Prevalence and outcomes of air leak syndrome and subtypes in neonates in the United States

Abstract

Objectives

To describe the prevalence, risk factors, and outcomes for neonatal air leak syndrome and its subtypes (pneumothorax, pneumomediastinum, pneumopericardium, pulmonary interstitial emphysema, and pneumoperitoneum), evaluate variables associated with the development of air leak, and analyze the national trend of neonatal air leak syndrome in the United States.

Methods

A retrospective cross-sectional analysis of neonates was performed using the Kids’ Inpatient Database. Univariate and multivariable analyses were used to compare neonates with and without air leak syndrome and its subtypes using the 2016 and 2019 data. Data from 1997 to 2019 were used for trend analysis.

Results

Of 7.7 million neonatal discharges, 41,814 developed air leak syndrome (5.41/1,000 discharges) in 2016 and 2019, with a mortality rate of 7.2%. Pneumothorax was the most common air leak syndrome (4.3/1,000 discharges). The risk of air leak syndrome increased with decreasing gestational age on univariate analysis. On multivariable analysis, gestational age had a variable effect on developing air leaks depending on the type of air leak syndrome. The presence of meconium aspiration syndrome, congenital diaphragmatic hernia, persistent pulmonary hypertension, and the use of invasive mechanical ventilation were associated with an increased risk of air leak syndrome. There was an increased linear trend in neonatal air leak syndrome prevalence in the United States from 1997 to 2019.

Conclusions

Air leak syndrome remains a serious and significant issue among the neonatal population and is associated with high morbidity and mortality. We present a national prevalence and outcomes of various neonatal air leak syndromes in the United States.

Keywords

air leak neonate pneumomediastinum pneumopericardium pneumothorax pulmonary interstitial emphysema

Introduction

Air leak syndrome represents a spectrum of phenomena caused by the rupture of alveoli wherein escaped air from the tracheobronchial tree collects in body spaces where it is not ordinarily present.^1,2 Neonatal air leak syndrome encompasses pneumothorax, pneumomediastinum, pneumopericardium, and pulmonary interstitial emphysema (PIE).^2,3 Commonly, air leak syndrome results from barotrauma or volutrauma on injured lungs, with a higher incidence occurring in the neonatal period compared to other age groups.^2,4 However, the development of air leak syndrome in neonates is often multifactorial regardless of the presence or absence of mechanical ventilation.⁵ Multiple studies have identified the presence of underlying lung pathology, gestational age, male sex, and exposure to positive pressure ventilation as contributing factors.^3–7

Over the past 50 years, clinical practice changes in the management of neonates, both term and preterm, have led to a decreased incidence of neonatal air leak syndrome in some reports.^2,3 Examples of these changes include the use of nasal continuous positive airway pressure, exogenous surfactant therapy, antenatal steroid administration, and practice changes in both conventional and high-frequency oscillatory ventilation.^2,8 Despite these changes, neonatal air leak syndrome remains a significant cause of morbidity and mortality.^6,7 A wide range of incidences of air leak syndrome have been reported in the past.⁹ The rate of pneumothorax in neonates admitted to intensive care units participating in the Canadian Neonatal Network was unchanged from 2005 to 2011, with bimodal peaks at term and 32 weeks gestational age.¹⁰ A review of a single institution’s perinatal data over 10 years showed that the incidence of pneumothorax was 0.27% in neonates who weighed more than 2,500 grams and 2.5% in neonates who weighed less than 2,500 grams.¹¹ In a regional study from Denmark over 9 years, the incidence of pneumothorax in live births was 0.14%, with transient tachypnea of the newborn being the most common underlying respiratory illness.¹² In a previous study from the Kids’ Inpatient Database (KID), the prevalence of pneumothorax was 0.034% among neonates from 2006 to 2012, with the highest prevalence at ≤24 weeks gestation (0.67%) and lowest at term (0.02%).¹³ The overall mortality rate of neonates with pneumothorax was 8.8%, and greater in preterm (16.3%) versus term infants (2.7%).¹³ The incidence of asymptomatic pneumothorax may be higher than has been reported.⁹

Mortality rates for neonatal air leak syndrome range from 10% to 68%. However, mortality rates vary by subtype.^4,6,7 To date, many studies have focused on one subtype of air leak syndrome, with pneumothorax being the most studied. The primary objective of our study aimed to utilize data collected from a large national discharge database to describe the prevalence and outcomes of neonatal air leak syndrome and its subtypes during 2016 and 2019 and to evaluate variables associated with air leak syndrome and their outcome. The secondary objective was to evaluate the trend in the prevalence of neonatal air leak syndrome from 1997 to 2019 in the United States.

