ReviewLung Ultrasound, a Better Choice for Neonatal Pneumothorax: A Systematic Review and Meta-analysis
Introduction
Neonatal pneumothorax (PTX), an air leak that occurs when air accumulates between the parietal and visceral pleura, is a common and critical illness in neonatal intensive care units (NICUs) (Bruschettini et al. 2019). The incidence of PTX is 1%–2% in full-term infants (Al Tawil et al. 2004; Edwards et al. 2013), and because of poor lung compliance, the incidence increases to 6% in premature infants (Horbar et al. 2002). The mortality rate of neonatal PTX is approximately 20% (Joshi et al. 2020). Risk factors for PTX include pre-term birth, mechanical ventilation, low weight, suffocation and meconium inhalation (Boo et al. 2011; Aly et al. 2014; Bhat Yellanthoor and Ramdas 2014; Joshi et al. 2020), among which prematurity and mechanical ventilation are considered the most important. PTX is usually accompanied by signs of respiratory distress; however, >26% of newborns have been reported to have no obvious clinical symptoms at an early stage, making it difficult to make a diagnose (Duong et al. 2014). PTX is a common cause of death in neonates, especially in premature infants (Bhatia et al. 2011; Duong et al. 2014), and can cause hypotension, bradycardia and an increased risk of intraventricular hemorrhage (Pishva et al. 2012). Moreover, it is also a risk factor for cerebral palsy (Aly et al. 2014). Therefore, timely and accurate diagnosis of PTX is of great importance for emergency treatment.
Chest computed tomography (CT) is known as the “gold standard” for the diagnosis of PTX in adults, while it is rarely used in neonates because of its strong ionizing radiation (Abdalla et al. 2016). At present, chest X-ray (CXR) and chest transillumination are the main diagnostic methods for neonatal PTX and are always the reference standards. However, these methods still use ionizing radiation. Compared with other age groups, newborns, especially those who require repeated examinations, are at a greater risk from latent effects of CXR owing to their small size and the close proximity of radiosensitive tissues and organs. Moreover, CXR and chest transillumination have the potential for misdiagnosis; a small PTX may be difficult to detect during CXR examination, especially in low-birth-weight neonates and pre-term infants with young gestational age (Alrajhi et al. 2012; Alrajab et al. 2013; Liu et al. 2017; Deng et al. 2020). Furthermore, CXR is always used to confirm the endotracheal tube position for the infants with continuous positive airways pressure in the NICU who are at greater risk from latent effects compared with other age groups (Hall and Brenner 2012).
In adult critical care medicine, lung ultrasound (LUS) has been widely used in the diagnosis, differentiation and treatment of PTX (Nagarsheth and Kurek 2011; Staub et al. 2018; Han et al. 2020), and some meta-analyses have confirmed that LUS is superior to CXR (Alrajhi et al. 2012; Alrajab et al. 2013). LUS examination is independent of the patient's age (Volpicelli et al. 2012); moreover, anatomic features of newborn's lungs (thin chest wall and narrow thorax) are conducive to LUS imaging, LUS is a new choice for the diagnosis and treatment of neonatal PTX (Bourcier et al. 2014; Brat et al. 2015). Compared with CXR, ultrasound combines the advantages of bedside diagnosis, avoidance of irradiation, cost-effectiveness, high accuracy and reliability (Nagarsheth and Kurek 2011; Liu et al. 2019; Rea et al. 2019).
At the International Liaison Committee on Lung Ultrasound in 2012, a multi-disciplinary panel involving 28 experts from eight countries had developed a consensus on ultrasonic diagnosis of lung disease based on medical evidence (Volpicelli et al. 2012). In recent years, with the development of LUS and advances in knowledge, China and the United States have developed guidelines for neonatal LUS (Kurepa et al. 2018; Liu et al. 2019). Normal ultrasound images show hypoechoic lines, pleural lines and A-lines of clear, smooth, regular, linear high echo and a clearly visible lung slide (Kurepa et al. 2018; Liu et al. 2019). The main LUS manifestations of neonatal PTX include the disappearance of lung slide, the presence of pleural lines and A-lines, absence of B-lines and presence of lung points (Table 1, Fig. 1) (Volpicelli et al. 2012; Liu et al. 2017; Kurepa et al. 2018). Consequently, we aimed to compare the diagnostic accuracy of LUS with that of CXR in neonates with PTX.
Section snippets
Methods
We followed the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019) and completed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist (Liberati et al. 2009).
