Predictive value of combined detection of blood Urea nitrogen and Neutrophil-to-lymphocyte ratio for identifying severe pneumonia complicated with sepsis in neonates

Abstract

Background

Newborns with severe pneumonia are at a high risk of developing sepsis, and early identification of this risk can improve the prognosis for the affected children. The purpose of this study was to evaluate the predictive value of combining blood urea nitrogen (BUN) and neutrophil-to-lymphocyte ratio (NLR) in diagnosing neonatal severe pneumonia complicated with sepsis (NSPCS).

Methods

We retrospectively included 194 newborns hospitalized from January 2018 to December 2021. Of these, 51 newborns with severe pneumonia developed sepsis. Clinical and laboratory data were collected from electronic medical records. The newborns were divided into severe pneumonia and sepsis groups. Multivariate logistic regression analysis was performed to determine whether BUN and NLR were independent predictors of NSPCS. The predictive value of combining BUN and NLR was assessed using receiver operating characteristic (ROC) curve analysis.

Results

Newborns with severe pneumonia complicated by sepsis had elevated levels of BUN (P < 0.001) and NLR (P = 0.003). Correlation analysis indicated a positive correlation between NSPCS and levels of BUN (r = 0.341, P < 0.001) and NLR (r = 0.213, p = 0.003). Multiple logistic regression analysis revealed that BUN and NLR were independent risk factors for NSPCS. ROC curve analysis revealed that combining BUN and NLR had better efficacy in identifying NSPCS (AUC = 0.757, 95% CI: 0.681–0.834, P < 0.001), with significantly better discriminatory ability than either BUN (AUC = 0.724, 95% CI: 0.643–0.804, P < 0.001) or NLR (AUC = 0.640, 95% CI: 0.545–0.735, P = 0.003) alone.

Conclusion

The combined detection of BUN and NLR was a valuable biomarker for identifying NSPCS.

What is Known:

Newborns with severe pneumonia are at a high risk of developing sepsis, and early identification of this risk can improve the prognosis for the affected children.

The clinical symptoms of neonatal sepsis are nonspecific, and the diagnostic criteria are unclear. Identifying effective biomarkers for early detection of neonatal severe pneumonia complicated with sepsis (NSPCS) could significantly improve clinical outcomes.

What is New:

This study determined the risk factors for NSPCS.

The combined measurement of BUN and NLR levels may be a valuable biomarker for identifying NSPCS, providing clinicians with a reference for accurate diagnosis and treatment.

Introduction

Severe pneumonia is a serious infection of the lung parenchyma with a high mortality rate. Newborns, with their underdeveloped immune systems, are highly susceptible to severe pneumonia. It is one of the most causes of sepsis and is frequently seen in clinical practice [1, 2]. Once sepsis develops in children, the disease progresses rapidly, affecting multiple organs and elevating the mortality rate [3]. According to statistics, approximately 3 million newborns are diagnosed with sepsis annually, with a mortality rate ranging from 11 to 19% [4]. In 2019, there were 375,000 deaths due to sepsis [5, 6]. The clinical symptoms of neonatal sepsis are nonspecific, and the diagnostic criteria are unclear, which is challenging for clinicians [7, 8]. This highlights the importance of early diagnosis and prompt treatment when severe pneumonia leads to sepsis. However, there is a lack of research on neonatal severe pneumonia complicated with sepsis (NSPCS), and there are currently no effective biomarkers to accurately predict patient outcomes, resulting in delays in treatment. Identifying effective biomarkers for early detection of NSPCS could significantly improve clinical outcomes.

