Combining lung ultrasound and SNAPPE-II score to predict the severity and prognostic value of neonatal respiratory distress syndrome: an investigational study

Posted on by admin

RESEARCH ARTICLE

Hippokratia 2025, 29(1): 6-12

Ha XM1, Huang CY1,2, Zhang SQ1
1Department of Ultrasound, Yantaishan Hospital
2Medical Impact and Nuclear Medicine Program, Binzhou Medical University
Yantai, China

Abstract

Objective: This study aimed to investigate the usefulness of combining lung ultrasound (LUS) score with Score for Neonatal Acute Physiology Perinatal Extension II (SNAPPE-II) in assessing the severity and prognosis of neonatal respiratory distress syndrome (NRDS).

Methods: This retrospective study included 130 neonates, in total, diagnosed with NRDS between May 2020 and November 2022, and among them, 108 neonates showed improvement in their health status after treatment and were classified into the survival group, while 22 neonates did not survive, despite receiving treatment, and were classified into the non-survival group. Pearson correlation analysis evaluated the association between the SNAPPE-II and LUS scores in NRDS, and we used the receiver operating characteristic (ROC) curve to calculate the sensitivity, threshold, and specificity of NRDS.

Results: The non-survival group exhibited significantly higher LUS (27 ± 5.43) and SNAPPE-II (33.46 ± 7.35) scores compared to the survival group (p <0.001). LUS, SNAPPE-II, and their combined diagnostic methods had statistical significance in NRDS (p <0.001). ROC analysis showed that the cutoff value of the LUS score was 22.5, with a specificity of 0.696, sensitivity of 0.926, an area under the curve (AUC) of 0.892 [95 % confidence interval (CI): 0.838-0.946], and a Youden index of 0.703. The cutoff value for the SNAPPE-II score was 20.2, with a specificity of 0.826, sensitivity of 0.676, an AUC of 0.831 (95 % CI: 0.768-0.895), and a Youden index of 0.622. Notably, the combined diagnostic method using both scoring methods showed significantly better performance than using either method alone, with an AUC of 0.934.

Conclusion: The combination of LUS and SNAPPE-II scores exhibits excellent performance in diagnosing NRDS, providing better patient guidance and improving survival probability. HIPPOKRATIA 2025, 29 (1):6-12.

Keywords: Lung ultrasound score, SNAPPE-II, diagnosis, prognosis, neonatal respiratory distress syndrome

Corresponding author: Shaoqin Zhang, PhD, Department of Ultrasound, Yantaishan Hospital, Yantai, Shandong, China, e-mail: jsenwill@126.com

Introduction

Neonatal respiratory distress syndrome (NRDS) is a severe neonatal condition characterized by respiratory distress and severe hypoxemia, posing a significant threat to the neonate respiratory system1. Clinical manifestations include respiratory distress, inspiratory grunting, and wheezing, often accompanied by excessive ventilation, respiratory acidosis, and wheezing2. Early diagnosis and prompt implementation of appropriate interventions are critical in reducing the high mortality rate associated with NRDS.

Lung ultrasound (LUS) has emerged as a valuable diagnostic modality for neonatal respiratory diseases, including neonatal pneumonia, NRDS, and meconium aspiration syndrome. Due to the thin chest wall and small lung volume of neonates, LUS can detect pleural and lung lesions effectively. In recent years, LUS has complemented the diagnosis of lung diseases and has become an essential diagnostic tool for neonatal lung conditions in the neonatal intensive care unit (NICU)3. LUS has demonstrated higher accuracy and specificity in diagnosing NRDS compared to chest X-rays. Iovine et al4 reported that LUS enables rapid bedside evaluation of various pulmonary and pleural disorders in neonates. Its advantages include low cost, lack of ionizing radiation, and availability for bedside use. These features have significantly improved the diagnosis and prognosis of neonatal respiratory diseases, leading to substantial social and economic benefits. However, despite its advantages, Liu et al5 pointed out that LUS has limitations and cannot entirely replace chest X-rays in clinical practice. For example, LUS may have difficulty identifying specific abnormalities located deep in the lung parenchyma or outside the pleural space.