Methods

We performed a retrospective cross-sectional analysis of neonates utilizing the Healthcare Cost and Utilization Project’s (HCUP) Kids’ Inpatient Database (KID) for 2016 and 2019. HCUP has published the KID every 3 years since 1997. The KID is an ideal database to identify, track, and analyze demographic data, outcomes, and trends for a broad range of pediatric conditions and procedures on a national level.¹⁴ The KID for the year 2019 includes data from 4,000 US community hospitals (defined as short-term, non-Federal, general, and specialty hospitals, excluding hospital units of other institutions), excluding rehabilitation hospitals. The KID for the year 2016 includes data from 4200 community hospitals. The KID 2019 includes data from 48 states plus the District of Columbia, while the data in KID 2016 includes data from 46 states and the District of Columbia. Normal newborns were sampled at a rate of 10%, while complicated newborns were sampled at 80%. The KID includes pediatric discharges of all payers, including children covered by Medicaid, private insurance, and uninsured children. For trend analysis of neonatal air leak syndrome prevalence, we used KID databases from 1997 to 2019 (1997, 2000, 2003, 2006, 2009, 2012, 2016, and 2019). International Classification of Diseases 9th revision (ICD-9-CM) groups neonatal pneumothorax, pneumomediastinum, pneumpericardium, and PIE into one code (770.2). Hence, we were only able to trend neonatal air leak syndrome and not the subtypes.

We included all neonates, defined as patients aged 0–28 days at hospital admission. Exclusion criteria included patients with more than 28 days of life at admission. We identified neonates with air leak syndromes using ICD diagnosis codes. ICD-10-CM codes P25.0 through P25.8 were used to identify air leak subtypes originating in the perinatal period and J93 through J93.9 for other air leak subtypes. ICD-10-CM codes for PIE and related conditions, P25.0, P25.8, and J98.2, were combined under PIE. For patients with multiple air leak syndrome diagnoses, for example, pneumothorax and PIE, the patient was grouped into the most severe form of air leak based on the case fatality rates. We used ICD-9 codes to identify variables in KID databases before 2016. There are no specific ICD-9 codes for subtypes of neonatal air leak syndrome. Hence, we used aggregate air leak syndrome from 1997 to 2019 for trend analysis. The Clinical Classification Software (CCS) condenses several ICD-10 and ICD-9 diagnosis and procedure codes into a smaller set of clinically relevant groups and is used when appropriate.¹⁵ Other associated conditions, procedures, and risk factors were identified using relevant ICD-10-CM, ICD-10-PCS, or CCS codes. Gestational age was subdivided by weeks of gestation into extremely preterm (<28 weeks), very preterm (28 to <32 weeks), moderate preterm (32 to <34 weeks), late preterm (34 to <37 weeks), and term (≥37 weeks).^16,17 Morbidity (non-routine discharges) was defined as those patients whose discharge status included disposition to a short-term hospital, skilled nursing facility, or with home health services, per HCUP data elements classifications.¹⁸

The KID combines race and ethnicity data into one variable. We have regrouped race/ethnicity data into White, Black, Hispanic, and others (Asian and Pacific Islanders, Native Americans, and unspecified). The payor status was regrouped into government insurance, private insurance, and other forms of coverage. Median household income is categorized into quartiles based on estimated income data from residents’ ZIP codes. These quartiles range from 1 (lowest) to 4 (highest), indicating the poorest to wealthiest populations, with the income ranges for each quartile varying annually due to updates in ZIP code-based demographic information.¹⁹ The KID reports total charges as submitted by the source hospital, excluding the professional fees and non-covered charges. Total charges were adjusted for inflation to 2019 values.²⁰

Statistical analysis

All analyses employed complex sampling and data weighting for national estimates. Demographic, clinical, and outcome variables were compared between neonates with and without air leak syndromes. The outcomes analyzed were length of stay, total charges, and mortality. The demographic characteristics, hospital type, gestational age, various neonatal disease processes, use of mechanical ventilation, and mortality were compared between neonates with and without air leaks using chi-square tests. The Mann–Whitney U test was used to compare continuous variables between the two groups. Chi-square and Kruskal–Wallis tests were used to compare categorical and continuous variables among air leak syndrome subtypes. Multivariable analyses were performed to assess the risk for air leaks in all neonates and the risk factors for mortality in neonates with air leak syndrome. In all multivariable analyses, a priori selected independent variables included in the models were gestational age groups, various pulmonary diseases (respiratory distress syndrome (RDS), meconium aspiration syndrome (MAS), congenital diaphragmatic hernia (CDH), persistent pulmonary hypertension (PPHN), and transient tachypnea of newborn (TTN)), any operating procedure for cardiac conditions, and invasive and non-invasive mechanical ventilation. In the binary regression analysis, to determine the mortality risk in neonates with air leak, we included various subtypes of air leaks in the model in addition to the independent variables mentioned above. The model’s overall performance is presented as chi-square (df) for significance, and Nagelkerke’s R² is used to explain the variation. The risks for the outcome variable are presented as adjusted odds ratio and 95% confidence intervals (CIs).

Categorical data are summarized using numerical values and percentages with 95% CI, while continuous data are depicted using medians and interquartile ranges (IQRs). Two-by-two analyses and regression results are presented as p-values alongside odds ratios (ORs) and 95% CI. For multiple analyses, the significance level was reported after the Bonferroni correction. The extended Mantel–Haenszel test was used for linear trend analysis. The linear trend analysis was performed using Epi Info™ (CDC, Atlanta, GA). IBM^® SPSS^® Statistics version 28.0 (IBM, Armonk, New York) was used for all other statistical analyses. The Institutional Review Board approved the study as exempt. All analyses were performed per the HCUP guidelines, and any data with values <11 patients were not reported.