Blood urea nitrogen (BUN) is a waste product of protein metabolism produced by the liver. It is transported to the kidneys through the bloodstream and ultimately eliminated from the body. When the body has a high protein intake or the glomerular filtration rate decreases, the BUN level increases. BUN is commonly used as an indicator to assess protein catabolism and kidney function. It can also be a useful tool in predicting the prognosis of critically ill patients, as high BUN levels have been linked to adverse outcomes. BUN is associated with mortality in various diseases, including heart failure, pneumonia, kidney injury, and sepsis [9–12]. A cohort study found that elevated BUN was associated with longer hospital stays in patients with chronic obstructive pulmonary disease (COPD) [13]. Previous studies have demonstrated that sepsis patients often experience increased protein metabolism, leading to elevated BUN levels [14–17]. Additionally, BUN exhibits significant predictive value in diagnosing sepsis. However, there is currently no known correlation between BUN and NSPCS. The neutrophil-to-lymphocyte ratio (NLR), calculated by dividing the neutrophil count by the lymphocyte count from a routine blood test, has been reported to be a useful indicator of inflammatory status in patients [18]. NLR has been shown to aid in the diagnosis of sepsis, coronary heart disease, and COPD [19]. Studies have also found that a high NLR is an independent risk factor for mortality in infectious diseases [20, 21]. In children, an increased NLR has been linked to a higher risk of mortality during hospitalization [22, 23]. The aim of this study was to investigate the predictive value of combining BUN and NLR in diagnosing NSPCS, providing clinicians with a reference for accurate diagnosis and treatment.

Materials and methods

Study population

This study retrospectively analyzed the clinical and laboratory data of 194 newborns admitted to the Children’s Hospital Affiliated to Zhengzhou University from January 2018 to December 2021, including 143 cases with severe pneumonia and 51 cases with severe pneumonia complicated with sepsis. The inclusion criteria were as follows: (1) Neonates diagnosed with severe pneumonia; (2) Newborns diagnosed with sepsis; (3) Age ≤ 28 days. The exclusion criteria were as follows: (1) Newborns with congenital diseases, hematological diseases, heart disease, tumors, liver/kidney related diseases, or other diseases; (2) Incomplete clinical and laboratory data. This research protocol followed the Declaration of Helsinki and was approved by the Ethics Review Board of the Children’s Hospital Affiliated to Zhengzhou University. All data in this study was anonymous. Given the retrospective nature of this study, informed consent was not required.

Clinical definition

According to the British Thoracic Society’s severity assessment for community-acquired pneumonia in children, the following manifestations indicate severe pneumonia: fever > 38.5 ℃, respiratory rate > 70 times/min, moderate to severe chest wall depression, nasal wing flapping, cyanosis, intermittent apnea, groans, refusal to eat, tachycardia, and capillary filling time ≥ 2 s [24]. The International Pediatric Sepsis Consensus defines neonatal sepsis as the presence of a suspected or confirmed infection with systemic inflammatory response syndrome [25]. The diagnosis of severe pneumonia and sepsis was made by two researchers.

Collection of clinical data

Demographic and laboratory data were collected from the patient’s electronic medical records during their initial admission. Demographic data included age, sex, weight, body temperature, respiratory rate, heart rate, systolic blood pressure, and diastolic blood pressure. Laboratory data consisted of procalcitonin (PCT), C-reactive protein (CRP), lymphocyte count, neutrophil count, platelet (PLT) count, white blood cell (WBC) count, creatinine (CREA), uric acid (UA), and BUN nitrogen (BUN). The serum PCT concentration was measured using the Cobas 8000 modular analyzer (Roche Diagnostic, Rotkreuz, Switzerland). CRP levels were quantified using the UPPER Analyzer (ultra-sensitive CRP Kit, Upper Bio-Tech, Shanghai, China). Lymphocyte, neutrophil, PLT, and WBC counts were measured using an automated hemocytometer (Sysmex Corporation, Kobe, Japan). Serum levels of CREA, UA, and BUN were measured using an automatic biochemical analyzer (AU5800 Clinical Chemistry Analyzer, Beckman Coulter, CA, USA).

Statistical analysis

All data processing and statistical analysis were conducted using IBM SPSS version 26.0 software (SPSS Inc., Chicago, Illinois, USA) and R 4.3.3. Continuous variables that followed a normal distribution were represented as mean ± standard deviation (SD) and analyzed using independent t-tests. Non-normally distributed continuous variables were presented as medians (interquartile range) and analyzed using the Mann-Whitney U test. Categorical variables were expressed as numbers and percentages (n, %), and analyzed using Chi-square tests or Fisher’s exact test. Correlations between laboratory indicators were assessed using Pearson or Spearman’s correlation analysis. Binary logistic regression analysis was used to identify independent risk factors for NSPCS. Variables with P-values < 0.05 in the univariate logistic analysis were included in the multivariate logistic regression analysis. The predictive value of BUN, NLR, and their combined test for NSPCS was evaluated using receiver operating characteristic (ROC) curve analysis. The optimal diagnostic cut-off point was determined using the Youden’s index (sensitivity + specificity − 1). Delong’s test was used to compare the area under the ROC curve (AUC) of the two variables. A P-value of less than 0.05 was considered statistically significant.

