Abstract
Background
Early onset sepsis (EOS) is a potentially fatal condition in neonates and its correct management is still a challenge for neonatologists. Early antibiotic administration in the neonatal period may carry risks and reducing unnecessary antibiotic therapy remains a key objective. It is well known that EOS risk calculator has reduced investigations and antibiotics for suspected EOS, but many limitations have been identified (e.g. strong value to respiratory symptoms, lack of standardized neonatal monitoring). To develop a new protocol, called Neosafe, for the management of newborn ≥ 35 weeks at risk of EOS based on the EOS risk calculator and serial physical examination with the aim of reducing blood investigations and unnecessary antibiotic therapy.
Methods
In this single-center observational retrospective study 2000 Neonates ≥ 35 weeks were enrolled. A new protocol called Neosafe has been developed, which deviates from the original for two major changes: (1) different serial physical examination (SPE) recommendation for each risk situations (2) clinical scenario revisited with multiple signs. Comparison between Neosafe protocol and EOS risk calculator simulated data were performed using McNemar’s test (statistical significance p < 0,05) and the Cohen’s kappa coefficient was used to assess the degree of agreement among them.
Results
48/2000 infants (2,4%) would have performed blood investigations versus 82/2000 (4,1%) whether Neosafe protocol would have been applied (p < 0,0001). 43/2000 infants (2,2%) with Neosafe protocol would have received antibiotics versus 80/2000 (4,0%) with EOS risk calculator (p < 0,0001). All infants who would not receive antibiotics for Neosafe protocol were infants with equivocal signs and isolated mild to moderate respiratory distress, but predisposed to advanced medical monitoring. Proven sepsis during the study period were 4, 2 of them were lost by both (no symptoms at birth, blood tests performed for non-infectious reasons). 1829 infants (91,5%) would receive routinal nursing SPE monitoring, 128 advanced medical SPE monitoring and 43 advanced medical SPE+vital for neonatal intensive care unit (NICU).
Conclusions
In our study Neosafe protocol could have reduced blood investigations and antibiotic therapy. Neosafe protocol could be a promising tool that integrates the use of digital technologies and direct attention to the patients. More trials are needed for further improvements.
Background
The management of neonates ≥ 35 weeks’ gestation at risk for early-onset sepsis (EOS) remains a crucial aspect of neonatal practice. While the incidence of EOS in term and late preterm infants is lower than in preterm infants, the potential for severe complications warrants a low threshold for evaluation and treatment. EOS is defined as the onset of symptoms within the first 72 h of life [1, 2], typically resulting from vertical transmission through contaminated amniotic fluid or exposure to bacteria in the maternal genital tract during delivery [3]. Key risk factors include maternal Group B Streptococcus (GBS) colonization, prolonged rupture of membranes (PROM), maternal fever and intra-amniotic infection [3, 4].
In high-income countries, the incidence of EOS has declined, largely due to intrapartum antibiotic prophylaxis (IAP) aimed at preventing GBS infection and now stands at approximately 0.5 per 1,000 live births in term infants [5]. However, rates can be significantly higher in late preterm infants, reaching up to 4 per 1,000 [6]. EOS often presents with subtle, non-specific signs that can progress rapidly and early recognition is vital. Symptoms may include temperature instability (hypothermia or fever), irritability, lethargy, respiratory distress (e.g., tachypnea, grunting, hypoxia), poor feeding, tachycardia, poor perfusion, and hypotension [7]. Due to the non-specific nature of these signs, clinicians must maintain a high index of suspicion when a newborn shows deviations from their typical behavior or condition.
Diagnosing neonatal sepsis remains difficult, as many of its signs overlap with non-infectious conditions [8]. For instance, respiratory distress is often the first symptom of sepsis, but it can also result from benign conditions such as transient tachypnea of the newborn (TTN), which is far more common and occurring in approximately 60 per 1,000 live births [9, 10]. This overlap can lead to overdiagnosis and unnecessary antibiotic use. Clinicians must carefully weigh the risks of starting empirical antibiotics against close monitoring, especially in infants without clear perinatal risk factors. Although no universal definition of neonatal sepsis exists, it is generally understood as infection-induced organ dysfunction resulting from a dysregulated immune response [11]. Therefore, a thorough clinical and laboratory evaluation is essential before confirming sepsis as the underlying cause. Currently, the gold standard remains a positive blood culture [12]. However, its reliability can be affected by factors such as low-level bacteremia (≤ 4 CFU/mL) and small sample volumes (< 1 mL) [8, 13]. Some studies suggest that blood culture sensitivity can approach 100% if at least 1 mL of blood is collected and processed appropriately [14–16].
