Clinical utility of the Platelet-to-neutrophil ratio in differentiating sepsis from neonatal pneumonia: an observational study

Abstract

Background

In some cases, neonates with pneumonia can develop sepsis. The timely and accurate diagnosis of sepsis due to neonatal pneumonia poses a challenge for many clinicians. Therefore, this study aimed to evaluate the potential of the platelet-to-neutrophil ratio (PNR), as an early biomarker for clinically differentiating neonatal sepsis among pneumonia cases.

Methods

We performed a cross-sectional analysis of clinical and laboratory data collected from 1,103 patients with neonatal pneumonia [773 with pneumonia and 330 with pneumonia and sepsis (PS)].

Results

The PS group had a lower PNR than the pneumonia group. Further analysis revealed that the incidence of PS was significantly higher in the low PNR group than in the high PNR group (43.9% vs. 15.9%, p < 0.01). Multivariate logistic regression analysis revealed that the PNR was an independent risk factor for sepsis in neonatal pneumonia. Receiver operating characteristic curve analysis demonstrated that the PNR has good efficacy in diagnosing neonatal patients with PS (area under curve = 0.76, 95% confidence interval: 0.73–0.80, p < 0.001).

Conclusions

PNR can serve as an early biomarker for differentiating neonatal sepsis among pneumonia cases. Despite its potential, the PNR requires validation in multicenter studies.

Introduction

Neonatal pneumonia, resulting from the high vulnerability of neonatal lungs to bacterial and viral infections, is the leading cause of morbidity and mortality among newborns globally [1]. Each year, between 152,000 and 490,000 infants under the age of one year die from pneumonia [2]. The term neonatal sepsis is used to designate a systemic condition caused by bacteria, viruses, or fungi (yeast) that results in hemodynamic changes and other clinical symptoms, leading to substantial morbidity and mortality [3]. It is important to note that neonatal pneumonia and sepsis are major contributors to neonatal mortality [4], and that newborns with pneumonia may also have undiagnosed sepsis. Additionally, newborns with both pneumonia and sepsis may miss the opportunity for timely treatment if sepsis is not managed according to recommended guidelines [5]. Traditionally, laboratory-confirmed neonatal sepsis is diagnosed by isolating the causative agent from a normally sterile body site, such as blood, cerebrospinal fluid, urine, pleural fluid, joint fluid, or peritoneal fluid [3]. Although blood culture is considered the gold standard for diagnosing neonatal sepsis, it has notable limitations, including prolonged turnaround time and a low positive rate for pathogen detection [6]. Determining whether a neonate with pneumonia also has sepsis is a difficult task for clinicians, as it requires both speed and accuracy. In resource-limited settings, there is a growing need for biomarkers that can provide rapid and reliable results. Camacho-Gonzalez et al. [7] have reported advancements in biomarkers such as complete blood count, C-reactive protein, procalcitonin, mannose-binding lectin, and cytokine profiles for neonatal sepsis. We also focus our attention on other novel biomarkers.

Platelets play an important role in anti-infection immunity and coagulation in patients with sepsis, and are closely related to sepsis progression [8,9]. Neutrophils are the innate immune cells that drive the initial inflammatory response during sepsis [10]. The platelet-to-neutrophil ratio (PNR) is calculated as the ratio of the platelet count (×10⁹ cells/L) to neutrophil count (×10⁹ cells/L). The PNR is increasingly considered as a driver of inflammation and thrombosis [11,12]. Sreeramkumar et al. [13] reported that the PNR was reduced in patients with septic shock and activated platelets interacted with neutrophils, both of which resulted in vascular and vital organ damage that may trigger shock. However, the association between PNR and neonatal sepsis has received little attention in the current literature.

Therefore, this study aimed to evaluate the potential of the PNR in clinically differentiating sepsis from pneumonia in neonates.

