An Analysis of Predictive Factors for Severe Neonatal Infection and the Construction of a Prediction Model

Introduction

Neonatal infection is one of the most common causes of morbidity and mortality among newborns, who are especially susceptible to infection due to the immaturity of their immune systems. Neonates most likely acquire infections before, during or after delivery. The symptoms of neonatal infection are elusive and often difficult to detect. Many severely infected neonates present only non-specific manifestations, including fever, hypothermia, depression, decreased responsiveness, insufficient milk intake and lethargy. This makes the diagnosis of neonatal infection extremely difficult, and a missed or delayed diagnosis can lead to rapid progression of the illness and even death.1 At present, methods of diagnosing neonatal infection include growing pathogenic microorganism cultures, chest radiography, peripheral blood and umbilical cord blood tests and monitoring C-reactive protein (CRP) and procalcitonin (PCT) levels. Pathogenic microorganism culturing is the gold standard for diagnosing neonatal infection, but its low positivity rate and time-consuming process do not meet the current demand for clinical prediction and early diagnosis of infection.2 Thus, identifying clinical indicators that can predict severe neonatal infection early and effectively and establishing predictive models based on these indicators is of great clinical significance for the early diagnosis, control and improved prognosis of neonatal infection.3–5

White blood cell count (WBC), CRP and PCT levels are currently used to diagnose and monitor neonatal infection,6 white blood cells are essential components of the human immune system, responsible for identifying and eliminating pathogens such as bacteria, viruses, and parasites, when an infection occurs, white blood cells increase to counter the infection, platelets play a key role in coagulation and hemostasis while also participating in inflammatory responses. During an infection, the inflammatory response can lead to platelet activation and consumption, potentially resulting in a decreased platelet count, additionally, some infections (particularly bacterial infections) can cause a reduction in platelets as toxins or immune mediators produced by the pathogens may directly damage platelets or interfere with their production, but the sensitivity and specificity of any single peripheral blood index need to be studied further. Neonatal leukocyte levels fluctuate by the day, and although leukocyte detection is convenient, it is greatly affected by non-infectious factors, such as emotions. Studies have shown that the sensitivity and specificity of leukocyte and neutrophil counts are low. PCT is a calcitonin precursor and an important indicator of bacterial infection with high sensitivity and specificity. In addition, its concentration in adults is positively correlated with the degree of bacterial infection; however, PCT levels fluctuate greatly in the first 48 h after birth, so their value for the diagnosis of early neonatal infection needs to be investigated. A meta-analysis, performed by Vouloumanou et al,7 suggests that PCT has great diagnostic value after the first 48 h of life. Meanwhile, Zhao et al8 found that PCT was useful for the early diagnosis of severe infection but had limited value in the diagnosis of mild infection. CRP is an acute, time-phase protein that begins to rise 4–6 h after bacterial infection and usually reaches an effective threshold after 24–48 h. However, CRP fluctuates greatly in the first 3 days after birth, so errors arise in the diagnosis of neonatal infection based on CRP alone. Nevertheless, the combination of CRP and other peripheral blood indicators helps to improve diagnostic accuracy. Wu et al9 found that the detection of PCT and high-sensitivity CRP in combination could promote diagnostic specificity. Previous studies have focused on diagnosis in the early stage of infection, but few studies have reported on changes in infection indicators during severe infection or in the progression from mild to severe infection. Therefore, this study retrospectively analyses clinical data related to the diagnosis and treatment of neonatal infection to explore changes in infection-related indicators over the course of a severe infection, find primary independent predictors of neonatal infection and establish predictive models based on these predictors. It will serve as a reference for clinicians concerning the timely identification and intervention of severe neonatal infections.