In the work-up of neonatal sepsis, the most commonly used laboratory biomarkers include complete blood count (CBC) with differential, C-reactive protein (CRP), procalcitonin and presepsin. However, these markers generally have low specificity and positive predictive value (PPV), meaning that while normal results may offer some reassurance, abnormal values must be interpreted with caution. Various non-infectious conditions (such as maternal preeclampsia, chorioamnionitis, hypoxic-ischemic injury, and intrauterine growth restriction) can elevate these markers in the absence of true sepsis [17–22].
The lack of consensus in existing guidelines for management of suspected EOS and in general, a low threshold for neonatologist in evaluation and treating a neonate with suspected EOS, lead to unnecessary treatment and hospitalization resulting in an disproportionate overtreatment [23, 24].
The american academy of pediatrics (AAP) recently outlined three main strategies for preventing EOS in newborns ≥ 35 weeks’ gestation [3]:
The categorical risk-based approach relies on predefined maternal and perinatal risk factors [25, 26]. The multivariate risk-assessment model (EOS Calculator) estimates individualized EOS probability by integrating perinatal variables with the infant’s clinical status [27–30]. The clinical-condition–based strategy, centered on serial physical examinations (SPE), focuses primarily on the newborn’s evolving clinical presentation over the first 48 h [31, 32].
Each approach has limitations. Categorical risk assessment tends to over-treat uninfected infants, while the EOS calculator promotes selective blood testing and treatment but it may still lead to overtreatment when mild, non-specific symptoms (mainly respiratory) are present [33]. Instead, SPE strategies, while promising, depends heavily on consistent and standardized clinical monitoring.
In a previous study, our team retrospectively applied with simulation the EOS calculator to newborn managed per CDC-based guidelines. We observed a significant reduction in blood testing, but not in antibiotic use [34]. Two major issues emerged: (1) respiratory signs often triggered antibiotic treatment in uninfected infants and (2) the absence of a structured protocol for serial clinical assessment. Another concern was the lack of attention to suspected chorioamnionitis, a known major risk factor for EOS. Therefore, our aim was to develop a tool that optimally integrates risk stratification at birth based on maternal-neonatal risk factors with careful clinical observation of all patients, under the assumption that, in our setting, this approach may be more accurate than the EOS calculator and currently more applicable than routine clinical observation alone.
Based on these findings, a new protocol called NeoSafe was developed to address these gaps and enhance both safety and antibiotic stewardship. The aim of the study is to assess whether Neosafe can reduce blood investigations and unnecessary antibiotic therapy.
Methods
This was a single-center observational retrospective study conducted at the Division of Neonatology at Filippo del Ponte Hospital in Varese, Italy. The study population consists of 2000 newborns with GA ≥ 35 weeks (the first 1,000 infants born in 2021 and the first 1,000 born in 2023) admitted to our department and managed according to the previous EOS prevention local hospital guideline (CDC algorithm [26]). Patients with major congenital malformations, major neurological conditions and major congenital heart diseases were excluded from the study. A comparison was made for the assessment and management of newborns at risk for early-onset sepsis through a simulation between the hypothetical application of the EOS risk calculator and the hypothetical application of the new protocol, which integrates the EOS risk calculator with serial physical examination (Neosafe protocol). Data collected included the following: neonatal data, mode of delivery, maternal and neonatal risk factors (GBS maternal status, amniotic membrane rupture ≥ 18 h, maternal fever ≥ 38 °C, intrapartum or pre-partum antibiotic therapy, suspected or confirmed intrauterine inflammation, infection, or both (Triple I) [35]. We also described the presence and type of clinical signs of EOS, laboratory tests and confirmed proven sepsis discovered in the study population. For the application of both protocol risk score and subsequent management recommendation were determined using an EOS incidence rate of 0.5/1000 live births. For EOS risk calculator each newborn has been recorded in a multivariate risk prediction model that combined neonatal/maternal data and neonatal clinical condition during the first 48 h after birth. Each newborn was classified as “well appearing”, “equivocal” or “clinical illness” based on the characteristics reported on the EOS calculator website. (https://neonatalsepsiscalculator.kaiserpermanente.org).