Materials and methods

Study population

We retrospectively analyzed clinical and laboratory data from 1,103 neonates with pneumonia hospitalized at Children’s Hospital Affiliated to Zhengzhou University from January 2015 to March 2022, including 773 cases of pneumonia and 330 cases of PS. The first test results obtained after the hospitalization of these neonates were collected for analysis. The inclusion criteria were as follows: (1) 1–28 days of age, and (2) diagnosis of pneumonia. The exclusion criteria were: (1) incomplete laboratory data, and (2) other diseases such as cancer, hematological diseases, congenital lung anomalies, congenital heart disease, and major congenital abnormalities. The study protocol adhered to the principles outlined in the Declaration of Helsinki and was approved by the Clinical Ethics Committee of the Children’s Hospital Affiliated to Zhengzhou University (2024-K-054). We confirm that all data were confidential and anonymized. Informed consent requirement was waived by the Clinical Ethics Committee of the Children’s Hospital Affiliated to Zhengzhou University due to the retrospective nature of the current study. According to the requirements of the Clinical Ethics Committee, the information that could identify patients in this study was strictly confidential. The study did not have any adverse effects on the patients.

Clinical evaluation and definition

Clinical pneumonia was defined by the presence of characteristic respiratory symptoms-such as cough, wheezing, sputum production, tachypnea, cyanosis, or fever-in combination with abnormal lung sounds (rales or rhonchi) identified on chest auscultation [14,15]. The diagnosis of pneumonia relied primarily on clinical manifestations, supported by imaging evidence and laboratory findings [16–18]. The clinical manifestations included abnormal body temperature, respiratory distress, and cough. Imaging evidence consisted of chest radiographs indicative of pneumonia, typically demonstrating infiltrates or consolidation. Laboratory examinations included routine blood tests, and blood gas analysis to support the diagnosis of pulmonary infection. Neonatal sepsis was defined as a systemic inflammatory response syndrome in the presence or result of a suspected or proven infection, according to the published International Pediatric Sepsis Consensus [19]. Diagnosis of pneumonia and sepsis by two separate researchers.

Data collection

The electronic medical records of hospitalized patients included data on age, sex, weight, body temperature, respiratory rate, heart rate, systolic blood pressure (SBP), and diastolic blood pressure (DBP). Moreover, data on the laboratory examinations of hospitalized patients before clinical intervention, including procalcitonin, C-reactive protein (CRP), total bilirubin (TBIL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), total protein (TP), albumin (ALB), urea nitrogen (UREA), creatinine (CREA), and uric acid (UA) were collected. The methods used to detect laboratory indicators were described in our previously published article [20].

Statistical analysis

This study utilized SPSS 25.0 for statistical analyses, and GraphPad Prism 8 was employed to generate the figures. Neonates were classified into two groups according to their PNR levels. The median PNR was chosen as the cutoff point because it equalized the sample size and minimized bias. Continuous variables with a normal distribution were analyzed using independent t-tests and expressed as mean ± standard deviation (SD). Non-normally distributed continuous variables were analyzed using the Mann-Whitney U test and reported as median (interquartile range). Categorical variables were analyzed using chi-square test or Fisher’s exact test and presented as numbers and percentages (n, %). The correlation between two variables was assessed using the Spearman correlation test. Independent risk factors for pneumonia and sepsis in neonates were identified through multivariate logistic regression analysis. The predictive accuracy was assessed using the area under the ROC curve. Statistical significance was defined as p < 0.05.

Results

Baseline clinical characteristics of the study population

A total of 1,103 neonates with pneumonia were included in this study. Of these, 330 were diagnosed with sepsis and assigned to the PS group, whereas the remaining 773 were assigned to the pneumonia group (Figure S1). Compared with the pneumonia group, neonates in the PS group had a significantly lower age, weight, temperature, systolic blood pressure (SBP), and diastolic blood pressure (DBP) (p < 0.001; Table 1). However, the respiratory rate and levels of procalcitonin and C-reactive protein (CRP) were significantly higher in the PS group than in the pneumonia group (p < 0.05; Table 1). Additionally, the levels of urea nitrogen (UREA) and creatinine (CREA) were significantly higher in the PS group (p < 0.001), whereas the levels of alanine aminotransferase (ALT), total protein (TP), and albumin (ALB) were significantly lower (p < 0.05; Table 1). Regarding the hematological parameters, the white blood cell (WBC) and neutrophil counts were significantly higher in the PS group than in the pneumonia group (p < 0.05), whereas the platelet count and PNR were significantly lower (p < 0.001; Table 1).

Table 1.

Baseline characteristics of mild-to-moderate pneumonia and severe pneumonia.

