Variables related to bronchopulmonary dysplasia severity: a Six-Year retrospective study

Abstract Objectives This was a retrospective observational study conducted in a tertiary neonatal intensive care unit, in order to investigate factors which influenced the severity of bronchopulmonary dysplasia under NICHD new classification. Methods Six years of clinical data with different grades of bronchopulmonary dysplasia patients were collected and analyzed, bivariate ordinal logistic regression model and multivariable ordinal logistic regression model were used with sensitivity analyses. Results We identified seven variables were associated with the severity of BPD via a bivariate ordinal logistic regression model, including the level of referral hospital (OR 0.273；95% CI 0.117, 0.636), method of caffeine administration (OR 00.418；95% CI 0.177, 0.991), more than two occurrences of reintubation (OR 4.925；95% CI 1.878, 12.915), CPAP reapplication (OR 2.255；95% CI 1.059, 4.802), presence of positive sputum cultures (OR 2.574；95% CI 1.200, 5.519), the cumulative duration of invasive ventilation (OR 1.047；95% CI 1.017, 1.078), and postmenstrual age at the discontinuation of oxygen supplementation (OR 1.190；95% CI 1.027, 1.38). These seven variables were further analyzed via all multivariable ordinal logistic regression models, and we found that tertiary hospital birth and early administration of caffeine could reduce the severity of BPD by approximately 70% (OR 0.263；95% CI 0.090, 0.770) and 60% (OR 0.371；95% CI 0.138, 0.995), respectively. In contrast, multiple reintubations were related to higher BPD severity with an OR of 3.358 (95% CI 1.002, 11.252). Conclusion Improving perinatal care in level II hospitals, standardized caffeine administration, and optimized extubation strategy could potentially decrease the severity of BPD.


Introduction
In 1967, Northway and colleagues introduced the definition of bronchopulmonary dysplasia (BPD), which was described as a lung injury in preterm infants resulting from oxygen and mechanical ventilation [1].Half a century later, the pathophysiology of BPD was recognized as the result of an aberrant reparative response to both antenatal and repetitive postnatal injury to the developing lungs [2], and it has become one of the most common adverse outcomes of premature infants [3].Unlike other prematurity-related morbidities, the overall incidence of BPD remains constant at around 40% for preterm infants of 28 weeks gestational age (GA) [4].BPD is independently associated with long-term neurological impairment and poor respiratory outcomes [5].The cost of caring for these patients is an enormous burden on both families and society [6].
It is generally acknowledged that the severity of BPD is independently related to both morbidity and mortality, as well as to post-discharge use of the healthcare system, and these outcomes present in a stepwise manner as BPD severity increases [7][8][9].With the development of neonatology, the management of premature newborns has made great advances; however, minimizing the injury to these immature lungs and optimizing medical interventions remains a challenge.
In 2018, the National Institute of Child Health and Human Development (NICHD) put forward new BPD diagnostic criteria with classification for patients who CONTACT Xiaoying Li sdeyneo@163.comDepartment of Neonatology, Children's Hospital Affiliated to Shandong University, Jinan, Shandong, P.R.

China
Supplemental data for this article can be accessed online at https://doi.org/10.1080/14767058.2023.2248335.
were born with GA of 32 weeks.Subsequently, Chinese data showed that these new criteria had a better correlation with the observed morbidity and mortality [10].However, we continue to lack an optimal prediction model for the development of BPD with good predictive accuracy and calibration [11][12][13], not to mention a prediction model of BPD severity.Since clinical analysis is aimed to identify influencing factors that could contribute to the development of reliable prediction models, patients who develop BPD could benefit from interventions that could reduce the severity of BPD [14].We summarized and analyzed the data of Chinese patients with BPD according to this new classification, trying to recognize influencing factors relating to the severity of BPD.This approach should optimize medical care and interventions for patients with more severe BPD, in order to improve outcomes in this population.