Results

Study Population Characteristics

This study examined 143 newborns diagnosed with severe pneumonia (severe pneumonia group) and 51 newborns diagnosed with severe pneumonia complicated with sepsis (sepsis group; Table 1). Except for age and SBP, there were no significant differences between the two groups in terms of gender, weight, temperature, respiratory rate, heart rate, and DBP (P > 0.05; Table 1). Biochemical analysis revealed that newborns in the sepsis group had higher levels of PCT, CRP, neutrophil count, PLT counts, CREA, UREA, as well as NLR compared to those in the severe pneumonia group (Fig. 1). There was no significant difference between the two groups in terms of lymphocyte count, WBC count, and UA.

Table 1.

Basic characteristics of study subjects by groups

Vanables	Severe pneumonia group (n = 143)	Sepsis group (n = 51)	P
Age (days)	10.0 (4.0, 26.0)	9.0 (3.0, 13.0)	0.022
Male, n (%)	86 (60.1%)	35 (68.6%)	0.283
Weight (kg)	3.130 ± 0.598	2.976 ± 1.067	0.333
Temperature (℃)	36.918 ± 0.568	36.853 ± 0.925	0.638
Respiratory (rate/minute)	58.76 ± 11.93	58.73 ± 11.99	0.988
Heart rate (bpm)	151.70 ± 16.08	152.18 ± 17.95	0.860
SBP (mm Hg)	75.27 ± 11.43	71.41 ± 11.43	0.040
DBP (mm Hg)	45.90 ± 9.08	43.14 ± 10.83	0.078
Biochemical parameters
PCT (ng/mL)	0.132 (0.086, 0.817)	0.636 (0.269, 2.850)	< 0.001
CRP (mg/L)	0.80 (0.50, 3.66)	2.51 (0.80, 28.60)	0.006
Lymphocyte (×109 cells/L)	4.07 (2.83, 5.64)	3.41 (1.67, 6.07)	0.099
Neutrophil (×109 cells/L)	4.92 (3.02, 7.30)	6.79 (4.74, 9.47)	0.003
PLT (×109 cells/L)	283.0 (202.0, 376.0)	195.0 (84.2, 291.0)	0.006
WBC (×109 cells/L)	9.81 (7.53, 12.91)	11.63 (7.60, 14.80)	0.139
CREA (mmol/L)	41.60 (32.90, 63.70)	55.80 (40.0, 82.2)	0.001
UA (mmol/L)	152.6 (115.1, 217.3)	185.9 (118.0, 234.7)	0.066
UREA (mmol/L)	2.80 (1.80, 4.30)	4.70 (3.20, 6.40)	< 0.001
NLR	1.102 (0.498, 2.059)	1.755 (0.838, 3.966)	0.003

Abbreviations: SBP Systolic blood pressure, DBP Diastolic blood pressure, PCT Procalcitonin, CRP C-reactive protein, PLT Platelet, WBC White blood cell, CREA Creatinine, UA Uric acid, UREA Urea nitrogen. NLR Neutrophil-to-lymphocyte ratio

Fig. 1.

The levels of PCT, CRP, neutrophil count, PLT counts, CREA, UREA, and NLR in two group of data. ***P < 0.001 vs. control group, **P < 0.01 vs. control group

Correlation between clinical parameters and the presence of NSPCS

The Spearman correlation analysis was utilized to investigate the relationship between clinical parameters and NSPCS. PCT (r = 0.307, P < 0.001), CRP (r = 0.200, P = 0.005), neutrophils (r = 0.214, P = 0.003), CREA (r = 0.230, P = 0.001), UREA (r = 0.341, P < 0.001), and NLR (r = 0.213, P = 0.003) exhibited a positive correlation with the presence of NSPCS (Fig. 2; Table 2). Additionally, age (r=−0.164, P = 0.022), SBP (r=−0.157, P = 0.029), and PLT (r=−0.198, P = 0.006) were found to have a negative association with the presence of NSPCS.

Fig. 2.