Methods Study Subjects

Neonates hospitalised in the Department of Neonatology at Suixi County Hospital between January 2019 and June 2022 were retrospectively analysed and divided into two groups, based on the degree of infection as follows: a mild infection group and a severe infection group, using a 1:1 ratio. Samples were randomly selected using computer-generated random numbers. The minimum sample requirement was at least 40 neonates per group. The neonatologists in charge of the cases diagnosed cases of severe infections. The hospital has 1200 beds, 10 neonatologists on staff, and is a teaching and referral hospital. The inclusion criteria were as follows:

Diagnosis of neonatal infection according to Practical Neonatology,10 with two or more of the following clinical signs: Body temperature >37.5°C Increased heart rate Unstable or decreased percutaneous oxygen saturation (SPO2) Decreased milk intake Abnormalities in one or more of the following: blood pressure; respiration; partial pressure of oxygen in arterial blood; pH, serum sodium, serum potassium, creatinine, blood urea nitrogen, haematocrit levels (all neonates included in the study underwent pathogenic microorganism culture analysis and were diagnosed with a neonatal infection); Gestational age of <42 weeks and ≥24 weeks; Admission age of 1–28 days; Newborns who need to be treated with special grade antibiotics are considered severe infections, otherwise they are considered mild infections. Mild infections are treated with non-special grade antibiotics.

The purpose of this study was to explore the predictors of different severities of neonatal infection. Thus, study participants were divided into two groups: a severe infection group and a mild infection group. In the severe infection group, neonates were diagnosed with conditions such as sepsis, pneumonia, meningitis, and necrotising enterocolitis. In contrast, neonates in the mild infection group were diagnosed with less severe infections, including upper respiratory tract infections, urinary tract infections, and localised skin infections. Severe infections were defined as infections that are difficult to control with “restricted use”-grade antibiotics (ie, signs and symptoms of infection and/or abnormal indicators of infection were aggravated during the treatment) and require treatment with “special use” antibiotics. Restricted use antibiotics are defined as antibiotics that are reserved for specific situations or severe infections to prevent the development of antibiotic resistance. The World Health Organization’s AWARE (“access, watch, and reserve”) classification categorises antibiotics into three groups, with Reserve antibiotics being those that should be used as a last resort for treating multi-drug resistant infections. The remainder of the study population was treated as the mild-infection group. Non-infectious diseases, such as neonatal asphyxia, haematological diseases and immune abnormalities, were excluded. The exclusion criteria were (1) congenital abnormalities of platelets or coagulation factors; (2) the mother’s uterine or placental dysfunction (eg, severe malnutrition, gestational hypertension, preeclampsia, placenta previa or placental abruption); and (3) instances where the mother suffered from known thrombotic tendencies or autoimmune diseases and used anticoagulants, antiplatelet drugs, steroidal anti-inflammatory drugs and/or blood products in the third trimester of pregnancy or during delivery.

Relevant indicators and possible predictors of neonatal infection during analysis included fever, increased heart rate, unstable percutaneous SPO2 (often <90%), decreased milk intake, increased WBC, decreased WBC, increased neutrophils%, decreased neutrophils%, decreased platelet count (PLT), increased mean platelet volume (MPV), an increased MPV/PLT ratio and elevated CRP and PCT levels.11 The WBC, CRP and PCT levels often increase within the first 72 h after birth due to non-infectious factors, such as labour pressure, foetal distress, perinatal asphyxia and ischemic tissue injury, which affect the determination of infection. The upper bounds of the WBC, CRP and PCT levels should be specified; namely, the normal reference ranges within 72 h of birth are12,13 as follows: WBC: (5.0–25.0) × 109/L; PLT: (100–300) × 109/L; CRP: (0–20) mg/L; and PCT: (0–10) ng/mL. The study was approved by the medical ethics committee of the study hospital (ethics no.: KY-2019002), and informed consent was obtained from the families of all neonatal study participants. Although data collection was based on already completed medical records, we still made efforts to ensure informed consent was provided by the patients’ families. To do this, we contacted patient families either by phone or in writing, informing them of the purpose, methods, and potential risks of the study, and sought their consent to use patient data for research purposes. We respected the wishes of the families, and if they did not agree to participate, we refrained from using their related data. During data analysis, we ensured that all data were anonymised to protect patient privacy.

Data Collection

General data (eg, gender, age and body weight), clinical manifestations (eg, body temperature, heart rate, percutaneous SPO2 and milk intake volume), laboratory results (eg, WBC, N%, PLT, MPV, MPV/PLT, CRP and PCT levels) and clinical severity of infection were collected by two research assistants.