Possible management recommendations were as follows: (1) no culture, no antibiotics, routine vitals; (2) no culture, no antibiotics, vitals every 4 h for 24 h; (3) blood culture, vitals every 4 h for 24 h; (4) strongly consider starting empiric antibiotics, vitals for neonatal intensive care unit (NICU); (5) empiric antibiotics, vitals for NICU.
Otherwise, Neosafe protocol represents a modified version of the original EOS risk calculator from the Kaiser Permanente Division of Research. Neosafe stands for newborn and safe, highlighting the importance of safety in this population. The predictors were retained as in the original calculator: gestational age, the highest maternal antepartum temperature (measured only in Celsius), duration of membrane rupture, maternal GBS status and type and duration of antepartum antibiotics. Given its importance, a new predictor was added, considered an independent risk factor: suspected chorioamionitis using Triple I definition [35]. This condition is diagnosed if the maternal intrapartum temperature is > 39°C or between 38–39°C and persists for more than 30 minutes with one or more of the following clinical signs: maternal leukocytosis > 15,000/mm³, purulent cervical drainage, fetal tachycardia > 160 bpm for at least 10 minutes or more [35]. The answer options for Triple I is dichotomous “yes”, if present, or “no”, if absent. When present, this nullifies the numerical EOS risk composite value and different recommendations are indicated based on clinical evaluation.
Each newborn must also undergo a careful clinical examination at birth and during the scheduled monitoring evaluations, which further classifies the newborn into three clinical presentations: “well appearing,” “equivocal,” or “clinical illness.” The clinical patterns distinguishing the “equivocal” or “clinical illness” scenarios were modified from the original EOS risk calculator to better reflect multiple signs and symptoms of suspected sepsis and limit the weight of a single isolated respiratory symptoms. Particular attention was given to the type of severity and evolution of the respiratory symptoms: mild respiratory distress is defined by tachypnea and/or signs such as nasal flaring, grunting, and chest retractions, with or without low-flow supplemental oxygen. Moderate distress is characterized by the need for CPAP and/or HFNC with FiO₂ < 30%. Severe distress involves non-invasive respiratory support with CPAP/NIV and/or HFNC with FiO₂ > 30% and/or invasive support requiring endotracheal intubation.
Definitions of clinical presentations are described below in Table 1.
Table 1.
Clinical presentations: Neosafe protocol
| Clinical presentation | |
|---|---|
| Well appearing | In good condition, no persistent physiological abnormalities |
| Equivocal |
Single mild-to-moderate persistent physiological abnormality lasting > 2 h, including: - Mild respiratory distress (tachypnea > 60 breaths/min and/or signs of respiratory distress such as nasal flaring, grunting, and chest retractions) with or without low-flow oxygen supplementation; - Isolated moderate respiratory distress (need for CPAP and/or HFNC with FiO₂ < 30%); - Tachycardia (heart rate > 160 bpm); - Gastrointestinal signs (abdominal distention, poor feeding, vomiting); - Color skin changes (capillary refill > 2 s and/or pallor); - Tone changes (decreased tone) and/or irritability. |
| Clinical illness |
Two or more mild-to-moderate persistent physiological abnormalities lasting > 2 h, including: - Mild respiratory distress (tachypnea > 60 breaths/min and/or signs of respiratory distress such as nasal flaring, grunting, and chest retractions) with or without low-flow oxygen supplementation; - Isolated moderate respiratory distress (need for CPAP and/or HFNC with FiO₂ < 30%); - Tachycardia (heart rate > 160 bpm); - Gastrointestinal disturbances (abdominal distention, poor feeding, vomiting); - Color skin changes (capillary refill > 2 s and/or pallor); - Tone changes (decreased tone) and/or irritability; - Altered biomarkers (PCR > 15 mg/L) and/or white blood cell changes (leukocytes < 5000/µL, > 25,000/µL at birth, > 30,000/µL between 12–24 h, > 21,000/µL after 24 h, and neutrophils < 1800/µL or > 5400/µL) and platelets (< 150,000/µL) (performed only when indicated by the clinical recommendations) |
|
One or more severe persistent physiological abnormalities, including: - Severe respiratory distress (need for CPAP and/or HFNC with FiO₂ > 30%, need for endotracheal intubation); - Hemodynamic instability requiring vasoactive medications; - Central nervous system (CNS) alterations (changes in reactivity, neonatal seizures without another cause). |