Variables	Pneumonia (n = 773)	Pneumonia with Sepsis(n = 330)	P
Age (days)	18.0 (10.0,23.0)	11.0 (5.0,17.0)	<0.001
Male,n (%)	478 (61.8%)	200 (60.6%)	0.700
Weight (kg)	3.30 ± 0.84	2.58 ± 1.01	<0.001
Temperature (°C)	36.91 ± 0.49	36.65 ± 0.67	<0.001
Respiratory (rate/minute)	51.49 ± 10.39	53.25 ± 11.93	0.014
Heart rate (bpm)	148.3 ± 14.9	147.3 ± 20.5	0.390
SBP (mm Hg)	75.62 ± 8.57	71.33 ± 11.57	<0.001
DBP (mm Hg)	45.92 ± 7.98	42.38 ± 9.97	<0.001
procalcitonin (ng/mL)	0.11 (0.08,0.20)	0.40 (0.15,4.09)	<0.001
CRP (mg/L)	0.8 (0.5,0.8)	0.8 (0.8,18.6)	<0.001
TBIL (μmol/L)	87.3(42.9,139.5)	92.9(33.7,140.4)	0.504
AST (U/L)	36.1(28.1,47.6)	34.4(26.0,53.0)	0.638
ALT (U/L)	29.8(22.6,39.5)	26.6(20.5,38.0)	0.025
TP (g/L)	52.12 ± 5.81	49.61 ± 8.48	<0.001
ALB (g/L)	32.77 ± 4.95	28.37 ± 5.47	<0.001
UREA (mmol/L)	3.0(2.0,3.9)	4.0(2.6,6.3)	<0.001
CREA (mmol/L)	38.9(29.1,53.6)	52.2(36.4,76.6)	<0.001
UA (mmol/L)	146.1(113.7,185.4)	155.4(107.8,231.7)	0.106
WBC (109 cells/L)	9.64(7.87,11.99)	10.38(7.24,15.17)	0.018
Platelet (109cells/L)	324.0(327.5,403.5)	216.5(103.0,334.5)	<0.001
Neutrophils (109 cells/L)	3.57(2.39,5.85)	5.68(3.36,9.47)	<0.001
PNR	91.18(46.56,150.80)	36.69(14.05,78.36)	<0.001

All values are presented as the mean ± standard deviation (SD) (measurement data for normal distribution), n (%) (Counting data), or as the median (interquartile range) (Measurement data of non-normal distribution).

Association of the PNR with neonatal sepsis

To investigate the relationship between PNR and neonatal PS, we divided the newborns into a low PNR group (PNR ≤74.49) and a high PNR group (PNR >74.49), based on the median PNR (Table 2). We found that neonates in the low PNR group were younger and had higher levels of procalcitonin and CRP (p < 0.05) compared to those in the high PNR group. Additionally, the low PNR group had significantly higher levels of total bilirubin (TBIL) and CREA (p < 0.001), whereas ALT, TP, and ALB levels were significantly lower (p < 0.001; Table 2). Further analysis revealed that the proportion of patients with neonatal PS was significantly higher in the low PNR group compared to the high PNR group (p < 0.001), whereas the proportion of patients with neonatal pneumonia was significantly lower (p < 0.001).

Table 2.

Clinical and demographic characteristics according to the median platelet-to-neutrophil ratio (PNR).

Variables	Low PNR group (≤74.49) (n = 551)	High PNR group (>74.49) (n = 552)	P
procalcitonin (ng/mL)	0.17(0.09,0.55)	0.12(0.08,0.37)	<0.001
CRP (mg/L)	0.8 (0.5,2.0)	0.8 (0.5,1.1)	0.025
TBIL (μmol/L)	98.5(43.9,153.0)	80.8(37.7,128.3)	<0.001
AST (U/L)	34.2(26.4,48.1)	37.1(28.8,50.0)	0.021
ALT (U/L)	27.1(20.2,37.0)	30.7(23.2,40.3)	<0.001
TP (g/L)	51.03 ± 8.64	53.35 ± 7.17	<0.001
ALB (g/L)	30.88 ± 9.22	33.23 ± 10.66	<0.001
UREA (mmol/L)	3.3(2.1,4.8)	3.1(2.1,4.08)	0.010
CREA (mmol/L)	44.3(32.1,65.5)	38.9(28.7,56.0)	<0.001
UA (mmol/L)	148.7(106.8,196.3)	147.9(117.3,194.5)	0.217
Pneumonia, n (%)	309(56.1%)	464(84.1%)	<0.001
Pneumonia with sepsis, n (%)	242(43.9%)	88(15.9%)	<0.001