Specimen Collection and Detection

Two millilitres of peripheral venous blood were drawn from each neonatal study participant and placed in a heparin lithium anticoagulant tube and an EDTA2K anticoagulant tube. A Mindray BC-5390 Auto Haematology Analyzer and a Roche Cobas® 8000 biochemical immune-detection system were used to measure WBC, N%, PLT, MPV, CRP and PCT levels. All blood specimens were tested by the professionals of the laboratory department of the study hospital according to strict operating instructions. The data collection and laboratory results were obtained within 72 hours after birth. For serial data, the highest values within the 72-hour time frame were used in the analysis.

The normal reference ranges 72 h after birth10 were as follows: WBC (5.0–14.5) × 109/L; PLT (100–300) × 109/L; CRP (0–10) mg/L; and PCT (0–0. 5) ng/mL.

Statistical Methods

Statistical analysis of the data was performed using SPSS v. 21.0 statistical software and the R3.6.1 software and extension package. Measurement data that did not conform to a Gaussian distribution were expressed by M (P25, P75), and a Mann–Whitney U-test was used. Categorical variable data are expressed as n (%), and a chi-squared (χ²) test or Fisher exact probability method was adopted. Considering both the statistical significance and clinical relevance of the variables to identify the primary predictive factors for the occurrence of severe neonatal infection, multivariate analysis was performed using logistic regression (progressive forward, α in: 0.05, α out: 0.10). A receiver operating characteristic (ROC) curve was used to obtain the area under the curve (AUC) in the analysis of predictive efficacy. Logistic regression was used to predict the probability of the occurrence of an event by fitting data to a logit function.14 R3.6.1 software was used to draw a nomogram model, and the calibration curve was drawn with a bootstrap method (1000 sampling repeats) for the internal verification of the established model and to test the accuracy of the prediction model. The results were considered statistically significant at P < 0.05.

Results

From January 2019 to June 2022, consecutive neonates with sepsis hospitalised at Suixi County Hospital were enrolled in this retrospective study. According to the inclusion criteria, out of the initial 200 neonates, 40 were excluded for reasons such as a lack of complete clinical and laboratory data; 80 newborns diagnosed with severe infection and admitted to the neonatal intensive care unit (NICU) were placed in the severe infection group, and 80 neonates diagnosed with mild infection during the same period (none of whom were admitted to the NICU) were randomly selected at a ratio of 1:1 and placed in the mild infection group (Figure 1). Microbiological culture results in the severe infection group revealed that most infections were due to gram-negative rods. Positive and negative likelihood ratios were 5.67 and 0.42, respectively.

Figure 1 Flow chart of patient recruitment.

The Comparison of General Clinical Characteristics Between the Two Groups in the Early Stage of Infection

Eighty neonates with severe infection and 80 with mild infection were included in this study. There was no statistical difference in gender, age or weight distribution between the two groups. The male-to-female ratios of the severe infection group and the mild infection group were 54:26 and 51:29, respectively, and the average body weight was 2.3 kg and 3 kg, respectively. The average age of neonates in both groups was 1 day. Newborns with severe infections were more likely to have an increased heart rate and decreased milk intake. Routine analysis of blood data showed that, compared with the mild infection group, neonates in the severe infection group were more likely to have decreased WBC and N% levels and a significantly increased MPV/PLT ratio. Compared with the recovery stage, the CRP and PCT levels in both groups were increased in the early stage of infection and the difference was statistically significant (P < 0.05), as shown in Table 1 and Table 2. There was no significant change in MPV in the early stage of infection in either group.

Table 1 Comparison of Relevant Indicators Between the Early Stage and the Recovery Stage of Neonatal Infection in the Two Groups M(P25,P75)

Table 2 Comparison of Clinical Indicators in the Early Stage of Neonatal Infection Between the Two Groups

The Comparison of Infection-Related Indicators in the Severe Infection Group Between the Early and Recovery Stages of Infection

The mean routine blood data of the severe infection group in the early stage of infection showed a significant decrease in WBC and PLT levels, while the MPV/PLT ratio, as well as CRP and PCT levels, was elevated. The above indicators tended to be within the normal range during the recovery stage. The MPV level during the early stage of infection was normal, but it increased during the recovery period, as shown in Table 1.

The Comparison of Infection-Related Indicators in the Mild Infection Group Between the Early and Recovery Stages of Infection

Similar to severely infected patients, the WBC, CRP and PCT levels in the mild infection group increased in the early stage of infection and tended to be normal in the recovery stage. The MPV level was normal in the early stage of infection but increased significantly in the recovery stage. Moreover, the MPV/PLT ratio in the early stage of infection did not differ from that of the recovery stage (see Table 1).