A further change is the introduction of standardized SPE through two different monitoring sheet to be completed in the first 48 h of life. First, a standard monitoring sheet, that has been adapted from the postnatal surveillance sheet al.ready used in the Italian Society of Neonatology’s recommendations for the prevention and management of sudden unexpected postnatal collapse (SUPC) for the newborn during rooming-in practice [36]. It is filled out by obstetric and nursing staff with two initial assessments between the 3rd and 12th hours after birth, followed by further assessments approximately every 6–8 h for the first 48 h of life. If altered parameters are detected in the standard monitoring sheet, the attending physician is called to re-assess the newborn. Second, an advanced monitoring sheet contains more detailed clinical information that has been created by us. It is filled out by medical staff with two assessments between the 3rd and 12th hours of life, followed by approximately every 6–8 h for the first 48 h of life. Any change in scenario allows for the reassignment of a new score in a case of neonatal clinical deterioration (Figs. 1 and 2).
Fig. 1.
Standard monitoring sheet-derived from SUPC monitoring sheet during rooming-in
Fig. 2.
Advanced monitoring sheet
With the score obtained, different scenarios can be identified, seven of which have been defined, five of which were already present in the original EOS risk calculator. For each of these scenarios, clinical recommendations and an appropriate monitoring approach have been provided to emphasize the importance of repeated clinical observation of the newborn. When suggested by the clinical recommendations, the exams include blood tests (blood culture, complete blood count with differential and CRP) (Fig. 3).
Fig. 3.
Scenarios, clinical recommendations and an appropriate monitoring approach adopted in Neosafe protocol
If the EOS risk after clinical examination is < 1 and the EOS risk at birth is < 1: clinical recommendations indicate no blood tests or antibiotics, and monitoring follows the standard monitoring sheet;
If the EOS risk after clinical examination is < 1 and the EOS risk at birth is > 1: clinical recommendations indicate no blood tests or antibiotics but, given the importance of risk factors for EOS at birth, the newborn undergoes advanced monitoring sheet;
If the EOS risk after clinical examination is between 1 and 3: clinical recommendations lead to performing blood tests and monitoring is done with the advanced monitoring sheet;
If the EOS risk after clinical examination is > 3: the patient undergoes blood tests and receives antibiotics based on our current recommendations. The advanced monitoring sheet is used with continuous vital sign monitoring;
If the EOS risk after clinical examination is < 3, but the newborn has a “clinical illness” presentation: clinical recommendations involve performing blood tests and considering the possibility of administering antibiotics. Monitoring is done with the advanced monitoring sheet with continuous vital sign monitoring.
A sixth scenario was added to emphasize the importance of the “equivocal” clinical presentation in low-risk contexts:
If the EOS risk after clinical examination is < 1, but the newborn has an “equivocal” presentation: Clinical recommendations are similar to the first two scenarios, with no blood tests or antibiotics, but we increase the alert level, similarly to the fifth scenario, and place the newborn under advanced monitoring sheet to avoid missing borderline cases.
The seventh and final scenario added refers to clinical recommendations and monitoring in the case of suspected Triple I in each of the three clinical presentations:
Suspected Triple I and “well-appearing” newborn: the patient undergoes neither blood tests nor antibiotics. Monitoring follows the advanced monitoring sheet;
Suspected Triple I and “equivocal” newborn: the patient undergoes blood tests and the possibility of antibiotics is considered. Monitoring follows the advanced monitoring sheet with continuous vital sign monitoring;
Suspected Triple I and “clinical illness” newborn: the patient undergoes blood tests and receives antibiotics based on our current recommendations, with the advanced monitoring sheet used and continuous vital sign monitoring added.
Data were collected into a designated database and a comparison between the different protocols were performed using McNemar’s test (statistical significance p < 0.05) and the Cohen’s kappa coefficient was used to assess the degree of agreement among them.