Correlation between the PNR and clinical parameters

Spearman correlation analysis was used to further investigate the association between clinical indicators and the PNR. The PNR was positively correlated with age (r = 0.226, p < 0.001), ALT (r = 0.137, p < 0.001), TP (r = 0.173, p < 0.001), and ALB (r = 0.207, p < 0.001), but negatively correlated with procalcitonin (r = −0.176, p < 0.001), CRP (r = −0.128, p < 0.001), TBIL (r = −0.136, p < 0.001), UREA (r = −0.102, p = 0.001), and CREA (r = −0.179, p < 0.001; Table 3). Moreover, the PNR was not significantly correlated with aspartate aminotransferase (AST) or uric acid (UA) levels (p > 0.05; Table 3).

Table 3.

Spearman correlation between the platelet-to-neutrophil ratio (PNR) and clinical parameters.

Variables	r	P
procalcitonin (ng/mL)	−0.176	<0.001
CRP (mg/L)	−0.128	<0.001
TBIL (μmol/L)	−0.136	<0.001
AST (U/L)	0.027	0.382
ALT (U/L)	0.137	<0.001
TP (g/L)	0.173	<0.001
ALB (g/L)	0.207	<0.001
UREA (mmol/L)	−0.102	0.001
CREA (mmol/L)	−0.179	<0.001
UA (mmol/L)	−0.004	0.905

The PNR has potential for differentiating sepsis from pneumonia in neonates

Univariate and multivariate binary logistic regression analyses were used to assess the predictive value of PNR in identifying sepsis in neonates with pneumonia. Age, weight, temperature, respiratory rate, CRP, TP, ALB, and UREA were significantly associated with the PNR in the univariate logistic analysis (p < 0.05). After adjusting for these factors, the multivariate analysis showed that PNR was still an independent predictor of sepsis in neonatal pneumonia patients (odds ratio (OR) = 0.995, 95% confidence interval (CI): 0.992–0.998, p = 0.001). Further analysis revealed an independent association between median PNR and increased prevalence of neonatal PS (Table 4). These results indicated that the PNR has the potential to differentiate sepsis from pneumonia in neonates, which is consistent with a previous report [21].

Table 4.

Predictive value of the platelet-to-neutrophil ratio (PNR) for sepsis from pneumonia in neonates.

Variables	Univariate OR (95% CI)	P	Multivariate∗ OR (95% CI)	P
Presence of sepsis
PNR	0.920 (0.877–0.965)	<0.001	0.995 (0.992–0.998)	0.001
PNR group
High PNR	1		1
Low PNR	4.266 (3.165–5.751)	<0.001	2.211 (1.504–3.251)	<0.001

^*Adjusted for age, weight, temperature, respiratory rate, C-reactive protein (CRP), total protein (TP), albumin (ALB), and urea nitrogen (UREA).

Potential of the PNR for diagnosing neonatal sepsis

The receiver operating characteristic (ROC) curve analysis showed that the PNR (area under the curve (AUC) = 0.76, 95% CI: 0.73–0.80, p < 0.001; Table 5) had better diagnostic ability for neonatal PS than the platelet count and neutrophil count (Figure 1). The sensitivity of the PNR in predicting neonatal PS was 68.18%, and the specificity was 74.40% when the cutoff value was 60.91 (Table 5).

Table 5.

The efficacy of the platelet count (PLT), neutrophil count, and the platelet-to-neutrophil ratio (PNR) in predicting sepsis from pneumonia in neonates.

Variables	AUC	95% CI	cut-0ff	Sensitivity (%)	Specificity (%)	P
PLT	0.68a	0.65–0.72	236.50 × 109/L	55.76	71.29	<0.001
Neutrophil	0.65b	0.61–0.69	6.16 × 109/L	46.67	78.14	<0.001
PNR	0.76	0.73–0.80	60.91	68.18	74.40	<0.001

^a P < 0.05 for PLT vs. PNR.

^b P < 0.05 for Neutrophil vs. PNR.

Figure 1.

Receiver operating characteristic (ROC) curve for the platelet count, neutrophil count, and the platelet-to-neutrophil ratio (PNR) in predicting sepsis from pneumonia in neonates.