Heat map of the correlations between clinical parameters and the presence of NSPCS. The bar on the right side of the plot shows the value of the coefficient. The blue color indicates a positive correlation while the red color indicates a negative relationship. The deepness of the color indicates the strength of the correlation

Table 2.

Correlation between the neonatal severe pneumonia complicated with sepsis and clinical parameters

Variables	r	P
Age (days)	−0.164	0.022
SBP (mm Hg)	−0.157	0.029
PCT (ng/mL)	0.307	< 0.001
CRP (mg/L)	0.200	0.005
Neutrophil (×109 cells/L)	0.214	0.003
PLT (×109 cells/L)	−0.198	0.006
CREA (mmol/L)	0.230	0.001
UREA (mmol/L)	0.341	< 0.001
NLR	0.213	0.003

Abbreviations: SBP Systolic blood pressure, PCT Procalcitonin, CRP C-reactive protein, PLT Platelet, CREA Creatinine, UREA Urea nitrogen, NLR Neutrophil-to-lymphocyte ratio

Independent risk factors for NSPCS

In the univariate logistic analysis, variables with a P-value less than 0.05 were selected for inclusion in the multivariate logistic regression analysis. These variables included age, SBP, PCT, PLT, CREA, BUN, and NLR. After adjusting for these variables, BUN (OR = 1.281, 95% CI 1.083–1.516, P = 0.004) and NLR (OR = 1.239, 95% CI 1.063–1.443, P = 0.006) were identified as independent risk factors for NSPCS (Table 3). Comparative analysis showed neonates with NSPCS had significantly higher BUN and NLR than those with severe pneumonia.

Table 3.

Regression analyses to determine the independent predictor of neonatal severe pneumonia with sepsis

Variables	Univariate		Multivariate
	OR (95% CI)	P	OR (95% CI)	P
Age (days)	0.952 (0.917–0.988)	0.009	0.976 (0.932–1.021)	0.285
SBP (mm Hg)	0.971 (0.944–0.999)	0.042	0.997 (0.964–1.030)	0.854
PCT (ng/mL)	1.037 (1.003–1.071)	0.030	0.989 (0.950–1.030)	0.585
PLT (×109 cells/L)	0.997 (0.995–1.000)	0.026	0.999 (0.996–1.001)	0.376
CREA (mmol/L)	1.018 (1.007–1.029)	0.001	1.003 (0.989–1.018)	0.657
UREA (mmol/L)	1.373 (1.183–1.594)	< 0.001	1.281 (1.083–1.516)	0.004
NLR	1.284 (1.110–1.486)	0.001	1.239 (1.063–1.443)	0.006

Abbreviations: SBP Systolic blood pressure, PCT Procalcitonin, PLT Platelet, CREA Creatinine, UREA Urea nitrogen, NLR Neutrophil-to-lymphocyte ratio

The predictive value of BUN and NLR combined detection for NSPCS

A ROC curve analysis was used to evaluate the ability of BUN combined with NLR to predict the presence of NSPCS. The higher area under the curve (AUC) for the combination indicated superior discriminative ability. The AUC demonstrated that the combination of BUN and NLR had a more significant discriminative ability than either BUN or NLR alone (AUC = 0.757, 95% CI: 0.681–0.834, P < 0.001; Fig. 3). The ROC curve displayed that BUN had a sensitivity of 84% and specificity of 53% in predicting the presence of NSPCS, while NLR had a sensitivity of 41% and specificity of 85%. When the optimal cut-off value was set at 0.19, the combined detection of BUN and NLR had a sensitivity of 86% and specificity of 59% in predicting the presence of NSPCS (Table 4). The combined detection of BUN and NLR achieved higher sensitivity than NLR alone while maintaining better specificity than BUN alone, striking an optimal balance for clinical detection.

Fig. 3.

ROC curves of UREA and NLR separately and in combination to predict the presence of NSPCS

Table 4.

