Data Due to Changes in Respiratory Support Management in the 3 Time Periods
We observed a marked reduction in the duration of invasive mechanical ventilation. In addition, the median duration of NIRS also decreased over time. The use of invasive mechanical ventilation for > 7 d decreased from 33.2% in period A to 15.5% and 15.6% in periods B and C, respectively, during the study period. Conversely, the use of NIRS as the sole means of respiratory support during hospitalization increased from 22.9% in period A to 45.1% in period C, and the incidence of exclusive use of NIRS nearly doubled in period C compared to period A. Over the course of their entire hospital stay, the need for oxygen therapy decreased (Table 4).
Comparison of Morbidities in 3 Periods
The incidence of infant morbidities and their temporal trends are reported in Table 5. The incidence of all forms of BPD decreased from 32.4% in period A to 23.9% and 25.4% in periods B and C, respectively. Moderate-to-severe BPD was less common in period C (7.6%) than in period A (12.8%). During the 3 periods, the moderate-to-severe BPD rate decreased over time in the subgroup of infants born at < 28 weeks (35.7% in period A, 21.0% in period B, and 17.6% in period C, respectively) and at 30–31 6/7 weeks (8.8% in period A, 3.3% in period B, and 2.0% in period C, respectively, P < .001) although there was no significant difference in the proportion of infants born at < 28 weeks. The infant mortality rate during hospitalization decreased significantly from 22.9% in period A to 12.6% and 11.6% in period B and period C, respectively, (P < .001).
Trends in the Incidence of Bronchopulmonary Dysplasia and Other Complications for the Study Cohort
The incidence of other neonatal morbidities, such as pneumothorax, pulmonary hemorrhage, persistent pulmonary hypertension of newborns, surgical necrotizing enterocolitis, IVH grade III/IV, and periventricular leukomalacia, was not significantly different in the 3 periods. Among those infants who survived to 36 weeks of post-menstrual age, multivariable analyses indicated that the use of invasive mechanical ventilation > 7 d was associated with an increased risk for BPD. The invasive mechanical ventilation group was associated with an increased risk for all forms of BPD but not moderate-to-severe BPD. Conversely, the exclusive NIRS use during the hospital stay was a protective factor against BPD morbidity. Logistic regression models for the incidence of BPD are showed in Table 6.
Logistic Regression Models for the Incidence of Bronchopulmonary Dysplasia*
Discussion
In our study, changes implemented in the respiratory support management strategy were documented, with significantly decreased intubation in the delivery room and increased use of NIRS. The main differences included early CPAP in the delivery room, different levels of respiratory support modes in the unit, surfactant administration (MIST technology, surfactant dose), and transportation using mobile NICU equipment. Since the introduction of early CPAP and MIST and success in early NIRS, there have been decreases in endotracheal intubation need and the duration of invasive mechanical ventilation, with NIRS considered to be a safe alternative to intubation in preterm infants. Once the alveoli are aerated, breathing requires less effort, and distress symptoms can diminish. Many studies have demonstrated the positive effects of NIRS with NIV/CPAP in preference over intubation for the reduction of respiratory distress, ventilator-induced lung injury, and the prevention of death and BPD.16-22 These results are consistent with our study.
Avoidance of intubation in the delivery room did not increase surfactant use. Compared to period A, the need for surfactant decreased significantly in periods B and C. Not all preterm infants require surfactant treatment, and NIRS is a safe and effective approach.23 Dunn et al24 suggested that the use of early CPAP leads to a reduction in the number of infants who are intubated and receive surfactant. CPAP is an alternative to intubation and surfactant in preterm infants.25 The percentage of infants exposed to antenatal corticosteroids increased significantly. The percentage of infants exposed to antenatal corticosteroids in periods A and B was similar to a study that reported that antenatal corticosteroids were administered to 53.3% of infants born at 24–34 completed weeks in China between 2013–2014.26 The use of antenatal corticosteroids was associated with significant reductions in infant mortality and morbidity.27 Eighty percent of women at high risk for preterm birth in high-income countries currently receive antenatal corticosteroids, and the proportions in middle-income countries range from 30–50%.28 Compared to the high percentage of use of antenatal corticosteroids in high-income countries, the proportion was 74% in period C in our study. To achieve this goal, we will work together with obstetricians to increase the use antenatal corticosteroids for women at high risk for preterm birth.