Results

Prevalence

We analyzed 41,814 neonates with air leak syndrome out of 7.7 million neonatal discharges during 2016 and 2019. The overall prevalence of neonatal air leak syndrome in 2016 and 2019 was 5.41 per 1,000 discharges in the United States (Table 1).

Table 1.

Prevalence and in-hospital mortality of neonatal air leak syndrome.

	Total patients (n = 44,814)	Prevalence per 1000 discharges (95% CI)	Mortality % (95% CI)	p value*
Pneumothorax	33,140	4.30 (4.23–4.34)	6.47 (6.17–6.78)	<0.001
Pneumomediastinum	5537	0.72 (0.69–0.74)	3.42 (2.91–4.01)
Pulmonary interstitial emphysema	2869	0.37 (0.36–0.39)	20.72 (19.06–22.48)
Pneumopericardium	268	0.03 (0.03–0.04)	23.60 (18.25–29.95)
Any air leak	44,814	5.41 (5.35–5.47)	7.2 (6.9–7.4)	<0.001**

*Indicates the p values belong to mortality; ** compared to no air leaks.

Demographic characteristics

Males comprised 66.3% of neonates with air leak syndrome compared to 51.4% of total discharges among neonates without air leak (OR: 1.86; 95% CI: 1.82–1.90). The prevalence of air leak syndrome was 6.97/1,000 discharges in males compared to 3.77/1,000 discharges in females. The prevalence of neonatal air leak syndrome was higher in children’s and urban teaching hospitals. Table 2 provides a comparison of demographic data between those neonates with and without air leak syndrome.

Table 2.

Demographic characteristics of neonates with and without any air leak syndrome.

Variable	Neonates with air leak (n = 41,814)		Neonates without air leak (n = 7,690,200)		Significance
Variable	%	95% CI	%	95% CI	Significance
Sex (%)					1.86 (1.82–1.90)
Male	66.3	65.8–66.8	51.4	51.3–51.5	<0.001
Race/ethnicity (%)
White	52.1	51.6–52.7	46.3	46.2–46.4
Black	8.9	8.6–9.2	13.3	13.2–13.4
Hispanic	16.2	15.8–16.6	18.1	18.0–18.2	<0.001
Other	22.8	22.3–23.3	22.3	22.2–22.4
Payor (%)
Government	45.2	44.7–45.8	45.7	45.6–45.8
Private	47.0	46.4–47.5	46.4	46.3–46.5	0.112
Other	7.8	7.5–8.1	7.9	7.8–7.9
Median household income
1st quartile	27.4	26.9–27.9	28.4	28.3–28.4	p < 0.001
2nd quartile	25.4	24.9–25.8	24.6	24.5–24.6
3rd quartile	25.7	25.2–26.1	25.0	24.9–25.1
4th quartile	21,6	21.1–22.0	22.1	22.0–22.2
Hospital mortality rate (%)	7.2	6.9–7.4	0.31	0.31–0.32	24.4 (23.5–25.4) < 0.001
Length of stay (days)	6 (3–17)		2 (2,3)		<0.001
Charges ($)	53,770 (20,472–176,059)		4,678 (3,062–8256)		<0.001

Air leak syndromes and gestation age

Pneumothorax was the most common air leak syndrome, with a prevalence of 4.3/1,000 hospital discharges. The prevalences by subtype of air leak syndrome are presented in Table 1. By univariate analysis, the prevalence of air leak syndrome was significantly associated with earlier gestational age and pathological disease processes related to prematurity. The prevalence of air leak syndrome increased significantly with lower gestational age (Figure 1). The prevalence of each subtype of air leak syndrome also increased as gestation age decreased. The prevalence of air leak syndrome was highest in the extremely premature population, accounting for 101.2/1,000 discharges (Table 3).

Figure 1.

The prevalence of various neonatal air leak syndromes by gestational age in the United States during 2016 and 2019. Prevalence per 1,000 discharges of neonatal air leak syndrome by gestational age and subtype. Extremely preterm: <28 weeks gestation; very preterm: 28 to <32 weeks gestation; moderate preterm: 32 to <34 weeks gestation; late preterm: 34 to <37 weeks gestation; term: ≥37 weeks gestation. The values for pneumomediastinum and pneumopericardium are shown on the secondary axis.

Table 3.

Prevalence of air leak syndrome subtype per 1,000 discharges by gestational age.