The Univariate Analysis of Possible Predictors of Severe Neonatal Infection

As shown in Table 2, there was no difference in sex, weight or day age between the two groups (P > 0.05). Possible predictors of severe neonatal infection were determined, based on clinical symptoms, as well as a comparison of baseline levels of routine bloodwork in patients with severe and mild infections. The proportion of neonates with increased heart rate, decreased milk intake, decreased WBC, N% and PLT levels and an increased MPV/PLT ratio in the severe infection group was found to be higher than in the mild infection group. However, the proportion of neonates with increased WBC and N% levels (with correction) was lower in the severe infection group than in the mild infection group (P < 0.05) (see Table 2).

Independent Predictors of Severe Neonatal Infection and Associated Dichotomous Variable Model Equation

Crucial indicators from the univariate analysis were subjected to multivariate logistic regression analysis to identify independent predictors of severe infection. Decreased WBC, decreased PLT and increased CRP levels were found to be the three major independent predictors of severe neonatal infection, as shown in Table 3. Model equation (dichotomous variable): logitP = 7.402 × decreased WBC (no = 0, yes = 1) + 5.134 × decreased PLT (no = 0, yes = 1) + 3.148 × increased CRP (no = 0, yes = 1) − 3.787.

Table 3 Multivariate Logistic Regression Analysis of Predictors of Severe Neonatal Infection

As logitP = ln(P / [1 – P]), the predicted probability is P = 1/(1 + 2.718−logitP). The corresponding values of WBC, PLT and CRP were inserted into the equation to obtain a logitP score for each patient. Each score was then inserted into P = 1/[1 + 2.718−logitP] to calculate the predicted probability of severe infection for each neonate. An ROC analysis of logitP scores yielded an AUC of 0.958, a sensitivity of 85.0% and a specificity of 96.2%, which showed that the model equation has good predictive value for severe neonatal infection (see Figure 2).

Figure 2 ROC curves of the model equation for dichotomy variables (decreased WBC, decreased PLT, increased CRP) for risk of severe neonatal infection The AUC was 0.958.

An Analysis of the Predictive Value of Independent Indicators and Their Combined Prediction of Severe Neonatal Infection

ROC curves were drawn for the three independent predictors (WBC, PLT and CRP levels), as well as the three in combination, to analyse their predictive value for severe neonatal infection (see Figure 2). The AUCs of decreased WBC and PLT levels, increased CRP levels and the combination of these three indicators were 0.881, 0.798, 0.523 and 0.914, respectively, among of which that decreased WBC, decreased PLT and the three indicators together show the predictive value with statistical significance for severe neonatal infection (P < 0.05). The combination of the three indexes had the highest predictive value for severe neonatal infection, and its sensitivity and specificity were 91.3% and 77.5%, respectively (see Table 4 and Figure 3).

Table 4 ROC Analysis of Predictive Efficacy for Predictors of Severe Neonatal Infection

Figure 3 ROC curves of continuous numerical variables (decreased WBC, decreased PLT, increased CRP) and their combination for risk of severe neonatal infection. The AUC was 0.881, 0.798, 0.523, and 0.914, respectively.

The Establishment and Validation of a Nomogram Model for the Prediction of Severe Neonatal Infection

All independent risk predictors (continuous numerical variables) were used to establish a nomogram for the risk estimation of severe neonatal infection, as shown in Figure 4. The accuracy of the model was validated using the bootstrap technique (see Figure 5). After obtaining WBC, PLT and CRP values for each newborn, corresponding nomogram segments were checked. The corresponding scores were obtained, and a total score was calculated. After that, the total score and predicted probability segments of the nomogram were checked to obtain the predicted probability of severe infection for each newborn (Figure 4). The internal verification results of the nomogram model show the consistency index of the calibration curve to be 0.908 (95% confidence interval [CI] [0.862, 0.954]), which indicates that the predictions of the nomogram model (AUC = 0.914; 95% CI [0.872, 0.956]) were highly consistent with the results, thus demonstrating the model’s excellent predictive performance and high accuracy (see Figure 5).