Discussion

Neonates, owing to their immature immune systems, are particularly vulnerable to infections, which can easily progress to neonatal pneumonia or sepsis [7]. Neonatal sepsis poses a significant threat to newborn health, contributing to 15.2% of the 2.76 million global neonatal deaths annually [22]. At present, the diagnosis of neonatal sepsis primarily relies on nonspecific clinical signs [21]. Whereas blood culture is the gold standard, it takes up to 48 h to obtain results, and is not always sensitive due to effect by several factors, such as maternal antimicrobial drug therapy, insufficient blood volume, and contamination [6]. Therefore, there is an urgent need to establish multivariate predictive models capable of identifying novel biomarkers that offer rapid, sensitive, and specific detection of neonatal sepsis [21]. In our study, we found that the PNR had a better diagnostic ability for neonatal PS than the platelet and neutrophil counts (p < 0.05; Figure 1).

Sepsis is closely linked to immune and coagulation disorders, which greatly impact its progression and prognosis [23,24]. Particularly, the immune system plays a crucial role in both progression and regression of sepsis [25]. Emerging studies have defined platelets as a type of immune and inflammatory cell [26]. Platelets are not only related to coagulation but also participate in the immune response by forming immune thrombi [27,28]. Platelets play key roles in coagulation and anti-infection immunity. A decrease in the platelet count is considered a risk factor for mortality in sepsis, and disease severity is often evaluated based on the degree of platelet reduction [29]. Conversely, neutrophils are the main effectors of the immune response against invading pathogens and regulate adaptive immunity by regulating other immune cells [30,31]. Neutrophils play a vital role in the innate immune response during sepsis by releasing inflammatory cytokines, chemokines, and regulatory cytokines, as well as by engulfing and killing pathogens through the use of antimicrobial peptides, proteases, and oxidants [32]. Neutrophils have been identified as biomarkers of sepsis [33], and the absolute neutrophil count and immature/total neutrophil ratio are used to predict early onset sepsis [34]. It is noteworthy that the observed decrease in platelet count in the sepsis group may suggest bone marrow suppression, which could potentially affect the total leukocyte count (TLC). However, the inverse relationship between TLC and platelet count observed in our study may also be influenced by other factors. For instance, in sepsis, inflammatory responses can lead to leukocyte mobilization and proliferation despite concurrent bone marrow dysfunction [35,36]. Additionally, variations in the timing and severity of sepsis, as well as individual immune responses, may contribute to this inverse relationship. While our current data provide some insights, further investigation is necessary to fully elucidate the underlying mechanisms.

We introduced the PNR, which combines platelet and neutrophil counts to enhance the predictive value of the individual markers. In clinical practice, our findings indicate that a PNR value below 60.91 corresponds to a 76% probability of identifying neonatal sepsis among pneumonia cases. The PNR has garnered growing attention in recent years as a novel marker of inflammation. The PNR has also been demonstrated as a novel indicator of disease severity and prognosis in many diseases, such as ovarian cancer, acute ischemic stroke, rheumatic disease, and preeclampsia from gestational hypertension [37–39]. Furthermore, the PNR was positively associated with the long-term prognosis of ST-segment elevation myocardial infarction after successful primary percutaneous coronary intervention [40]. The PNR has a high diagnostic value for peripheral blood parameters in neonatal pneumonia [41]. A previous study evaluated the prognostic value of the PNR in predicting the outcome of severe sepsis and found that it was an independent predictor of mortality [42]. However, no study has reported a correlation between the PNR and neonatal sepsis. It has been suggested that the platelet–neutrophil complex interacts with bacterial sepsis, triggering the release of neutrophil extracellular traps and reactive oxygen species [43], which cause organ tissue damage while destroying pathogens. An extensive inflammatory response during the early stages of sepsis leads to rapid platelet depletion. Simultaneously, neutrophil counts are elevated in patients with sepsis as a result of pathogenic infections, suggesting that these patients may have lower levels of the PNR [33,44–47].