	Extremely preterm (n = 43,726)	Very preterm (n = 77,831)	Moderate preterm (n = 98,317)	Late preterm (n = 512,387)	Term (n = 6,999,752)	All (n = 7,732,014)
Pneumothorax	52.1 (49.8–54.6)	23.6 (22.5–24.9)	12.9 (12.2–13.8)	10.2 (9.9–10.5)	3.2 (3.2–3.2)	4.3 (4.2–4.3)
Pneumomediastinum	5.4 (4.6–6.2)	2.9 (2.5–3.4)	2.3 (2.0–2.7)	2.1 (1.9–2.2)	0.5 (0.5–0.6)	0.7 (0.7–0.7)
Pulmonary interstitial emphysema	42.3 (40.2–44.5)	4.48 (3.9–5.0)	0.78 (0.61–1.01)	0.18 (0.14–0.23)	0.07 (0.06–0.08)	0.37 (0.36–0.39)
Pneumopericardium	1.37 (1.03–1.83)	0.39 (0.26–0.59)	0.11 (0.05–0.22)	0.07 (0.05–0.1)	0.02 (0.02–0.02)	0.03 (0.03–0.04)
Any air leak	101.2 (98–104.5)	31.4 (30.1–32.9)	16.1 (15.3–17.1)	12.5 (12.2–12.9)	3.8 (3.8–3.9)	5.4 (5.3–5.5)

Subtypes of air leak syndromes

The data from 2016 to 2019 was used to analyze subtypes of air leaks. Pneumothorax was present in 36,670 neonates out of 7,732,014 neonatal discharges. Descriptive data of various subtypes of air leak syndromes are presented in Table 4.

Table 4.

Characteristics and outcomes of neonatal air leak syndrome subtypes in the United States during 2016 and 2019.

Variable	Pneumothorax	Pneumomediastinum	Pulmonary interstitial emphysema	Pneumopericardium
Number (95% CI)	36,670 (36,241–37,099)	5747 (5577–5917)	2869 (2748–2989)	277 (240–315)
Gestational age	—	—	—	—
Term	67.2 (66.7–67.8)	67.2 (65.7–68.5)	17.4 (15.9–19.1)	47.9 (41.2–54.8)
Late preterm	15.7 (15.3–16.1)	19.3 (18.2–20.5)	3.2 (2.5–4.0)	12.8 (8.8–18.2)
Moderate preterm	3.9 (3.7–4.1)	4.3 (3.7–4.9)	2.7 (2.1–3.5)	3.9 (1.9–7.6)
Very preterm	5.7 (5.5–5.6)	4.4 (3.8–5.0)	12.2 (10.8–13.6)	11.9 (8.1–17.1)
Extremely preterm	7.5 (7.2–7.8)	4.9 (4.3–5.6)	64.5 (62.5–66.5)	23.5 (18.2–29.8)
Associated other subtypes of air leaks	9.6 (9.3–10.0)	50.2 (48.7–51.7)	28.3 (26.5–30.3)	51.6 (44.7–58.3)
Pneumothorax	—	48.7 (47.2–50.1)	26.0 (24.2–27.9)	40.7 (34.1–47.5)
Pneumomediastinum	7.6 (7.3–7.9)	—	5.5 (4.7–6.6)	20.0 (15.1–26.0)
Pulmonary interstitial emphysema	2.0 (1.9–2.2)	2.8 (2.3–3.3)	—	**
Pneumopericardium	0.3 (0.2–0.4)	1.0 (0.7–1.3)	**	—
Invasive mechanical ventilation	37.3 (36.7–37.8)	35.9 (34.5–37.3)	89.9 (88.5–91.1)	68.2 (61.5–74.1)
Non-invasive mechanical ventilation	31.0 (30.5–31.6)	35.4 (34.0–36.9)	25.4 (23.6–27.3)	22.1 (17.0–28.3)
Mortality rate	6.7 (6.4–7.0)	3.8 (3.3–4.5)	20.7 (19.1–22.5)	23.9 (18.5–30.2)

The values are presented as % (95% confidence intervals). ** Too small (<11) to report.

Extreme preterm: <28 weeks gestation; very preterm: 28 to <32 weeks gestation; moderate preterm: 32 to <34 weeks gestation; late preterm: 34 to <37 weeks gestation; term: ≥37 weeks gestation.

Associated risk factors for developing air leaks

The adjusted odds ratios of developing various air leak syndromes are presented in Table 5. The adjusted odds of association between gestational age and air leak syndrome varied with the type of air leak syndrome (Table 5).

Table 5.

Variables associated with adjusted risk for developing air leak syndrome and its subtypes.