The efficacy of UREA, NLR alone and combined detection in predicting the neonatal severe pneumonia with sepsis

Variables	AUC	Cut-off	Sensitivity (%)	Sepicificit (%)	95% CI	P
UREA	0.724	2.95	0.84	0.53	0.643–0.804	< 0.001
NLR	0.64	2.66	0.41	0.85	0.545–0.735	0.003
Combined	0.757	0.19	0.86	0.59	0.681–0.834	< 0.001

Abbreviations: UREA Urea nitrogen, NLR Neutrophil-to-lymphocyte ratio

Discussion

Newborns, with underdeveloped immune systems, are prone to infections, particularly respiratory infections. The early identification of NSPCS holds significant clinical urgency. Firstly, severe pneumonia in newborns may quickly progress to sepsis, which, if not recognized and promptly treated, can be life-threatening [26, 27]. Newborns with severe pneumonia may have unidentified sepsis [28]. Early intervention may prevent this catastrophic cascade. Secondly, the 72-hour survival rate of sepsis decreases by approximately 7.7% per hour, emphasizing the exponential mortality increase with diagnostic delay necessitates time-sensitive interventions [29]. Thirdly, with the continuous advancement of medical and health technology, significant strides have been made in the diagnosis of neonatal sepsis, yet numerous challenges remain. The clinical manifestations of sepsis are nonspecific, making it difficult for clinicians to diagnose based on symptoms alone [6]. The blood culture detection is time-consuming, which delays diagnosis and treatment. Additionally, pre-hospital antibacterial treatment in affected children can reduce the positive rate of blood cultures [30, 31]. Accurate early identification helps avoid unnecessary broad-spectrum antibiotic use in culture-negative cases, reducing antimicrobial resistance risks. What’s more, timely diagnosis allows for appropriate ICU triage and resource allocation, particularly valuable in resource-limited settings. Therefore, there is a critical need for biomarkers with high sensitivity and specificity to identify NSPCS, enabling doctors to intervene effectively.

Biomarkers play an important role in the occurrence and progression of severe pneumonia and sepsis, and there are many studies related to them in clinical practice, including hematological indicators (such as WBC, neutrophil count, and lymphocyte count), inflammatory indicators (such as PCT and interleukin-6), and acute phase reactants (such as CRP and albumin) [32–34]. Among these, hematological and inflammatory indicators have been most extensively studied, but they exhibit low sensitivity and specificity, limiting their diagnostic utility. Furthermore, there are few indicators available for detecting NSPCS in clinical practice, highlighting the need to enhance their diagnostic value in the management of neonatal cases. In this study, higher BUN and NLR levels were associated with an increased likelihood of NSPCS. In the multivariate analysis (adjusted for age, SBP, PCT, PLT, CREA), both BUN and NLR remained independently associated with NSPCS. The combined detection of BUN and NLR effectively identified the risk of NSPCS. Compared with infection and inflammatory markers, BUN is a convenient and cost-effective indicator that has been demonstrated to be involved in the development of various diseases. BUN not only predicts renal function with high accuracy but has also been studied in evaluating heart failure and metabolic disorders [35, 36]. It has been found that BUN is an independent risk factor for predicting acute necrotizing pancreatitis [10]. Hong et al. [37] found that higher BUN levels are associated with higher cardiovascular disease and all-cause mortality. BUN is an indicator of severe renal insufficiency and is closely linked to sepsis mortality [38, 39]. Considering the relationship between BUN and early organ dysfunction in sepsis, it may offer superior diagnostic value in identifying NSPCS [40]. It is widely recognized that the inflammatory response stimulates the production of neutrophils and accelerates the apoptosis of lymphocytes. The NLR, which takes into account both neutrophil and lymphocyte parameters, represents a novel systemic inflammatory biomarker that is extensively employed in clinical practice and acknowledged by clinicians and researchers [41, 42]. De Jager et al. [43] noted that NLR more accurately predicts the severity and prognosis of sepsis compared to traditional biomarkers. An NLR range of 1.24 to 6.76 is diagnostic of neonatal sepsis [44]. High NLR levels are closely associated with sepsis shock [45]. In this study, the roles of BUN and NLR in NSPCS were consistent with those in previous studies. These findings indicate that BUN and NLR may serve as important indicators for independently identifying NSPCS.