Figure 4 Nomogram model for predicting severe neonatal infection Check the corresponding nomograms of WBC, PLT and CRP values respectively to get the corresponding scores, and add the scores to calculate the total score. Then check the total score and predicted probability segments of the nomogram to obtain the predicted probability of severe infection of each newborn.

Figure 5 Internal validation calibration curve of the Nomograph model The validation result showed that the consistency index (C index) of the calibration curve is 0.908 (95% CI: 0.862~0.954).

Discussion

By comparing the routine blood results and clinical symptoms of severe and mild infection in neonates, this study found that decreased WBC and PLT levels and elevated CRP levels were the main independent predictors of severe neonatal infection. Two prediction models (a dichotomous variable model equation and a nomogram model) were established to predict the incidence of severe neonatal infection. While both models were found to be highly accurate, the visualisation of the nomogram was more intuitive and convenient to use than the dichotomous variable model.

Newborns are particularly susceptible to pathogenic infections, and significant rates of morbidity have been observed worldwide. This may be due to an immature immune system. The pathogen spectrum varies with the healthcare setting and pathogen complexity, and the early symptoms of neonatal infection are insidious and atypical. Some newborns exhibit only non-specific manifestations, such as decreased milk intake. The presence of infection in newborns with low body weight is particularly difficult to detect and diagnose. Missed diagnoses and delays in treatment can result in serious infections and have a great impact on the health of the child. Therefore, early detection of severe neonatal infection is particularly important for early infection control and improved prognosis. WBC, CRP and PCT levels have some value in the diagnosis of neonatal infection,6 but there are few studies on the trends of infection indicators during severe infection or during the progression of infection from mild to severe.

As reported in the literature, the AUC of increased peripheral WBC during neonatal infection was 0.833, with a sensitivity of 64.0% and a specificity of 94.0%.6 This study found that the proportion of newborns with an increased WBC level in the severe neonatal infection group (8.8%) was lower than that in the mild neonatal infection group (23.8%); however, the proportion of newborns with a decreased WBC level in the severe neonatal infection group was higher (65.0%) than that in the mild neonatal infection group (1.3%). The AUC of decreased WBC levels predicting severe neonatal infection was 0.881 with a sensitivity of 98.8% and a specificity of 68.7%. This study achieved good diagnostic efficacy using decreased WBC levels as an early independent predictor of severe neonatal infection or progression from mild to severe infection. The results of this study show that WBC levels decreased in a majority of cases of severe neonatal infection, which differs from the trend of increasing WBC levels that were previously reported concerning neonatal infection. The reasons for this may be as follows: (1) the increase in neonatal WBC levels (up to 25,000–30,000/mm3) in the first 72 h after birth which was reported in the literature was due to perinatal risk factors (such as neonatal asphyxia), and they gradually recovered to the normal range after the 72-h period; the increase was not associated with an increased risk of infection; 13 (2) increased WBC levels 72 h after birth indicate mild infection, this finding is consistent with existing literature, which points out that an early increase in WBC levels may be related to infection; 15 and (3) in severe infection, a decrease in WBC levels is due to low immunity and bone marrow depletion. As reported in Austria, the low WBC levels associated with neonatal infection have high probability, high specificity and a negative predictive value (73.3–99.9% and >99.8%, respectively).16 A decreased WBC level is reported in approximately 35% of neonatal patients with sepsis. In contrast, the probability of an increased WBC level in neonatal patients with sepsis is only 17.7% or even as low as 4%.17 A decreased WBC level occurs in more than 40% of neonatal patients with sepsis caused by Escherichia coli, Klebsiella pneumoniae (K. pneumoniae) or Candida spp. and in approximately 30% of neonatal patients with septicaemia caused by Streptococcus agalactiae or Enterococcus spp. An increased WBC level was observed in only 22% of neonatal patients with coagulase-negative staphylococcal sepsis.18 The results of this study show that the probability of a decreased WBC level in severe neonatal infection is higher than the probability of an increased WBC level, which is consistent with the results of studies from other countries.19