	Air leak syndrome	Pneumothorax	Pneumomediastinum	Pneumopericardium	Interstitial emphysema
Gestational age
Term	Ref.	Ref.	Ref.	Ref.	Ref.
Late preterm	0.96 (0.93–0.99)	0.92 (0.89–0.95)	0.97 (0.90–1.05)	1.25 (0.83–1.88)	0.69 (0.54–0.87)
Moderate preterm	0.44 (0.42–0.47)	0.43 (0.40–0.45)	0.40 (0.35–0.46)	0.77 (0.39–1.50)	1.10 (0.85–1.43)
Very preterm	0.38 (0.36–0.40)	0.35 (0.33–0.37)	0.23 (0.20–0.26)	1.24 (0.77–2.01)	2.50 (2.10–2.98)
Extremely preterm	0.91 (0.87–0.95)	0.57 (0.54–0.60)	0.33 (0.28–0.37)	2.28 (1.51–3.45)	12.34 (10.67–14.28)
Associated diagnoses
Neonatal RDS	0.85 (0.81–0.89)	0.86 (0.82–0.91)	1.07 (0.96–1.19)	0.60 (0.34–1.05)	1.46 (1.10–1.92)
MAS	7.12 (6.78–7.47)	7.24 (6.88–7.62)	7.97 (7.08–8.97)	3.79 (2.18–6.60)	1.22 (0.92–1.61)
CDH	3.90 (3.46–4.41)	4.47 (3.95–5.05)	1.16 (0.73–1.84)	4.90 (2.35–10.21)	0.51 (0.23–1.09)
PPHN	2.66 (2.55–2.78)	2.48 (2.37–2.60)	2.12 (1.90–2.37)	2.64 (1.84–3.81)	3.31 (2.98–3.68)
TTN	1.11 (1.06–1.16)	1.08 (1.03–1.14)	1.14 (1.02–1.27)	0.71 (0.39–1.28)	0.96 (0.75–1.21)
Cardiac surgical procedures	2.14 (1.96–2.34)	2.14 (1.95–2.35)	1.20 (0.91–1.59)	3.34 (1.88–5.94)	1.74 (1.35–2.23)
Positive pressure ventilation
NIV	2.92 (2.85–3.00)	3.05 (2.97–3.14)	3.30 (3.10–3.52)	1.17 (0.86–1.58)	1.02 (0.93–1.11)
IMV	8.85 (8.59–9.13)	8.48 (8.21–8.76)	6.90 (6.39–7.46)	32.55 (22.31–47.48)	64.55 (54.37–76.62)
Chi-square (df)	111,683 (12)	92,741 (12)	15,752 (12)	1199 (12)	21,635 (12)
Nagelkerke R²	0.221	0.204	0.168	0.192	0.425
Length of stay (days)**	6 (3–17)	6 (3–14)	6 (3–14)	9 (2–27)	74 (19–121)
Charges ($)**	53,770 (20,472–176,059)	46,615 (18,461–137,934)	53,759 (23,289–132,695)	96,935 (32,504–392,112)	673,018 (239,446–1,299,506)

Extremely preterm: <28 weeks gestation; very preterm: 28 to <32 weeks gestation; moderate preterm: 32 to <34 weeks gestation; late preterm: 34 to <37 weeks gestation; term: ≥37 weeks gestation. CDH = congenital diaphragmatic hernia. IMV = invasive mechanical ventilation, MAS = meconium aspiration syndrome. NIV = non-invasive ventilation, PPHN = persistent pulmonary hypertension of the newborn, RDS = respiratory distress syndrome, TTN = transient tachypnea of the newborn.

**Continuous variables are presented as median (interquartile range), p < 0.001. The adjusted risk is presented as an odds ratio (95% CI).

Outcomes

The length of stay and hospital charges were higher in neonates with air leak syndrome (Table 2) and differed significantly among air leak syndrome subtypes (Table 5). Compared to neonates without air leak syndrome, routine discharge was significantly lower (74.1% vs 96.2%), while morbidity (18.7% vs 3.5%) and mortality (7.2% vs 0.3%; OR: 24.4; 95% CI: 23.5–25.4) were significantly higher in neonates with air leak syndrome. Males with air leak syndrome had a lower mortality risk than females (6.3% vs 8.7%; OR 0.71; 95% CI: 0.65–0.77). In univariate analysis, the mortality rate varied significantly with the air leak syndrome subtype (Table 1). The highest mortality rate among subtypes was pneumopericardium at 23.6%, although it had the lowest prevalence at 0.034/1,000 discharges. PIE had the second-highest mortality rate at 20.7%, with a prevalence of 0.37/1,000 discharges.

Logistic regression analysis determined the variables associated with mortality in neonates with air leak syndromes (Table 6). The adjusted risk of any mortality was significantly increased with lower gestational age compared with term gestation and with invasive mechanical ventilation, CDH, PPHN, and cardiac surgical procedures. It was lower with non-invasive ventilation, MAS, and TTN (Table 6).

Table 6.

Variables associated with adjusted risk of mortality in neonatal air leak syndrome.

	Adjusted odds ratio (95% CI)	p value
Gestational age
Term	Ref.
Late preterm	1.24 (1.07–1.44)	<0.004
Moderate preterm	1.89 (1.53–2.33)	0.001
Very preterm	2.49 (2.12–2.93)	<0.001
Extremely preterm	7.53 (6.63–8.54)	<0.001
Associated diagnoses
Neonatal RDS	1.31 (1.00–1.71)	0.053
Meconium	0.40 (0.31–0.52)	<0.001
Congenital diaphragmatic hernia	2.56 (2.0–3.29)	<0.001
PPHN	1.97 (1.76–2.19)	<0.001
TTN	0.17 (0.11–0.26)	<0.001
Cardiac surgical procedures	1.61 (1.27–2.03)	<0.001
Positive pressure ventilation
NIV	0.19 (0.16–0.22)	<0.001
IMV	9.02 (7.74–10.50)	<0.001
Air leak syndromes
Pneumothorax	Ref.
Pneumomediastinum	0.63 (0.53–0.74)	<0.001
Pneumopericardium	1.99 (1.42–2.8)	<0.001
Interstitial emphysema	0.66 (0.58–0.75)	<0.001
Chi-square (df)	6915 (15)	<0.001
Nagelkerke R²	0.379