Identification of studies
We completed our initial search strategy in February 2020 and identified 528 potential articles. There were 112 duplicate studies. Based on manuscript titles and abstracts, we excluded 401 papers. After reviewing the full text of 15 studies that were potentially eligible for inclusion, we found eight studies that met all the inclusion criteria and were considered eligible for meta-analysis (Cattarossi et al. 2016; Raimondi et al. 2016; Liu et al. 2017; Corsini et al. 2019; Grimaldi et al. 2019;
Overall Meta-analysis
Across the eight studies, the pooled sensitivity and specificity for the diagnosis of PTX were 98% (95% CI: 0.94–0.99) and 99% (95% CI: 0.98–1.00), respectively (Table 4, Fig. 4a, 4b). The I2 statistic values were 46.7% and 9.1%, respectively. The DOR was 920.01 (95% CI: 265.81–3184.33), showing a relatively high accuracy of LUS in diagnosing PTX in neonates, with an I2 value of 0.0%. Since these I2 statistic values were <50%, there was considered to be no significant heterogeneity based on the
CXR in diagnosing PTX
Analyzing studies indicated that the sensitivity and specificity of CXR for the diagnosis of PTX were 82% (95% CI: 0.72–0.90) and 96% (95% CI: 0.90–0.99), respectively (Table 5, Fig. 6a, 6b). DOR was 44.54 (95% CI: 4.30–460.98), showing a lower accuracy than LUS in diagnosing PTX.
LUS in diagnosing PTX except CXR as the single diagnosis criteria
Under this condition, analyzing studies indicated that the sensitivity and specificity of LUS for the diagnosis of PTX were 98% (95% CI: 0.93–1.00) and 100% (95% CI: 0.96–1.00) (Table 5, Fig. 7a, 7b). DOR was 965.39
CXR for diagnosing PTX
CXR is associated with a certain rate of misdiagnosis and is less sensitive than LUS for the diagnosis of mild-to-moderate PTX (Deng et al. 2020; Grimaldi et al. 2019), especially in premature infants (Cattarossi et al. 2016). It may miss 30% of patients with PTX in the supine position, while in the upright position, it can reduce the rate of misdiagnosis. However, it is almost impossible to carry out in NICU newborns (Abdalla et al. 2016). Other possible reasons include small lesions,
Review Limitations
This review has some limitations. The characteristics of ultrasound examinations were extremely heterogeneous across the studies, with different machines, probes and operator expertise levels. Because of the lack of prospective studies on the analysis of ultrasonic signs (presence of A-lines or lung point, absence of lung sliding or B-lines), a reasonable and effective meta-analysis on these factors could not be conducted.
Conclusions
LUS is an emerging technology that has better sensitivity and specificity than CXR in the diagnosis of PTX. Under the precondition of no diagnosis of pulmonary consolidation, some ultrasonic signs (absence of lung sliding or B-lines) had a high sensitivity in the diagnosis of PTX. The lung point could help to determine the severity of PTX. Its presence indicates that PTX is mild to moderate; otherwise, PTX will be severe.
Acknowledgments
We gratefully acknowledge Yu Hui Min, Prof., Dr., from the Department of Neonatology, the Children's Hospital of Zhejiang University School of Medicine, for supervising our work, reviewing the manuscript and for conceptual advice.
Conflict of interest disclosure
None of the authors have any financial relationships relevant to this article or potential conflicts of interest to disclose. No external funding was received for this paper.
References (42)
- et al.
Test characteristics of ultrasonography for the detection of pneumothorax: A systematic review and meta-analysis
Chest
(2012) - et al.
Performance comparison of lung ultrasound and chest x-ray for the diagnosis of pneumonia in the ED
Am J Emerg Med
(2014) - et al.
Meta-analysis in clinical trials
Control Clin Trials
(1986) - et al.
Respiratory distress of the term newborn infant
Paediatr Respir Rev
(2013) - et al.
Thoracic ultrasound accuracy for the investigation of initial neonatal respiratory distress
Arch Pediatr
(2019) - et al.
The diagnosis of neonatal pulmonary atelectasis using lung ultrasonography
Chest
(2015) - et al.
Lung ultrasonography to diagnose pneumothorax of the newborn
Am J Emerg Med
(2017) - et al.
Chest ultrasonography for the emergency diagnosis of traumatic pneumothorax and haemothorax: A systematic review and meta-analysis
Injury
(2018) - et al.
Lung ultrasound versus chest radiography for the diagnosis of pneumothorax in critically ill patients: A prospective, single-blind study
Saudi J Anaesth
(2016) - et al.
Symptomatic spontaneous pneumothorax in term newborn infants
Pediatr Pulmonol
(2004)