To evaluate the potential of the PNR to differentiate sepsis from pneumonia in neonates, we analyzed data from a substantial neonatal population comprising 1,103 cases. Our data showed that patients in the neonatal PS group had a significantly lower PNR than those in the neonatal pneumonia group. The low PNR group exhibited a significantly higher prevalence of neonatal PS than the high PNR group (p < 0.001). Further multifactorial analysis revealed PNR as an independent risk factor for neonatal PS. The ROC curve showed that the PNR was superior to platelet and neutrophil counts alone in differentiating sepsis from pneumonia in neonates. The AUC value of PNR was 0.76, indicating fair diagnostic accuracy as a standalone biomarker for neonatal sepsis, and notably superior to other biomarkers previously reported by our research team, such as the lymphocyte-to-C-reactive protein ratio (LCR, AUC = 0.72) [48] and the C-reactive protein-to-platelet ratio (CPR, AUC = 0.68) [49]. Although all three studies share a common clinical goal of improving early diagnosis of neonatal sepsis, they differ in biomarker selection, underlying immune mechanisms, and study designs. LCR reflects the balance between immune regulation and systemic inflammation, CPR indicates the interaction between inflammatory burden and thrombocytopenia, while PNR integrates aspects of both inflammation and coagulation. The higher AUC observed for PNR suggests that it may provide a more comprehensive indicator in this clinical context. On the other hand, a comparison was made between the specificity and sensitivity of PNR and neutrophil-to-lymphocyte ratio (NLR) as biomarkers for predicting neonatal sepsis. The sensitivity of PNR in predicting neonatal sepsis was 68.18%, with a specificity of 74.40% at a cutoff value of 60.91. In our team’s previous study by Li et al. [20], the sensitivity of NLR in predicting neonatal sepsis was 51%, with a specificity of 68% at a cutoff value of 1.62. This comparison shows that PNR has significantly higher sensitivity and specificity than NLR. In addition, our current study included a larger and more balanced sample of neonates with pneumonia and PS, which may have contributed to the improved diagnostic accuracy of PNR compared to the previous studies. Differences in study design, population characteristics, and sepsis severity may also account for variations in the predictive performance of each biomarker. These discrepancies highlight the complexity of neonatal immune responses and the need for multifaceted diagnostic tools.

From a clinical perspective, these findings suggest that PNR may offer superior utility in early sepsis detection among neonates with pneumonia. However, given that each biomarker reflects different aspects of the host response to infection, integrating them into a unified diagnostic model may yield greater accuracy. We recommend that future studies incorporate LCR, CPR, and NLR into the current PNR-based model to assess whether the combined use of these biomarkers can enhance the predictive power for neonatal sepsis.

Nonetheless, our study had several limitations that should be addressed in the future. Firstly, although we collected data from 1,103 cases, these data were obtained from a single hospital and lacked validation in a multicenter study. This may limit the generalizability of our findings. Secondly, there was a lack of follow-up on future clinical outcomes, which could provide valuable information on the long-term effects of the PNR. Additionally, the PNR was measured at a single time point, and it is possible that the sampling time may have influenced the results at different stages of sepsis. Serial measurements at multiple stages of disease progression would have allowed for a more comprehensive understanding of the PNR’s role in differentiating neonatal sepsis from pneumonia. Therefore, we suggest that future studies include continuous measurements at various stages of disease progression to better assess the PNR’s potential as an early biomarker for differentiating sepsis among pneumonia cases. Large-scale, multi-center studies would also be beneficial in determining the diagnostic value of the PNR when measured consecutively at multiple stages of sepsis progression.

Conclusions

PNR was an independent predictor of the presence of neonatal sepsis. The PNR can serve as an early biomarker for identifying neonatal sepsis among pneumonia cases. The incorporation of novel biomarkers into clinical protocols is imperative [50], and PNR, as a recently identified biomarker, has the potential to inform the development of therapeutic strategies for neonatal sepsis. However, longitudinal studies and multicenter validation are needed to evaluate changes in PNR over time.

Funding Statement

This work was supported by the Science and Technology Project of Henan Province [232102310234]; the Key Research, Development, and Promotion Projects of Henan Province (Scientific and Technological Tackling) [222102310328]; and the Medical Science and Technology (joint construction) Project of Henan Province [LHGJ20220750, LHGJ20230586].

Ethical approval

The study was approved by the Clinical Ethics Committee of the Children’s Hospital Affiliated with Zhengzhou University [2024-K-054]. In accordance with the Committee’s requirements, any information that could identify patients was kept strictly confidential. We confirm that all data were confidential and anonymized. The requirement for informed consent was waived by the Clinical Ethics Committee of the Children’s Hospital Affiliated with Zhengzhou University due to the retrospective nature of this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data supporting the findings of this study are available from the corresponding author (JY) upon request.