IMV = invasive mechanical ventilation. NIV = non-invasive ventilation. PPHN = persistent pulmonary hypertension of the newborn. RDS = respiratory distress syndrome. TTN = transient tachypnea of newborn. Extreme preterm: <28 weeks gestation; very preterm: 28 to <32 weeks gestation; moderate preterm: 32 to <34 weeks gestation; late preterm: 34 to <37 weeks gestation; term: ≥37 weeks gestation.

Trend analysis

A linear trend analysis was performed to determine the prevalence of air leak syndrome in all neonates from 1997 to 2019. The prevalence ranged from 5.1 to 5.8 neonates with air leak syndrome for every 1,000 neonatal discharges. During the analysis period, the linear trend for neonatal air leak syndrome prevalence increased significantly (Figure 2).

Figure 2.

The prevalence of neonatal air leak syndrome from 1997 to 2019 in the United States. Prevalence trend of neonatal air leak syndrome from 1997 to 2019.

Discussion

Our study from a national discharge database illustrates that air leak syndrome remains pervasive in the neonatal population and is also increasing, with an overall prevalence of 0.53%. In this analysis, we present the national prevalence of neonatal air leak syndrome, various subtypes of air leak syndromes, and the risk factors associated with the development of neonatal air leak syndromes and associated outcomes. Neonatal air leak syndrome has been well described in the medical literature for decades. While there is data to suggest that the prevalence of air leak syndrome was higher in previous decades, we report an increase in prevalence over the last 2 decades.^2,3,5

The study by Acun et al. using the KID database for the years 2006, 2009, and 2012 reported a prevalence rate of 0.034% compared to the prevalence in our report of 0.53%.¹³ Several methodologic differences may account for the difference in the reported prevalence of air leak syndromes. We have used all ICD codes for pneumothorax, including neonatal air leaks. The previous study did not report the specific codes used to identify pneumothorax but may have only used neonatal air leaks (ICD-9 code 770.2). In addition, we have included all neonates in our cohort, whereas the authors have excluded neonates with congenital anomalies of the diaphragm, lung, and abdominal wall in the previous study.

Prior studies have shown an increased risk of neonatal air leak syndromes in the male population compared to females, particularly pneumothorax, similar to our study.^4,7,21,22 Our data supports this relationship across all neonatal air leak syndrome subtypes. Male gender is known to have a higher risk for mortality and morbidity in those born prematurely.²² Although the risk of developing air leak syndrome was higher in male neonates, the associated mortality risk was lower in univariate analysis in our study.

Gestational age and risk factors

Prematurity is a well-documented risk factor associated with the development of air leak syndrome.^1,4,7,23,24 Our data supports that of previously published literature that air leak syndrome and all subtypes of air leak syndromes are more prevalent in premature than in term babies. This increased risk has been associated with numerous factors, including underlying pulmonary pathology associated with prematurity, such as RDS, MAS, and PPHN. The literature documents the association between underlying pulmonary pathology and the development of air leak syndrome.^4,7 Our data support prior publications and show a statistically significant increased risk for air leak syndrome among neonates with the presence of MAS and CDH. In addition, the need for positive pressure ventilation is also associated with higher odds of developing neonatal air leak syndrome, as noted in our study.^4,5,7,21 However, when adjusted for various risk factors such as underlying pulmonary pathology and the use of mechanical ventilation, the associated risk of gestational age for developing air leaks depends on the type of air leak syndrome.

As we have noted, PIE is more likely to develop in premature babies. In contrast, pneumothorax or any air leak syndrome is more likely to develop in a term baby after adjusting for other risk factors. The discordance of the effect of gestational age on the development of air leaks in univariate and multivariable analyses may be due to the risk factors included in the regression models. For example, positive pressure ventilation in the delivery room is a known risk factor for developing pneumothorax in term and near-term neonates.²⁵ Delivery by elective caesarian section is also known to be associated with developing pneumothorax in term babies.^26,27 Spontaneous pneumothorax at the time of delivery due to high pressure generated during initial breaths is known to occur.²⁸ Due to a lack of documentation, we could not include the use of positive pressure ventilation in the delivery room and elective caesarian section in our model. RDS and the use of invasive mechanical ventilation are strong risk factors for premature babies developing air leaks. We included those variables in our regression analyses.

Air leak syndrome subtypes

In our study, pneumothorax and pneumomediastinum had similar characteristics regarding gestational age, the use of mechanical ventilation, and mortality risk. Similar to pneumothorax, pneumomediastinum is known to occur spontaneously after birth.^23,29 Pathophysiologically and epidemiologically, pneumomediastinum and pneumothorax behave similarly, as seen in our study.