The Score for Neonatal Acute Physiology Perinatal Extension II (SNAPPE-II) scoring system is another widely used tool for assessing the severity of neonatal diseases and predicting mortality risk in neonates. This simplified scoring system, which includes nine indicators such as the lowest body temperature, lowest mean arterial pressure, and arterial oxygen tension to inspired oxygen fraction ratio, has gained international recognition for its utility in neonatal care6,7. The SNAPPE-II scoring system allows clinicians to evaluate the overall condition of neonates and provides theoretical guidance for prognosis and treatment strategies in conditions such as NRDS.

Various neonatal risk scoring systems have been developed and are currently being applied to predict neonatal disease incidence and case fatality rates8. Both LUS and SNAPPE-II scores have been widely used independently in previous studies to aid early intervention for NRDS. LUS score has proven effective for the diagnosis and initial assessment of NRDS, while the SNAPPE-II score has been employed to evaluate the severity and prognosis of neonatal diseases, including NRDS. However, few researchers have investigated the combined use of LUS and SNAPPE-II in assessing the severity and prognosis of NRDS9. The potential advantages of integrating these two tools include leveraging the diagnostic precision of LUS and the prognostic capabilities of SNAPPE-II to provide a more comprehensive assessment. This gap in the literature highlights the need for further research to explore the complementary roles of LUS and SNAPPE-II scores in NRDS management.

In this study, clinical data from NRDS patients admitted to our hospital over the past two years were analyzed to evaluate the combined value of LUS and SNAPPE-II scoring in assessing the severity and prognosis of NRDS. By addressing this gap, we aim to provide new insights for the severity assessment and prognosis prediction of NRDS.

Materials and Methods

Study design and patients

This retrospective observational study included 130 neonates diagnosed with NRDS in our institute between May 2020 and November 2022. The focus of this study was on neonates diagnosed with NRDS, considering their discharge or death after successful treatment as the respective endpoints. The study was approved by the Clinical Ethics Committee of Yantaishan Hospital, Shandong Province, China (Ethical Review No. 20220001), and was conducted according to the principles of the Declaration of Helsinki throughout. Informed consent was obtained from each child’s guardian, and all family members explicitly agreed to the anonymous use of the neonate’s personal and medical data, including demographics, gestational age, clinical presentation, ultrasound findings, and diagnostic results, provided they met the prespecified inclusion criteria.

The inclusion criteria for the study were: i) diagnosis of NRDS based on clinical manifestations, arterial blood gas analysis, and chest X-ray findings; ii) neonates between 6 and 12 hours after birth who exhibit symptoms such as severe breathing difficulties, cyanosis, moaning during exhalation, and three concave inspirations; iii) blood gas analysis results indicating hypoxemia [defined as arterial partial pressure of oxygen (PaO2) <60 mmHg] or hypercapnia [defined as partial pressure of carbon dioxide (PCO2) >50 mmHg]; iv) chest X-ray examinations revealed reduced lung transparency, increased lung texture, presence of granular and reticular shadows on the lung wall with average dispersion, blurred heart margin, and diaphragm margin, and bronchial inflation sign; v) availability of complete clinical medical records for the neonates.

The exclusion criteria for the study were: i) neonates with congenital diseases, such as complex congenital heart disease, congenital malformations of the respiratory system, chromosomal abnormalities, or congenital mental disorders; ii) Neonates displaying symptoms of persistent pulmonary hypertension or cardiogenic pulmonary edema; iii) neonates with lung sepsis, septic shock, severe intracranial hemorrhage, or severe hypoxic-ischemic encephalopathy; iv) neonates diagnosed with meconium aspiration syndrome, air leak syndrome, neonatal wet lung, or other respiratory conditions; v) cases where the ultrasonic image quality was significantly compromised due to various factors resulting in increased measurement errors.

Lung ultrasound and SNAPPE-II

The study employed the PHILIPS CX5010 portable ultrasound diagnostic instrument, which utilized a high-frequency linear array probe with a frequency range of 8-12 MHz10. The LUS examinations were performed by a radiologist in the NICU during routine clinical care with the assistance of nurses. For cases where unclear LUS images were obtained, a standard clinical protocol was followed, which included positioning the infant in a lateral recumbent position to improve image clarity. The radiologist recorded images of each lung region during the examination.