In addition to early initiation of NIRS, there was an increase in the proportion of preterm infants who were supported exclusively with NIRS during hospitalization. The increased use of NIRS in periods B and C can be attributed to the application of different levels of respiratory support modes and MIST procedures in this population. In our study, NIV and nasal HFOV replaced CPAP in the latter 2 periods as the main NIRS mode in the NICU. It has been suggested that early NIV was more successful than CPAP in preventing intubation and invasive mechanical ventilation among preterm infants with respiratory distress syndrome.21,22,29 Nasal HFOV significantly improved CO2 clearance and reduced the need for intubation compared with CPAP/bi-level positive airway pressure.30 Wang et al reported that MIST could decrease the incidences of intubation with mechanical ventilation and the composite outcome of death or BPD.31 MIST resulted in a rapid and homogenous increase in end-expiratory lung volume, associated with an improvement in oxygenation and a decrease in breathing effort in preterm infants with respiratory distress syndrome.32 We consider that the changed practices of NIRS and the introduction of MIST have become important factors in preventing tracheal intubation and the successful use of NIRS during hospitalization. Not only was there an increase in NIRS associated with a decrease in invasive mechanical ventilation, but there was also a decrease in both oxygen therapy duration, which led to decreases in BPD and mortality, and in moderate-to-severe BPD.
After adjustment for confounders, NIRS only during hospital stay dramatically decreased the relative risk of developing BPD compared with infants offered invasive mechanical ventilation treatment or not maintained successfully on NIRS therapy. Our findings should give impetus to preventing BPD by increasing the use of NIRS. Considered together, we believe that the increasing use of NIRS therapy and decreasing use of invasive mechanical ventilation, as well as lung-protective ventilation strategies, are important factors explaining the lower BPD rate.
Compared to period B, changes in period C contributed to the further increase in the use of NIRS and the reduction in invasive mechanical ventilation in period C, including prophylactic CPAP in the delivery room, widespread application of MIST technology, and mobile NICU equipment to refine precise pressure values and to prevent newborns from losing body heat. Shi et al suggested that, in extremely preterm infants, optimal pulmonary outcomes could be achieved by minimizing the duration of invasive ventilation.7 Even when the duration of invasive mechanical ventilation was shorter in period C, the BPD rate slightly increased from 23.9% in period B to 25.4% in period C (without statistical significance). Moderate-to-severe BPD still decreased over time, especially in the subgroup of infants born at < 28 weeks and 30–31 6/7 weeks, although there was no significant difference in the proportion of infants born at < 28 weeks. Despite continued efforts to limit the duration of invasive respiratory support, the decision to reduce the duration of mechanical ventilation for very preterm infants is always difficult since these infants have immature and injured lungs as well as poor respiratory drive. The optimal respiratory support practices of these infants should be further discussed. Additionally, our data only permitted assessment of the total length of invasive mechanical ventilation. We did not have information about the ventilator modes, the number of ventilation courses, and the timing or the reasons for re-intubation. The outcomes could have been affected by the above factors. Since the cause of BPD is multifactorial, we, therefore, assume that other unmeasured factors influenced the incidence of BPD. Exploring the reasons for this finding is beyond the scope of this retrospective analysis.
Among the morbidities observed in this cohort, other respiratory outcomes, such as pneumothorax, pulmonary hemorrhage, and persistent pulmonary hypertension of newborns, remained static. The incidence of IVH III/IV and the rate of periventricular leukomalacia remained stable over the 3 periods. Although the days of oxygen therapy decreased, the incidence of ROP was increased over time. The impact of this change could not be interpreted from this cohort. Further study is warranted to verify better approaches to decrease the incidence of ROP in preterm infants.
Our study has several limitations. First, due to the retrospective study design, it is difficult to reliably determine how particular changes have contributed to the observed improvements in patient outcomes. Second, some data on the use of other therapies that could influence pulmonary outcomes at least temporarily, such as post-natal corticosteroids, diuretics, bronchodilators, and nutritional management, were not collected. Finally, our study is limited by the lack of information about the long-term outcomes of our infants. Further studies are needed to evaluate the long‐term benefits and risks associated with evolving practices for respiratory support in premature infants. Although we did not aim to compare different noninvasive initial respiratory support techniques with each other and with invasive mechanical ventilation, it is required to accurately determine what proportions of these changes contributed the most and to define the ideal respiratory strategy. We look forward to conducting such comparative studies in the future.
Conclusions
The changes in early respiratory management of preterm infants introduced at our center over the last decade resulted in an improvement in respiratory outcomes. With the introduction of early CPAP and MIST technology and the increased use of NIRS, there were decreases in endotracheal intubation, surfactant therapy, and oxygen therapy along with a significant reduction in BPD. It is likely that our aim to reduce lung injury by improving our respiratory management has contributed to a favorable outcome.
Footnotes
- Correspondence: Weiwei Gao, Department of NICU, Guangdong Women and Children Hospital, No. 521, Xingnan Road, Panyu District, Guangzhou City, Guangdong Province, China. E-mail: mgbbbear{at}qq.com
The authors have disclosed no conflicts of interest.
This research was supported by the Medical Scientific Research Foundation of Guangdong Province, China, Early Respiratory Function and Growth in Preterm Infants With Bronchopulmonary Dysplasia, a Prospective Cohort Study from the Health Commission of Guangdong Province (Grant number A2020224). Funding agencies played no role in the design, collection, analyses, or interpretation of the results of this study.
- Copyright © 2022 by Daedalus Enterprises