PIE is uncommon and can occur at any age, but it is most frequent in neonates.³⁰ As our study shows, the prevalence of PIE is inversely related to gestational age.³¹ The risk factors for pulmonary PIE in preterm infants include higher oxygen use during resuscitation and increased surfactant use, as well as high ventilatory pressures.³² PIE continues to occur in premature babies who are mechanically ventilated despite advancements in neonatal management.

The diagnosis of PIE is fraught with inaccuracies.³⁰ PIE can be difficult to differentiate from lucent bronchiole overdistention. PIE can also be misinterpreted as pulmonary edema or aspiration syndrome. Air bronchograms of respiratory distress syndrome may be misdiagnosed as PIE. The airways and alveoli can appear distended in neonates on mechanical ventilators, as in PIE.

National trend

Surfactant therapy has been shown to improve pneumothorax in premature babies with RDS.^33–35 There are several reports of decreased prevalence of air leaks after surfactant therapy was approved. Our study period is several decades after surfactant therapy is in use. We find that air leak prevalence has increased nationally from 1997 to 2019. We could not find a similar study comparing the prevalence of air leaks at a national level before the surfactant was used. In a study from the Canadian Neonatal Network, the rate of pneumothorax in neonates admitted to intensive care units from 2005 to 2011 was unchanged.¹⁰ Although there is improved neonatal care, such as ventilator management, the increased trend in air leaks may be due to an increase in survival of extremely preterm and other contributing perinatal management practices. This is an area that needs further exploration.

Limitations

Our study has several limitations. Like many administrative databases, the KID is subject to coding errors, including incorrect coding and omission of relevant codes. The data provided by these databases often lack granularity by nature of how and what information is entered into the databases. Laboratory and physiologic data are not included in the KID. Several perinatal variables that may influence the development of air leaks, such as delivery room resuscitation and vigorous suctioning efforts, are not coded and were not used in the analyses. The KID does not provide temporal associations, so we could not evaluate temporal relationships. Our data may not apply to neonates outside of the United States. Due to transfers across the institutions, there is a potential for double counting of neonates with air leaks.

Conclusion

In summary, air leak syndrome remains a serious and significant medical condition among the neonatal population and is associated with high morbidity and mortality. Our national study presents the prevalence of air leak syndrome and its subtypes at different gestational ages, the risk factors for developing air leaks, and associated mortality. Since 1997, the prevalence trend of air leak syndrome in neonates has increased significantly in the United States, an area that needs further exploration.

Statements and declarations

Footnotes

Author contributions

Adam Beaton: conceptualization, formal analysis, and writing—original draft preparation. Prithvi Sendi: data curation, formal analysis, and writing—reviewing and editing. Paul Martinez: writing—reviewing and editing. Balagangadhar Totapally: conceptualization, formal analysis, supervision, and writing—reviewing and editing.

Conflicting interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Adam Beaton

Prithvi Sendi

Balagangadhar R. Totapally

References

Kim

. Development of pulmonary air leak in an extremely-low-birth-weight infant without mechanical ventilation: a case report. Korean J Anesthesiol 2017; 70(2): 217–220.

Jeng

Lee

Tsao

, et al. Neonatal air leak syndrome and the role of high-frequency ventilation in its prevention. J Chin Med Assoc 2012; 75(11): 551–559.

Roychoudhury

Kaur

Soraisham

. Neonatal pneumopericardium in a nonventilated term infant: a case report and review of the literature. Case Rep Pediatr 2017; 2017: 3149370.

Ilçe

Gündogdu

Kara

, et al. Which patients are at risk? Evaluation of the morbility and mortality in newborn pneumothorax. Indian Pediatr 2003; 40(4): 325–328.

Niwas

Nadroo

Sutija

, et al. Malposition of endotracheal tube: association with pneumothorax in ventilated neonates. Arch Dis Child Fetal Neonatal Ed 2007; 92(3): F233–F234.

Bhat Yellanthoor

Ramdas

. Frequency and intensive care related risk factors of pneumothorax in ventilated neonates. Pulm Med 2014; 2014: 727323.

Parekh

Maguire

Emery

, et al. Pneumothorax in neonates: complication during endotracheal intubation, diagnosis, and management. J Anaesthesiol Clin Pharmacol 2016; 32(3): 397–399.

Reuter

Moser

Baack

. Respiratory distress in the newborn. Pediatr Rev 2014; 35(10): 417–428.

Montero Gato

Sacristán

Vázquez

, et al. Incidence of ultrasonographic signs of pneumothorax in asymptomatic neonates. Pediatr Pulmonol 2023; 58(6): 1691–1696.

10.

Duong

Mirea

Shah

, et al. Pneumothorax in neonates: trends, predictors and outcomes. J Neonatal Perinat Med 2014; 7(1): 29–38.

11.

Aly

Massaro

Acun

, et al. Pneumothorax in the newborn: clinical presentation, risk factors and outcomes. J Matern Fetal Neonatal Med 2014; 27(4): 402–406.