Study subjects
This was a retrospective observational study in the neonatal intensive care unit (NICU) of the Children's Hospital Affiliated with Shandong University, between October 2015 and October 2021.The criteria for inclusion were as follows: (i) Gestational age (GA) less than 32 weeks; (ii) Admitted to our NICU within seven days of life; (iii) Development of BPD according to the diagnostic criteria of the NICHD 2018 classification system.The exclusion criteria included major congenital or genetic abnormalities or death before postmenstrual age (PMA) of 36 weeks GA.Enrolled patients were assigned to three groups graded to correlate with increasing severity of BPD based on the NICHD 2018 classifications: group 1 (Grade I/mild), group 2 (Grade II/moderate), and group 3 (Grade III/severe).

BPD definition and classification
According to the NICHD recommendations [15] for preterm infants born at a GA of less than 32 weeks, BPD was defined as: evidence of persistent parenchymal lung disease, radiographic confirmation of parenchymal lung disease, and the requirement at 36 weeks PMA of FiO 2 ranges/oxygen levels/O 2 concentrations for greater than or equal to three consecutive days to maintain arterial oxygen saturations between 90-95% (Table 1).We did not include the Grade IIIA patients as we were unable to obtain all the clinical data required for a complete analysis of this group.

Data collection
Demographic information collected included gender, GA, birth weight (BW) and age at the time of admission.Antenatal characteristics assessed included maternal age, the incidence of maternal gestational diabetes mellitus (GDM), the incidence of maternal pregnancy-induced hypertension (PIH), evidence of maternal infection prior to delivery (defined as fever or other infectious diseases), incidence of premature rupture of membranes (PROM), the incidence of stained amniotic fluid, frequency of administration of antenatal steroids, the incidence of multiple pregnancies (including twins and triplets), mode of delivery (spontaneous vaginal delivery (SVD) versus cesarean section (C-section)), Apgar score at 5 min of age, a size small for gestational age (SGA, based on Fenton growth chart), and the level of birth hospitals (based on Chinese national health commission website: nhc.gov.cn).Clinical characteristics during hospitalization included documentation of FiO 2 at the time of initial respiratory support, use and number of administrations of pulmonary surfactant (PS), use and time of initiation of caffeine therapy, the requirement of highfrequency oscillation ventilation (HFOV) mode as a rescue strategy, cumulative duration of mechanical ventilation (MV), total number of reintubations, reapplication of CPAP, patent ductus arteriosus (PDA), persistent pulmonary hypertension of the newborn (PPHN), and the occurrence of positive sputum culture.Early administration of caffeine was defined as administration of caffeine within the first 2 days of life.Age and PMA at the discontinuation of oxygen supplementation were assessed to evaluate respiratory outcomes.

Statistical analysis
Categorical variables are expressed as counts and percentages, which were compared across the three groups using the chi-squared test or Fisher's exact test.Continuous variables are expressed as the mean-± standard deviation or median (interquartile range [IQR]) in cases of non-normal distribution.The normality of distribution was checked graphically using the Shapiro-Wilk test.Normally distributed continuous variables were compared using analysis of variance; while non-normally distributed variables were compared using the Kruskal-Wallis H test.
The association between clinical variables and the BPD grading were examined using a bivariate ordinal logistic regression model with proportional odds ratios (OR).The proportional odds hypothesis was tested using a score test.Variables associated with BPD grade in bivariate analyses (p < 0.05) were included in a multivariable ordinal logistic regression model (all models) with a proportional OR.
Sensitivity analyses were conducted to assess the robustness and reliability of the multivariate ordinal logistic regression model.We combined the groups of Grade II and III patients, and compared these with the group of Grade I patients via univariate binary logistic regression.Variables with p < 0.05 in the three-group bivariate ordinal logistic regression model were further analyzed via multivariate binary logistic regressions to assess whether the results were consistent with the three-group results.
All statistical analyses were conducted using R version 4.1.1 and statistical significance was defined as by the presence of a P value of ＜0.05.