As about 75% of neonatal sepsis is caused by a gram-negative bacterial infection, and reduced PLT levels are associated with gram-negative sepsis, thrombocytopenia is a common problem in neonatal sepsis.17 Approximately 75% of neonatal sepsis cases present with reduced PLT levels, which increase mortality almost fourfold.17,20,21 Moreover, the MPV/PLT ratio in adult patients with infection is higher than that in neonates, and it is a predictor of the clinical severity and mortality of sepsis;21 however, this is rarely mentioned in current domestic reports of neonatal infection. This study shows that the proportion of patients with a decreased PLT level in the severe neonatal infection group in the early stage of infection was higher (36.3%) than the proportion in the mild neonatal infection group (2.5%). After logistic multivariate regression and ROC analysis, a decreased PLT level was an early independent predictor of severe neonatal infection, with an AUC of 0.798, 83.8% sensitivity, 70.0% specificity and good diagnostic efficacy. The reasons for reduced PLT levels may be as follows: (1) the direct pathophysiological role of endotoxins produced by gram-negative bacteria is to decrease PLT levels;17 (2) inflammatory mediators in neonatal infection lead to platelet activation and aggregation and induce platelet adhesion to the endothelial cells of blood vessels. Some platelets can cross endothelial cells and fail to return to the blood, while others are destroyed by megakaryocytes. Moreover, some platelets bind to inflammatory mediators and lipopolysaccharides, which accelerates their destruction. Decreased PLT levels were reported in 75% of neonatal patients with sepsis, and 72.0% of neonatal patients with sepsis exhibiting severe thrombocytopenia.17 In cases of neonatal sepsis caused by Candida species or K. pneumoniae, PLT levels were decreased in 53.0% and 65.0% of cases, respectively. PLT levels decreased by 17.0% in neonates with gram-positive septicaemia.19 In addition, the results of this study show that there was no difference in MPV levels between the two groups in the early stage of infection; MPV levels increased in both groups during the recovery stage (compared with the early stage) and MPV/PLT ratios increased in the early stage. However, since MPV in the early stages of infection did not differ, multivariate logistic regression analysis showed that the MVP/PLT ratio is not an early predictor of severe neonatal infection; this result is inconsistent with the findings of related studies.22,23 This change may be due to the following reasons: (1) PLT levels increase in the early stage of infection, while the number of immature platelets in the blood increases in the recovery stage; an increase in MPV levels compensates for the adverse consequences of decreased PLT;24 (2) relevant studies do not indicate clearly whether data were drawn from the early or the recovery stage of infection. Therefore, the increase in MPV levels may derive from the recovery stage of infection rather than the early stage, which necessitates further research and discussion.

Studies have reported that CRP levels increased in 38.5% and PCT levels increased in 98.8% of neonatal sepsis patients. This study found that CRP and PCT levels rose in 25.0% and 8.8% of severe neonatal infection cases, respectively. Logistic multivariate regression and ROC analyses found decreased CRP to be an early independent predictor of severe neonatal infection with an AUC of 0.523, 33.8% sensitivity, 78.7% specificity and low diagnostic efficacy. CRP is an acute-phase protein, and its concentration rapidly increases in response to infection and inflammation. However, in certain cases, a decrease in CRP levels may indicate that the infection is under control or that the inflammatory response within the body has subsided. Therefore, based on the trend of CRP changes, it can serve as an early independent predictor of neonatal infection.25 However, this study found that a PCT level is not an early predictor of severe neonatal infection, which is inconsistent with the related studies;26–28 this may be related to the characteristics of neonatal physiology and immune response, as well as the sample size and types of infections included in the study.29 The low sensitivity of CRP and PCT in predicting severe neonatal infection may be due to their physiological increase in the first three days of life,12,30 which is a topic that requires further exploration.

As a single-centre, retrospective study with a limited sample size, this study has certain limitations. First, as this study was a retrospective investigation conducted at a single hospital, its results may be influenced by sample selection bias and data completeness. Second, since our research was conducted at a regional hospital, the study findings might not be directly generalisable to different regions and types of healthcare institutions. Furthermore, we did not perform separate analyses for each type of infection, which may limit our understanding of the impact of various infection types on WBC, PLT, CRP, and PCT. Lastly, although we utilised multivariable regression analysis to identify independent predictors of severe infection, there may still be other unconsidered confounding factors. Therefore, future studies should be conducted in larger patient cohorts and different types of healthcare institutions to further validate and expand our findings.