This study has certain clinical significance and advantages. In this study, we initially assessed the clinical value of BUN and NLR in distinguishing NSPCS. Our results indicated that NSPCS had higher levels of BUN and NLR. The marked elevation of BUN in NSPCS likely reflects hemodynamic instability (e.g., reduced renal perfusion due to septic shock) and catabolic state induced by infection. Meanwhile, NLR’s increase mirrors the immune dysregulation in sepsis, where bacterial load triggers neutrophilia while suppressing lymphocyte counts. This divergence between newborns with severe pneumonia complicated by sepsis and severe pneumonia underscores their utility in early risk stratification. Correlation analysis revealed that PCT, CRP, neutrophils, CREA, BUN, and NLR were positively correlated with NSPCS. Multivariate logistic regression analysis demonstrated that BUN and NLR were independent risk factors for NSPCS. Based on our ROC curve analysis, the optimal cut-off values were determined by maximizing the Youden index (sensitivity + specificity − 1), balancing diagnostic accuracy and clinical utility. The ROC curve indicated that the combination of BUN and NLR achieved superior discriminative ability. The combined detection of BUN and NLR enhanced the predictive power for NSPCS, compared to detecting either biomarker alone. The high sensitivity (86%) of the combined detection meant that it can effectively rule out the NSPCS (with a low false negative rate). The moderate specificity (59%) of the combined test indicated that it may still yield some false positive results, but the AUC (0.757) supported its overall diagnostic value. This strategy could be valuable for early screening, where high sensitivity is prioritized to avoid missing severe cases. It’s worth noting that clinical interpretation should integrate other factors (e.g., comorbidities, dynamic trends) rather than relying solely on single measurements. Previous studies have focused on the use of various biomarkers to predict the occurrence of severe pneumonia or sepsis in newborns. In this study, we utilized a combination of serum biochemical and hematological indicators to more accurately predict the presence of NSPCS. Xiao Juan et al. reported that higher BUN levels were linked to the presence and severity of neonatal sepsis [17]. BUN was associated with poor prognosis in sepsis and represented an independent risk factor for sepsis-related organ damage [46, 47]. Researches has suggested that NLR may serve as an effective diagnostic and prognostic indicator for sepsis [48]. Scholars have found that NLR is used as an auxiliary diagnostic indicator for distinguishing the presence and severity of bacterial sepsis in infants [49]. These findings align with our conclusions. In summary, the combined detection of BUN and NLR improved the diagnostic sensitivity of NSPCS, aiding physicians in early identification of high-risk patients, prompt targeted treatment, and consequently reducing hospital mortality rates.

However, there are some limitations to our study. This is a retrospective single-center clinical study that does not track future clinical outcomes and requires prospective multicenter studies to validate the clinical finding. Moreover, the sample size included in this retrospective study was relatively small, requiring larger sample size data for validation. Additionally, BUN and NLR data were only collected at admission. Dynamic monitoring of these indicators across multiple clinical stages of NSPCS would further enhance their diagnostic value and treatment response. While our results hint at the potential of BUN and NLR as biomarkers, further neonatal-specific studies are needed to confirm their utility for risk stratification and prognosis in this population.

Conclusions

The presence of NSPCS was positively and independently correlated with elevated levels of BUN and NLR. The combined measurement of BUN and NLR levels may be a critical reference indicator for evaluating NSPCS. Our study indicated that the combined detection of BUN and NLR was a valuable biomarker for identifying NSPCS.

Abbreviations

BUN

Blood urea nitrogen

NLR

Neutrophil-to-lymphocyte ratio

NSPCS

Neonatal severe pneumonia complicated with sepsis

ROC

Receiver operating characteristic

AUC

Area under the ROC curve

COPD

Chronic obstructive pulmonary disease

PCT

Procalcitonin

CRP

C-reactive protein

PLT

Platelet

WBC

White blood cell

CREA

Creatinine

UA

Uric acid

SD

Standard deviation

SBP

Systolic blood pressure

DBP

Diastolic blood pressure

Authors’ contributions

WG wrote the main manuscript text. WG, KG, TL, LY, HS, ZS and CL downloaded and analyzed the data. JY and JN were responsible for project design, administration, statistical analysis, manuscript review and editing. All authors critically reviewed the manuscript. The fnal manuscript has been approved by all authors.

Funding

This work was supported by the Key Research, Development, and Promotion Projects of Henan Province [242102310047], and the Medical Science and Technology (Joint Construction) Project of Henan Province [LHGJ20220729 and LHGJ20230586].

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Ethics Review Board of the Children’s Hospital Affiliated to Zhengzhou University. We confirmed that all the data were anonymized and maintained with confidentiality. The requirement for informed consent has been waived because of the retrospective nature of the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.