Results
A total of 1544 infants with GA of ＜32 weeks were admitted to our NICU between October 2015 and October 2021, and 108 of them (7%) developed BPD, meeting the inclusion criteria.These infants with BPD were divided into three groups of patients assessed as Grade I, II, and III in severity, with percentages of 57%, 30%, and 13%, respectively (Figure 1).All continuous variables were normally distributed, except for birth weight and maternal age (supplementary files, Stable 1).
The lowest gestational age was 25 þ1 weeks with a mean GA of 28.78 weeks (±1.56), and the lowest birth weight was 730 g with a median birth weight of 1150 g (970, 1330).The number of male patients was double that of female patients.Half of Grade I infants were born in a level III hospital, while this percentage was reduced to 25% and 14.3% in the Grade II and III groups, respectively (Table 2).
The clinical characteristics and respiratory outcomes of infants with BPD of different severities are detailed in Table 3.Most infants with BPD (71.3%) required invasive ventilation for their initial respiratory support after birth, and Grade III patients required the highest percentage of HFOV as a rescue strategy.Caffeine was administered to 88.9% of infants (96/108), of which 71.9% received it within the first two days of life.Grade III patients had the lowest percentage of early caffeine administration among the three groups.Approximately half of all patients with BPD experienced extubation failure (EF) at least once, and half of patients with Grade III BPD required reintubation more than twice, which was significantly higher than that in patients with Grade I BPD.As the BPD grade increased, the cumulative duration of MV correspondingly increased.Grade III BPD infants were more likely to develop dyspnea post-CPAP, requiring reapplication of CPAP, and were not able to discontinue respiratory support until they were at an increased PMA.Echocardiography was performed within 2 days of birth.Although most patients (80.6%) had a PDA, only 17.6% developed PPHN at the early stage.The percentage of positive sputum culture(s) became higher as the severity of BPD increased.
All variables met the proportional odds hypothesis (supplementary files, S-table 2); therefore, we conducted a bivariate ordinal logistic regression to investigate the relationship between variables and BPD severity (Table 4).The results demonstrated that there were seven variables associated with the BPD severity, including the level of a referral hospital, timing of caffeine administration, multiple reintubations, CPAP reapplication, presence of positive sputum culture(s), cumulative duration of MV, and PMA when oxygen supplementation was discontinued.Infants who were born in a level III hospital and received caffeine therapy within two days of life were at a lower risk of developing Grade II/III BPD.In contrast, the rest five factors were related to an increased risk of severe BPD.
The multicollinearity of these seven influencing factors (p < 0.05) of BPD severity was tested (supplementary files, S-table 3), and a multivariable ordinal logistic regression analysis was conducted (Table 5).More than two occurrences of reintubation were significantly associated with BPD severity (OR 3.358；95% CI 1.002, 11.252).It was observed that a higher-level birth hospital could decrease the severity of BPD by more than 70% (95% CI 0.09, 0.77).Similarly, compared with late administration of caffeine, early caffeine administration was another protective factor, with an OR of 0.371 (95% CI 0.138, 0.995).
Since Grade II and Grade III patients with BPD are reported to have higher mortality and different clinical courses than Grade I patients [15], we combined the patients of the Grade II and III groups and compared them with the Grade I group to test the validity of our observations (Table 6).The results were the same as those of the three-group analysis, which verified that multiple reintubations, tertiary hospital births, and early caffeine administration were independent influencing factors of the severity of BPD.