Conclusion

Decreased WBC and PLT levels and elevated CRP levels are the primary independent predictors of severe neonatal infection. The two models (the dichotomous variable model equation and the nomogram model) that were constructed, based on the major independent predictors, performed comparably in predicting severe neonatal infection with good differentiation and high accuracy. However, the nomogram visualisation is more intuitive and convenient to use than the dichotomous variable model. Both models can effectively predict the occurrence of severe neonatal infections and provide a scientific basis for clinicians to identify such infections early and formulate treatment strategies.

Data Sharing Statement

All data generated or analyzed during this study are included in this published article.

Ethics Approval and Consent to Participate

This study was conducted in accordance with the declaration of Helsinki. This study was conducted with approval from the Ethics Committee of The Hospital of Suixi County. Written informed consent was obtained from all parents/local guardians.

Consent for Publication

The manuscript is not submitted for publication or consideration elsewhere.

Consent Statement

Written informed consent was obtained from all parents/local guardians.

Acknowledgments

No funding or sponsorship was received for this study or publication of this article.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Funding

This research did not receive any funding support.

Disclosure

The authors declare that they have no competing interests.

References

1. Aixing H, Qingcao L. Application of SAA, CRP and PCT in diagnosis of neonatal infections. Shanghai J Prev Med. 2019;31(5):413–416. doi:10.19428/j.cnki.sjpm.2019.18428

2. Camargo JF, Caldas JPS, Marba STM. Early neonatal sepsis: prevalence, complications and outcomes in newborns with 35 weeks of gestational age or more. Rev Paul Pediatr. 2021;40:e2020388. doi:10.1590/1984-0462/2022/40/2020388

3. Shipe ME, Deppen SA, Farjah F, Grogan EL. Developing prediction models for clinical use using logistic regression: an overview. J Thorac Dis. 2019;11(Suppl 4):S574–S584. doi:10.21037/jtd.2019.01.25

4. Beneyto-Ripoll C, Palazón-Bru A, Llópez-Espinós P, et al. A critical appraisal of the prognostic predictive models for patients with sepsis: which model can be applied in clinical practice? Int J Clin Pract. 2021;75(8):e14044. doi:10.1111/ijcp.14044

5. Li X, Liu C, Wang X, Mao Z, Yi H, Zhou F. Comparison of two predictive models of sepsis in critically ill patients based on the combined use of inflammatory markers. Int J Gen Med. 2022;15:1013–1022. doi:10.2147/IJGM.S348797

6. Li R, Yang X, Li Y, Xie G, Liu X, Huang HC. The diagnostic value of HBP, PCT, CRP, WBC, and NLR in neonatal infections. Lab Med Clin. 2021;18(5):632–634. doi:10.3969/j.issn.1672-9455.2021.05.015

7. Vouloumanou EK, Plessa E, Karageorgopoulos DE, Mantadakis E, Falagas ME. Serum procalcitonin as a diagnostic marker for neonatal sepsis: a systematic review and meta-analysis. Intensive Care Med. 2011;37(5):747–762. doi:10.1007/s00134-011-2174-8

8. Zhao J, Jiang Y, Hou X. Diagnostic value of procalcitonin in early neonatal infection disease. Chin J Appl Clin Pediatr. 2012;27(2):122–124. doi:10.3969/j.issn.1003-515X.2012.02.019

9. Wu Y, Wang X, Fang Z, Chen H. Applications of PCT, CRP, WBC count in the early Diagnosis of Neonatal Infection. Chin J Nosocomiol. 2014;24:6063–6065. doi:10.11816/cn.ni.2014-143826

10. Shao X, Ye H, Qiu X. Practical Neonatology. Beijing: People’s Health Publishing House; 2019.

11. Sahu P, Raj Stanly EA, Simon Lewis LE, Prabhu K, Rao M, Kunhikatta V. Prediction modelling in the early detection of neonatal sepsis. World J Pediatr. 2022;18(3):160–175. doi:10.1007/s12519-021-00505-1

12. Zhu Y, Huang H. Progress in international shock research in 2018. J Pract Shock. 2019;3(01):45–51.

13. Stocker M, van Herk W, El Helou S, et al. C-reactive protein, procalcitonin, and white blood count to rule out neonatal early-onset sepsis within 36 hours: a secondary analysis of the neonatal procalcitonin intervention study. Clin Infect Dis. 2021;73(2):e383–e390. doi:10.1093/cid/ciaa876