Results
Data population
This study enrolled 2,000 newborns ≥ 35 weeks’ gestation, all delivered at “F. del Ponte” Hospital in Varese, Northern Italy, and managed according to our department’s current EOS protocol (based on the CDC algorithm [26]).
The cohort consisted of 977 females (48.9%) and 1,023 males (51.1%), with a mean gestational age of 39.4 ± 1.4 weeks. Of these, 4,3% were born > 35 weeks and < 37 weeks and 12.6% ≥ 41 weeks. Mode of delivery included spontaneous vaginal birth (71.0%), operative vaginal delivery with vacuum extraction (5.4%), and cesarean Sect. (23.6%). Meconium-stained or foul-smelling amniotic fluid was present in 8.3% of births.
Average birth weight was 3,230 ± 465 g. Most newborns (83.3%) were appropriate for gestational age (AGA), while 11.3% were small for gestational age (SGA 2,548 ± 325 g) and 5.4% were large for gestational age (LGA 4,061 ± 265 g).
Regarding infection risk, 74.6% of mothers tested negative for GBS, 9.8% had unknown status, and 15.6% were GBS-positive. Among the 312 GBS-positive mothers, 49.7% received adequate intrapartum antibiotic prophylaxis (IAP), while 50.3% received incomplete or no IAP, according to the CDC 2010 revised guidelines [25].
Premature rupture of membranes (PROM > 18 h) occurred in 10% of cases; among these, 55.5% received antibiotic therapy, while 44.5% did not.
Simulated protocol application: EOS calculator vs. Neosafe
If the EOS risk calculator had been applied:
82 newborns (4.1%) would have undergone blood investigations (including blood cultures)
80 newborns eligible for antibiotics: 4.0% (“Strongly consider antibiotics” 69 cases and “Empirical antibiotics” 11 cases)
In contrast, under the Neosafe protocol:
48 newborns (2.4%) would have undergone a full sepsis evaluation (CBC, CRP, and blood culture).
43 newborns eligible for antibiotics: 2.2% (“Strongly consider antibiotics” 36 cases and “Empirical antibiotics” 7 cases)
The data are shown in Table 2.
Table 2.
Comparison of exams and antibiotic therapy between the EOS risk calculator and the Neosafe protocol
| EOS risk calculator | Neosafe protocol | ||
|---|---|---|---|
| Exams | 82 (4.1%) | 48 (2.4%) | |
| Antibiotic therapy | 80 (4.0%) | 43 (2.2%) |
The comparison between EOS risk calculator and Neosafe protocol regarding the performance of exams on newborns at risk for EOS was assessed using McNemar’s test (Table 3). The results show a significant difference between the two protocols (χ² = 24.08, p < 0.00001). Cohen’s kappa indicates a moderate-to-good agreement between the two protocols, with a value of 0.62 (95% CI: 0.52–0.72).
Table 3.
Comparison of exams between the EOS risk calculator and the Neosafe protocol
| Neosafe protocol | ||||
|---|---|---|---|---|
| EOS risk calculator | Exams | No | Yes | Total |
| No |
1959 (97,9%) |
7 (0,35%) |
||
| Yes |
41 (2,1%) |
41 (2,1%) |
82 (4,1%) |
|
| Total |
48 (2,4%) |
|||
McNemar’s test highlights a statistically significant difference in the probability of antibiotic administration between the two protocols, p < 0.0001 (Table 4). Cohen’s Kappa indicates a moderate agreement, with a value of 0.57 (95% CI: 0.47–0.68).
Table 4.
Comparison of antibiotic therapy between the EOS risk calculator and the Neosafe protocol
| Neosafe protocol | ||||
|---|---|---|---|---|
| EOS risk calculator | Antibiotic therapy | No | Yes | Total |
| No |
1913 (95,6%) |
7 (0,4%) |
||
| Yes |
44 (2,2%) |
36 (1,8%) |
80 (4,0%) |
|
| Total |
43 (2,2%) |
|||
A comparable proportion of newborns were identified as symptomatic (equivocal/clinical illness) by the two approaches: 5.6% (n = 111) using the EOS risk calculator and 6.8% (n = 135) with the Neosafe protocol. However, differences emerged in how each protocol categorized and interpreted clinical signs.