All infants included in the study underwent LUS as part of their clinical evaluation. Apgar scores were recorded at birth as part of standard care, and decisions regarding NICU admission and treatment were made based on the clinical judgment of the attending neonatologist and the clinical examination results. This study’s protocol did not influence these decisions, as all analyses were conducted retrospectively. Per LUS protocol, the neonate’s lungs were categorized into twelve regions: anterior upper, anterior lower, axillary, posterior upper, and posterior lower. Each lung region was assigned a score ranging from 0 to 3, resulting in a cumulative score ranging from 0 to 3611. Figure 1 is representative of the images we obtain in actual clinical practice. A score of zero indicates the presence of only A-lines or <3 B-lines (Figure 1A), a score of one indicates the presence of ≥3 well-spaced B-lines (Figure 1B), a score of two indicates dense and coalesced B-lines (Figure 1C), and a score of three indicates expansion and merger, exhibiting a beach-shaped pattern. Some lung consolidations presented bronchial air signs (Figure 1D).

Figure 1: Lung ultrasound images illustrating the four characteristic patterns of lung ultrasound scores: A) presence of only A-lines (arrows) or <3 B-lines (score: 0), B) presence of ≥3 well-spaced B-lines (arrows) (score: 1), C) dense and coalesced B-lines (arrows) (score: 2), D) lung consolidations presented bronchial air signs (circle).

The SNAPPE-II score for neonates was calculated based on clinical observations on admission; the PaO2 was evaluated under room air conditions, where the fraction of inspired oxygen (FiO2) is 0.21. The relevant parameters were minimum mean arterial pressure, minimum body temperature, birth weight, small for gestational age, PaO2/FiO2 ratio, minimum serum pH value, multiple seizures, urine output, and 5-minute Apgar score. Table 1 displays the SNAPPE-II scale, which covers a range of 0 to 162 points, where higher scores indicate more severe NRDS.

Diagnosis of NRDS and treatment

NRDS is a prevalent condition in neonatology, diagnosed through clinical presentation, arterial blood gas analysis, and chest X-ray. NRDS patients are divided into mild, moderate, and severe according to the international classification of diseases 9th and 10th Editions12,13. The main symptomatic manifestations are: i) symptoms of respiratory distress in neonates occurring within 0-12 hours of birth, including cyanosis, nasal flaring, grunting, inspiratory retractions, and tachypnea (respiratory rate >60 bpm), ii) arterial blood gas analysis often reveals hypoxemia (PaO2 <60 mmHg) or hypercapnia (PCO2 >50 mmHg), and iii) chest X-rays typically show reduced lung transparency, increased lung markings, and evenly distributed granular and reticular patterns, sometimes with blurred cardiac and diaphragmatic borders, and signs of bronchial inflation with severe inflammation in both lungs.

Statistical analysis

We analyzed and processed this study’s data using the IBM SPSS Statistics for Windows, Version 21.0 (IBM Inc, Armonk, NY, USA) software. The study’s numeric variables that conform to normal distribution are expressed as mean ± standard deviation, and comparisons between groups are made using the independent samples t-test. The non-normally distributed data are described by median (first quartile, third quartile), and comparisons between groups are made using non-parametric tests. We used regression analysis to evaluate the relationship between the LUS score or SNAPE-II score and mortality rate. We used non-parametric tests to compare the general data between the survival and non-survival groups and assessed the severity and prognosis of NRDS through Pearson correlation analysis. We evaluated the prognostic value of LUS, SNAPPE-II, and their combined use for NRDS using the receiver operating characteristic (ROC) curves. In this evaluation, the prognostic status of NRDS was considered the dependent variable, with one indicating a poor prognosis and zero indicating a favorable prognosis, while we tested LUS, SNAPPE-II, and their combination as independent variables. We defined statistical significance as a p-value less than 0.05. 