Discussion
Data were collected from a level III NICU over a period of six years, with clinical analysis based on patients with BPD at ＜32 weeks of gestation.Seven influencing factors were related to the severity of BPD, among which multiple reintubations were identified as significant risk factor for BPD severity.In contrast, tertiary    hospital births and early administration of caffeine could significantly reduce the severity of BPD.MV exposure and duration have been identified as important risk factors in previous studies [16][17][18].In this study, we found that patients with more severe BPD required a longer cumulative duration of invasive ventilation and higher PMA to discontinue oxygen supplementation.These patients also had a higher chance of experiencing multiple EF, reapplication of CPAP, and positive sputum cultures.Although all these factors were related to the severity of BPD, multiple EFs (more than twice) were an independent risk factor with nearly 3.4-fold increased odds.
Mounting evidence suggests that early extubation is associated with a shorter duration of MV, a lower incidence of BPD, and a shorter length of hospital stay [19][20][21].However, reintubation is associated with higher mortality and incidence of BPD [22], leading to a dilemma of early extubation and subsequent EF.Successful extubation is normally defined as not requiring reintubation during a pre-specified window of observation; however, this time window varies from 12 h to 7 days [23].Previous studies have found that EF is generally accompanied by a greater need for respiratory support, longer duration of oxygen therapy, increased risk of BPD, and longer length of hospital stay [24], with the greatest risk attributable to reintubation within the first 48 h after extubation [25].
Multiple factors contribute to EF, including low GA, low birth weight, maternal chorioamnionitis, high pressure during MV prior to extubation, low pH in pre-and post-extubation blood gases, low hemoglobin before extubation, and a low incidence of advanced noninvasive ventilation strategies (e.g.nasal intermittent positive pressure ventilation) [26][27][28].
The extubation policy of infants with GA less than 32 weeks in our center include: the patient is clinically stable for the past 24 h with good spontaneous breath; PIP is less than 15 H2O and FiO2 is less than 0.3; transcutaneous SpO2 can be maintained between 90-94% with arterial blood gas in the normal range.The extubation readiness and indications of reintubation need to be assessed by attending physicians.All patients were extubated to either CPAP or NIPPV as attending physicians' evaluations.In this study, 52.8% of patients with BPD required reintubation.We set the observation window at 48 h post-extubation, and the reintubation incidence of Grade III BPD was 70%, which was significantly higher than that in Grade I and II groups (43.3% and 41.2%, respectively).Noteworthily, the severity of BPD was related to the recurrence of extubation failure, and clinical approaches to increase the success of extubation are necessary for reducing BPD severity.
The hospital level was an independent factor influencing BPD severity in this study, and the results  In the present study, caffeine was another protective factor against BPD severity.The largest randomized controlled trial of caffeine was published in 2006, which determined the efficacy and safety of caffeine therapy in very low birth weight infants [29].In the past decade, caffeine has become one of the five most commonly prescribed medications in the NICU [30].Numerous studies have recommended early administration of high doses of caffeine in preterm infants [31][32][33].We instituted a prophylactic caffeine protocol in our NICU for infants with GA of ＜32 weeks and/or a birth weight of ＜1500 g, which consisted of a 20 mg/kg caffeine citrate loading dose followed by a maintenance dose of 10 mg/kg/day.Caffeine therapy was initiated immediately after birth.Twelve enrolled patients did not receive caffeine due to personal reasons.Among the 96 patients who received caffeine therapy, 28 were not prescribed caffeine within two days of life, which was a result of the lack of routine caffeine therapy for this population in some referral hospitals.Of the patients who did receive caffeine therapy in this study, all followed the above protocol, and BPD severity was subsequently reduced by approximately 70% compared with late administration.Based on these results, we emphasize the importance of early administration of caffeine in premature infants.
An interesting result in this study was that neither GA nor birth weight affected BPD severity.We analyzed the distribution of GA in these three groups and found that 60.2% of patients in this study were born with GA of 28 -30 weeks, which corresponded to the data of the Chinese Neonatal Network [34], and reflected the major preterm population in Chinese tertiary hospitals over the past decade.Because of the small sample size of extremely low-GA infants, GA did not play an essential role in this study.In addition, only 12% of the patients were SGA infants, birth weight could not become an influencing factor as the result showed.

Limitations and further study prospect
This study has several limitations.Firstly, the occurrence of multiple reintubations was an independent risk factor for BPD severity; however, we did not document the reasons for reintubation.Secondly, initial sufficient respiratory support with a pulmonary protective strategy is crucial to the respiratory outcomes of premature infants, we may need to analyze patients who were transferred to our center earlier (e.g.within 48 h of age) to verify these results with more homogeneous subsequent medical interventions.As mentioned above, the aim of this study was to investigate influencing factors of BPD severity based on previous experiences and supply evidence for more optimal medical care.We have established the provincial quality control center of neonatal care, aiming at optimizing perinatal management of premature infants in the whole province.This study was a pilot study of the respiratory aspect for further research.For example, factors leading to reintubation will be investigated in a subsequent study and a quality improvement project will be conducted to reduce the incidence of EF.