12.

Vibede

Bendtsen

, et al. Neonatal pneumothorax: a descriptive regional Danish study. Neonatology 2017; 111(4): 303–308.

13.

Acun

Nusairat

Kadri

, et al. Pneumothorax prevalence and mortality per gestational age in the newborn. Pediatr Pulmonol 2021; 56(8): 2583–2588.

14.

Agency for Healthcare Research and Quality . HCUP databases. Healthcare cost and utilization project (HCUP) 2022. Available from: https://hcup-us.ahrq.gov/kidoverview.jsp

15.

Agency for Healthcare Research and Quality . Clinical classifications software (CCS) for ICD-10-PCS (beta version). Healthcare cost and utilization project (HCUP) 2019. Available from: https://www.hcup-us.ahrq.gov/toolssoftware/ccs10/ccs10.jsp

16.

Shapiro-Mendoza

Lackritz

. Epidemiology of late and moderate preterm birth. Semin Fetal Neonatal Med 2012; 17(3): 120–125.

17.

Raju

Higgins

Stark

, et al. Optimizing care and outcome for late-preterm (near-term) infants: a summary of the workshop sponsored by the National Institute of Child Health and Human Development. Pediatrics 2006; 118(3): 1207–1214.

18.

Agency for Healthcare Research and Quality . KID description of data elements. Healthcare cost and utilization project (HCUP) 2023. Available from: https://www.hcup-us.ahrq.gov/db/nation/kid/kiddde.jsp

19.

Agency for Healthcare Research and Quality . HCUP KID description of data Elements2008 2024. Available from: https://hcup-us.ahrq.gov/db/vars/zipinc_qrtl/kidnote.jsp

20.

U.S. Bureau of Labor Statistics . Historical consumer price index for all urban consumers (CPI-U): U.S. City average, all items, by month. Archived Consumer Price Index Supplemental Files 2023. Available from: https://www.bls.gov/cpi/tables/supplemental-files/historical-cpi-u-202310.pdf

21.

Hadzic

Skokic

Husaric

, et al. Risk factors and outcome of neonatal pneumothorax in Tuzla Canton. Mater Sociomed 2019; 31(1): 66–70.

22.

Peacock

Marston

Marlow

, et al. Neonatal and infant outcome in boys and girls born very prematurely. Pediatr Res 2012; 71(3): 305–310.

23.

Rocha

Guimaraes

. Spontaneous pneumomediastinum in a term neonate - case report. Clin Case Rep 2018; 6(2): 314–316.

24.

Raissaki

Modatsou

Hatzidaki

. Spontaneous pneumomediastinum in a term newborn: atypical radiographic and ct appearances. BJR Case Rep 2019; 5(4): 20180081.

25.

Kim

E-A

Jung

J-H

Lee

S-Y

, et al. Neonatal pneumothorax in late preterm and full-term newborns with respiratory distress: a single-center experience. Neonatal Med 2022; 29(1): 18–27.

26.

Girard

Sommer

Dahan

, et al. Risk factors for developing pneumothorax in full-term neonates with respiratory distress. Arch Pediatr 2012; 19(4): 368–373.

27.

Hansen

Wisborg

Uldbjerg

, et al. Elective caesarean section and respiratory morbidity in the term and near-term neonate. Acta Obstet Gynecol Scand 2007; 86(4): 389–394.

28.

Ibrahim

CPH

Ganesan

Mann

, et al. Causes and management of pulmonary air leak in newborns. Paediatr Child Health 2009; 19(4): 165–170.

29.

Low

Tan-Kendrick

Loh

, et al. Spontaneous multiloculated multiseptated pneumomediastinum in a newborn baby: the spinnaker sail is rigged-CT features with pathologic correlation. Pediatr Radiol 2003; 33(10): 712–715.

30.

Jalota Sahota

Anjum

31.

Kastein

Coe

. Management of air leaks in the newborn. Adv Neonatal Care 2023; 23(2): 160–166.

32.

Nuñez-Ramiro

Aguar

Cernada

, et al. Oxygen needs during resuscitation and surfactant to achieve stabilisation were independent risks factors for pulmonary interstitial emphysema in preterm infants. Acta Paediatr 2018; 107(1): 28–32.

33.

Long

Corbet

Cotton

, et al. A controlled trial of synthetic surfactant in infants weighing 1250 g or more with respiratory distress syndrome. The American Exosurf Neonatal Study Group I, and the Canadian Exosurf Neonatal Study Group. N Engl J Med 1991; 325(24): 1696–1703.

34.

Horbar

Soll

Sutherland

, et al. A multicenter randomized, placebo-controlled trial of surfactant therapy for respiratory distress syndrome. N Engl J Med 1989; 320(15): 959–965.

35.

Soll

Hoekstra

Fangman

, et al. Multicenter trial of single-dose modified bovine surfactant extract (Survanta) for prevention of respiratory distress syndrome. Ross Collaborative Surfactant Prevention Study Group. Pediatrics 1990; 85(6): 1092–1102.