Results

Characteristics of the subjects

A total of 130 neonates who were delivered in the obstetrics department of Yantaishan Hospital between May 2020 and November 2022 were included in this retrospective study. All neonates who were diagnosed with NRDS were admitted to the NICU for treatment. The survival group consisted of 108 neonates with NRDS who demonstrated improved health following therapy, whereas the non-survival group comprised 22 neonates affected by NRDS who did not survive despite treatment. Figure 2 illustrates the comprehensive research flow diagram for NRDS patients. The overall mortality rate of NRDS neonates was 17 %. According to the analysis of our research data, it was found that the main causes of neonatal death were small gestational age, pulmonary infection, genetic primary diseases, and brain complications. Table 2 presented a comparison of the general data for the two groups of neonates. In the survival group, there were 71 male infants (65.7 %), whereas the non-survival group had 12 male infants (54.54 %). The birth weight (BW) in the survival group was 2,254 ± 726 g, compared to 1,865 ± 571 g in the non-survival group (p <0.001). The gestational age (GA) of the survival group was 34 ± 2.9 weeks, while that of the non-survival group was 31 ± 2.4 weeks (p =0.013). Cesarean section deliveries accounted for 56 cases (51.85 %) in the survival group and 23 cases (71.9 %) in the non-survival group (p =0.022). Among the infants in the survival group, 49 (45.3 %) had Apgar scores of ≤7 at one minute, and 21 (19.4 %) had similar scores at 5 minutes. In the non-survival group, there were 17 cases (77.3 %) with Apgar scores of ≤7 at one minute, and 13 cases (59.1 %) with similar scores at 5 minutes.

Figure 2: Flow diagram of the study design for this retrospective study that included 130 neonates diagnosed with neonatal respiratory distress syndrome out of the 207 newborns admitted with respiratory disorders between May 2020 and November 2022.

Lung ultrasound and SNAPPE-II score

Table 3 provides a comparative analysis of SNAPPE-II and LUS across the two groups under study. The non-survival group had patients with a LUS score of (27 ± 5.43), and the SNAPPE-II score was (33.46 ± 7.35), significantly elevated compared to those observed in the survival group (p <0.001). Furthermore, a correlation analysis between SNAPPE-II and LUS scores in NRDS patients was performed (Figure 3), revealing a significant positive correlation in neonates with NRDS (r =0.904, p <0.001).

Figure 3: Correlation analysis of lung ultrasound (LUS) and score for neonatal acute physiology perinatal extension II (SNAPPE-II) scores in neonatal respiratory distress syndrome.

Diagnostic value of SNAPPE-II and LUS scores in NRDS prognosis

The diagnostic and prognostic values of the three variable combinations are presented in Table 4. The values of sensitivity, specificity, cutoff value, 95 % CI, area under the curve (AUC), and Jorden index were determined for LUS in relation to NRDS, resulting in values of 0.926, 0.696, 22.5, 0.838-0.946, 0.892, and 0.703, respectively. In contrast, SNAPPE-II exhibited lower sensitivity (0.676), AUC (0.831), and Youden index (0.622). As shown in Figure 4, the combined diagnosis of the two is significantly better than the individual diagnosis, with the highest sensitivity (0.917), AUC (0.934), and Youden index (0.781).

Figure 4: The diagnostic value of lung ultrasound (LUS) and score for neonatal acute physiology perinatal extension II (SNAPPE-II) scores in the prognosis of neonatal respiratory distress syndrome.

Discussion

NRDS is a prevalent and severe condition in neonates, characterized by severe respiratory distress and hypoxemia14. Studies indicate that NRDS infants have elevated extravascular lung water and impaired gas exchange, contributing to lung consolidation15,16. Ferdian et al17 reported that RDS incidence inversely correlates with gestational age, with younger infants at higher risk due to low birth weight and underdeveloped lungs. This vulnerability results in higher infant mortality and increased risk of complications, including bronchopulmonary dysplasia. Recent studies highlight the critical role of early interventions, such as surfactant therapy, in enhancing survival and reducing NRDS-related morbidity18,19.