Conclusion
Our data suggest that tertiary hospital care and early administration of caffeine could favor less severity of BPD, which highlights the necessity to improve medical service in district or country-level hospitals.On the other hand, the occurrence of multiple reintubations is an independent risk factor for increased severity of BPD, leading to the requirement of a comprehensive evaluation and further research to reduce extubation failure.Although it is challenging to avoid all the risk factors, this study reminds Neonatologists to think more about these specific factors, trying to provide more optimal health care for premature infants.

Ethical approval
The study has been approved by the Ethics Committee of Children's Hospital affiliated to Shandong University (approval number: SDFE-IRB/T-2022011).

Figure 1 .
Figure 1.Flowchart of the study cohort.

Table 1 .
Definition and classification of BPD-NICHD 2018 recommendation.Grades Invasive IPPV Ã N-CPAP, NIPPV, or nasal cannula !3L/min Nasal cannula flow of 1-<3L/min Hood O 2 Nasal cannula flow of <1L/min Early death (between 14 days of postnatal age and 36 weeks) owing to persistent parenchymal lung disease and respiratory failure that cannot be attributable to other neonatal morbidities (e.g.necrotizing enterocolitis, intraventricular hemorrhage, redirection of care, episodes of sepsis, etc).Ã Excluding infants ventilated for primary airway disease or central respiratory control conditions.Values are percents.CPAP: continuous positive airway pressure; IPPV: intermittent positive pressure ventilation; N-CPAP: nasal continuous positive airway pressure; NIPPV: noninvasive positive pressure ventilation.

Table 2 .
Demographics, antenatal characteristics of BPD patients in different BPD grade data expressed in numbers (percentage) or mean (standard deviation) or median (interquartile range).GA: gestational age; BW: birth weight; GDM: gestational diabetes mellitus; PIH: pregnancy-induced hypertension; PROM: premature rupture of membranes; SGA: small for gestational age.Evidence of maternal infection: fever or other infectious diseases.Level of birth hospital: classified based on the Chinese national health commission website (nhc.gov.cn).

Table 3 .
Clinical characteristics of BPD patients in different grades data expressed in numbers (percentage) or mean (standard deviation) or median (interquartile range).PS: pulmonary surfactant; HFOV: high-frequency oscillation ventilation; CPAP: continuous positive airway pressure; MV: mechanical ventilation; PPHN: persistent pulmonary hypertension of the newborn; PDA: patent ductus arteriosus; PMA: postmenstrual age; HLOS: length of stay in hospital.Early administration of caffeine: administration of caffeine within the first 2 days of life.

Table 4 .
Relevant factors of BPD severity.
GA: gestational age; BW: birth weight; GDM: gestational diabetes mellitus; PIH: pregnancy-induced hypertension; PROM: premature rupture of membranes; SGA: small for gestational age; MV: mechanical ventilation; HFOV: high frequency oscillation ventilation; CPAP: continuous positive airway pressure; PPHN: persistent pulmonary hypertension of the newborn; PDA: patent ductus arteriosus; PMA: postmenstrual age; HLOS: length of stay in hospital.Evidence of maternal infection: fever or other infectious diseases.Tertiary hospital: classified based on Chinese national health commission website (nhc.gov.cn).Early administration of caffeine: administration of caffeine within the first 2 days of life.

Table 5 .
Multivariate ordinal logistic regression analysis of possible influencing factors for BPD severity.The variables of P value < 0.05 were shown by boldface.CPAP: continuous positive airway pressure; MV: mechanical ventilation; PMA: postmenstrual age;.Tertiary hospital: classified based on Chinese national health commission website (nhc.gov.cn).Early administration of caffeine: administration of caffeine within the first 2 days of life.

Table 6 .
Sensitive analyze of possible influence factors for BPD severity.The variables of P value < 0.05 were shown by boldface.CPAP: continuous positive airway pressure; MV: mechanical ventilation; PMA: postmenstrual age;.Tertiary hospital: classified based on Chinese national health commission website (nhc.gov.cn).Early administration of caffeine: administration of caffeine within the first 2 days of life.