14. Wu S. SPSS Actual Combat and Statistical Thinking. Beijing: Tsinghua University Press; 2019.

15. Rodwell RL, Leslie AL, Tudehope DI. Early diagnosis of neonatal sepsis using a hematologic scoring system. J Pediatr. 1988;122(5):761–767. doi:10.1016/S0022-3476(88)80699-1

16. Resch B. Comparison between pathogen associated laboratory and clinical parameters in early-onset sepsis of the newborn. Open Microbiol J. 2016;10:133–139. doi:10.2174/1874285801610010133

17. Utomo MT, Sumitro KR, Etika R, Widodo ADW. Current-proven neonatal sepsis in Indonesian tertiary neonatal intensive care unit: a hematological and microbiological profile. Iran J Microbiol. 2021;13(3):266–273. doi:10.18502/ijm.v13i3.6386

18. Guo J, Luo Y, Wu Y, Lai W, Mu X. Clinical characteristic and pathogen spectrum of neonatal sepsis in Guangzhou City from June 2011 to June 2017. Med Sci Monit. 2019;25:2296–2304. doi:10.12659/MSM.912375

19. Adane T, Worku M, Tigabu A, Aynalem M. Hematological abnormalities in culture positive neonatal sepsis. Pediatric Health Med Ther. 2022;13:217–225. doi:10.2147/PHMT.S361188

20. Ree IMC, Fustolo-Gunnink SF, Bekker V, Fijnvandraat KJ, Steggerda SJ, Lopriore E. Thrombocytopenia in neonatal sepsis: incidence, severity and risk factors. PLoS One. 2017;12(10):e0185581. doi:10.1371/journal.pone.0185581

21. Arabdin M, Khan A, Zia S, Khan S, Khan GS, Shahid M. Frequency and severity of thrombocytopenia in neonatal sepsis. Cureus. 2022;14(2):e22665. doi:10.7759/cureus.22665

22. Vélez-Páez JL, Legua P, Vélez-Páez P, et al. Mean platelet volume and mean platelet volume to platelet count ratio as predictors of severity and mortality in sepsis. PLoS One. 2022;17(1):e0262356. doi:10.1371/journal.pone.0262356

23. Panda SK, Nayak MK, Thangaraj J, Das P, Pugalia R. Platelet parameters as a diagnostic marker in early diagnosis of neonatal sepsis- Seeking newer answers for older problems. J Family Med Prim Care. 2022;11(5):1748–1754. doi:10.4103/jfmpc.jfmpc_1271_21

24. Nührenberg TG, Stöckle J, Marini F, et al. Impact of high platelet turnover on the platelet transcriptome: results from platelet RNA-sequencing in patients with sepsis. PLoS One. 2022;17(1):e0260222. doi:10.1371/journal.pone.0260222

25. Chiesa C, Panero A, Osborn JF, Simonetti AF, Pacifico L. Diagnosis of neonatal sepsis: a clinical and laboratory challenge. Clin Chem. 2004;50(2):279–287. doi:10.1373/clinchem.2003.025171

26. Li X, Ding X, Shi P, et al. Clinical features and antimicrobial susceptibility profiles of culture-proven neonatal sepsis in a tertiary children’s hospital, 2013 to 2017. Medicine. 2019;98(12):e14686. doi:10.1097/MD.0000000000014686

27. Habib A, Raza S, Ali U, Zubairi AM, Salim E. Diagnostic Accuracy of Serum Procalcitonin (PCT) as an Early biomarker of neonatal sepsis using blood culture as gold standard. J Coll Physicians Surg Pak. 2021;30(4):383–387. doi:10.29271/jcpsp.2021.04.383

28. Hu J, Qin X. Bacteria profiles and risk factors for proven early-onset sepsis in preterm neonates. Saudi Med J. 2021;42(12):1281–1288. doi:10.15537/smj.2021.42.12.20210430

29. Turner D, Hammerman C, Rudensky B, Schlesinger Y, Goia C. Procalcitonin in preterm infants during the first few days of life: introducing an age related nomogram. Arch Dis Child Fetal Neonatal Ed. 2006;91(4):F283–F286. doi:10.1136/adc.2005.085449

30. Delanghe JR, Speeckaert MM. Translational research and biomarkers in neonatal sepsis. Clin Chim Acta. 2015;451:46–64. doi:10.1016/j.cca.2015.01.031

留言 (0)

沒有登入
gif