Recent technological advancements have facilitated the evaluation of neonatal lung diseases. LUS and SNAPPE-II scores are widely employed worldwide for quantitative diagnosis in infants. Previous studies have shown that both LUS and SNAPPE-II scores possess high sensitivity and specificity in diagnosing NRDS20. Jiang et al21 reported that using the LUS score to predict NRDS severity yielded good results. Insufficient alveolar surfactant in infants with NRDS leads to alveolar collapse and lung consolidation, manifesting as varying numbers of B-lines on LUS. As the disease progresses, the number and extent of B-lines increase, A-lines disappear, and pleural lines become blurred, resulting in the characteristic “white lung” appearance9,22. These findings underscore the diagnostic advantages of LUS in early detection of NRDS. In our study, the LUS scores of deceased infants were significantly higher than those of the surviving infants. Further analysis of these findings revealed higher levels of pulmonary fluid in the deceased group, absence of A-lines, presence of pronounced B-lines in some cases, and slight pleural effusion in certain infants. These observations are consistent with pulmonary ultrasound findings in severe NRDS patients reported in previous studies.

Currently, neonates with NRDS typically require respiratory support to maintain alveolar expansion. The severity of the disease and its unfavorable prognosis are the primary factors contributing to neonatal mortality. Both LUS and SNAPPE-II scores are valid assessment methods used to evaluate the severity of NRDS. However, there is a lack of studies that combine these two methods to investigate NRDS. This study can fully verify and improve the application of LUS in NRDS, providing more effective support for treating and caring for NRDS patients. Correlation analysis showed that the LUS score increased with increasing SNAPPE-II score. Diagnostic value analysis showed that LUS and SNAPPE-II scores had high sensitivity and specificity in predicting the severity and prognosis of NRDS. The combined diagnosis of LUS and SNAPPE-II scores was significantly better than the separate tool, which can provide better guidance for NRDS patients and improve their survival probability.

However, our study has some limitations. This study did not evaluate the consistency of LUS and SNAPPE-II analysis performed by different operators during disease progression. No data are available for other neonatal outcomes relevant to NRDS (such as bronchopulmonary dysplasia, duration of mechanical ventilation, etc). The number of infant samples analyzed is relatively small, and the proportion of infants with gestational age <27 weeks and weight <1000 g is small, which may cause some bias in the research results. 

Conclusions

The higher the LUS and SNAPPE-II scores, the more severe the condition of the NDRS neonatal patient. It can be used as a reference to provide appropriate prolonged ventilation to reduce the risk of patient death. The combination of LUS and SNAP-II score in the diagnosis of NRDS shows excellent performance, which can provide better guidance for patients and improve their survival probability. 

Conflicts of interest

The authors declare no conflict of interest.

References

  1. Liu J, Qiu RX. Lung Ultrasound Monitoring of Legionella Ventilator-Associated Pneumonia in an Extremely Low-Birth-Weight Infant. Diagnostics (Basel). 2022; 12: 2253.
  2. Chowdhury N, Giles BL, Dell SD. Full-Term Neonatal Respiratory Distress and Chronic Lung Disease. Pediatr Ann. 2019; 48: e175-e181.
  3. Gao YQ, Qiu RX, Liu J, Zhang L, Ren XL, Qin SJ. Lung ultrasound completely replaced chest X-ray for diagnosing neonatal lung diseases: a 3-year clinical practice report from a neonatal intensive care unit in China. J Matern Fetal Neonatal Med. 2022; 35: 3565-3572.
  4. Iovine E, Nenna R, Bloise S, La Regina DP, Pepino D, Petrarca L, et al. Lung Ultrasound: Its Findings and New Applications in Neonatology and Pediatric Diseases. Diagnostics (Basel). 2021; 11: 652.
  5. Liu X, Si S, Guo Y, Wu H. Limitations of Bedside Lung Ultrasound in Neonatal Lung Diseases. Front Pediatr. 2022; 10: 855958.
  6. Chiu LW, Desai J, Shanti C, Rane S, Agarwal P, Thomas RL, et al. SNAPPE II Score As a Predictor of Survival in Neonates with Congenital Diaphragmatic Hernia: A Single Center Experience. Eur J Pediatr Surg. 2016; 26: 316-321.
  7. Hirakawa E, Ibara S, Yoshihara H, Kamitomo M, Kodaira Y, Kibe M, et al. Safety, speed, and effectiveness of air transportation for neonates. Pediatr Int. 2021; 63: 415-422.
  8. Chatziioannidis I, Samaras T, Nikolaidis N. Electrical Impedance Tomography: a new study method for neonatal Respiratory Distress Syndrome? Hippokratia. 2011; 15: 211-215.
  9. Ma H, Yan W, Liu J. Diagnostic value of lung ultrasound for neonatal respiratory distress syndrome: a meta-analysis and systematic review. Med Ultrason. 2020; 22: 325-333.
  10. Huang C, Ha X, Cui Y, Zhang H. A study of machine learning to predict NRDS severity based on lung ultrasound score and clinical indicators. Front Med (Lausanne). 2024; 11: 1481830.
  11. Popa AE, Popescu SD, Tecuci A, Bot M, Vladareanu S. Current Trends in the Imaging Diagnosis of Neonatal Respiratory Distress Syndrome (NRDS): Chest X-ray Versus Lung Ultrasound. Cureus. 2024; 16: e69787.
  12. Hirsch JA, Leslie-Mazwi TM, Nicola GN, Oklu R, Schoppe KA, Silva E 3rd, et al. The ICD-10 system: a gift that keeps on taking. J Neurointerv Surg. 2015; 7: 619-622.
  13. De Coster C, Quan H, Finlayson A, Gao M, Halfon P, Humphries KH, et al. Identifying priorities in methodological research using ICD-9-CM and ICD-10 administrative data: report from an international consortium. BMC Health Serv Res. 2006; 6: 77.
  14. De Luca D, Bonadies L, Alonso-Ojembarrena A, Martino D, Gutierrez-Rosa I, Loi B, et al. Quantitative Lung Ultrasonography to Guide Surfactant Therapy in Neonates Born Late Preterm and Later. JAMA Netw Open. 2024; 7: e2413446.
  15. Yang Z, Zheng H, Lin L, Hou J, Wen C, Wang Y, et al. Alterations in Respiratory Mechanics and Neural Respiratory Drive After Restoration of Spontaneous Circulation in a Porcine Model Subjected to Different Downtimes of Cardiac Arrest. J Am Heart Assoc. 2019; 8: e012441.
  16. Alonso-Ojembarrena A, Gregorio-Hernández R, Raimondi F. Neonatal point-of-care lung ultrasound: what should be known and done out of the NICU? Eur J Pediatr. 2024; 183: 1555-1565.
  17. Ferdian H, Wahid DI, Samad S, Wardani AE, Alam GS, Moelyo AG. Lung ultrasound in diagnosing neonatal respiratory distress syndrome: a meta-analysis. Paediatr Indones. 2019; 59: 340-348.
  18. Kim HC, Won YY. Clinical, technological, and economic issues associated with developing new lung surfactant therapeutics. Biotechnol Adv. 2018; 36: 1185-1193.
  19. Veldhuizen RAW, Zuo YY, Petersen NO, Lewis JF, Possmayer F. The COVID-19 pandemic: a target for surfactant therapy? Expert Rev Respir Med. 2021; 15: 597-608.
  20. Tochie JN, Sibetcheu AT, Arrey-Ebot PE, Choukem SP. Global, Regional and National Trends in the Burden of Neonatal Respiratory Failure and essentials of its diagnosis and management from 1992 to 2022: a scoping review. Eur J Pediatr. 2024; 183: 9-50.
  21. Jiang QX, Shi LJ, Shen LY, Li XQ, Huang RS, Chen LJ, et al. Application Value of a New Lung Ultrasound Scoring Method in Neonatal Respiratory Distress Syndrome Treatment. Ultrasound Med Biol. 2022; 48: 275-282.
  22. Quarato CMI, Verrotti di Pianella V, Sperandeo M. Transthoracic ultrasound in neonatal respiratory distress syndrome (NRDS): Complementary diagnostic tool. Eur J Radiol. 2019; 120: 108664.