In the EOS risk calculator, 31 have been categorized as having “equivocal” and 80 neonates as having “clinical illness,” of them 57 (71.3%) required respiratory support (HFNC, nCPAP, or mechanical ventilation), 10 needed supplemental oxygen ≥ 2 h (12.5%), 10 neonates developed encephalopathy with seizures or Apgar < 5 at 5 min and 3 required vasoactive drugs due to hemodynamic instability.
In contrast, the Neosafe protocol stratifies symptoms into “equivocal” (n = 92) and “clinical illness” (n = 43), considering a broader range of signs. Among the “equivocal” group:
44.5% had mild respiratory distress with or without low-flow oxygen
35.9% had isolated moderate respiratory distress
10.8% showed altered muscle tone
smaller percentages presented with skin color changes, gastrointestinal symptoms, or tachycardia
According to the Neosafe classification, the most frequently observed conditions among neonates with “clinical illness” were:
29% had severe respiratory distress
14,5% had CNS alterations
27,9% had mild or moderate respiratory distress plus other physiological abnormalities lasting more than 2 h
Many neonates labeled as “clinical illness” by the EOS calculator were considered “equivocal” by Neosafe, largely due to isolated mild/moderate respiratory symptoms. Conversely, some signs considered significant by Neosafe (e.g., altered tone or skin color) are not included in the EOS calculator algorithm, leading to divergent categorizations. Notably, 22 neonates categorized as “equivocal” by Neosafe were labeled as “well-appearing” by the EOS calculator, 50% underwent diagnostic investigations and 3 started antibiotic therapy; no cases of confirmed sepsis were observed.
The EOS risk calculator recommends assessing neonatal vital signs at predefined intervals based on risk stratification. In contrast, the Neosafe protocol integrates serial physical examinations (SPE) with standardized monitoring sheets tailored to risk categories.
When applied retrospectively to our cohort, the EOS risk calculator would have assigned:
1892 neonates to routine vital sign monitoring (details not clearly specified by the tool)
28 neonates to monitoring every 4 h for 24 h
80 neonates to intensive monitoring in the NICU or neonatal unit
In comparison, the Neosafe protocol would have required:
1829 neonates to undergo routine obstetric-nursing observation using standard monitoring sheets
128 neonates to be evaluated by a neonatologist with advanced monitoring sheets
43 neonates to receive continuous vital sign monitoring during admission in the NICU or neonatal unit
Simulated protocol application: EOS calculator vs. Neosafe in a subgroup with maternal fever
Fever during labor was reported in 42 women (2.1%), of whom 8 had isolated fever and 34 met clinical criteria for suspected Triple I, as defined by Higgins et al. [35]. Placental histopathology was performed in 32 cases, confirming Triple I in 22.
Within this subgroup, 3 neonates would have undergone diagnostic evaluation under the Neosafe protocol, consistent with the number identified by the EOS risk calculator (Table 5).
Table 5.
Comparison of exams between the EOS risk calculator and the Neosafe protocol in neonates with maternal fever
| Neosafe protocol | ||||
|---|---|---|---|---|
| EOS risk calculator | Exams | No | Yes | Total |
| No |
39 (92,8%) |
1 (2,4%) |
||
| Yes |
1 (2,4%) |
2 (4,7%) |
3 (7,1%) |
|
| Total |
3 (7,1%) |
|||
Additionally, 3 neonates would have received antibiotic therapy under the Neosafe protocol, compared to 2 under the EOS risk calculator (Table 6).
Table 6.
Comparison of antibiotic therapy between the EOS risk calculator and the Neosafe protocol in neonates with maternal fever
| Neosafe protocol | ||||
|---|---|---|---|---|
| EOS risk calculator | Antibiotic therapy | No | Sì | Totale |
| No |
39 (92,9%) |
1 (2,3%) |
||
| Sì |
0 (0,0%) |
2 (4,8%) |
2 (4,8%) |
|
| Totale |
3 (7,1%) |
|||
McNemar’s test does not reveal any statistically significant difference between the two protocols in terms of either diagnostic exams or antibiotic administration. Although the sample size was small, Cohen’s kappa indicates a moderate level of agreement for exams of 0.64 (95% CI: 0.18–1.10) and a good level of agreement for antibiotic therapy of 0.79 (95% CI: 0.39–1.19).
Positive blood cultures and case summaries
During the study period, four neonates had positive blood cultures: one for Enterococcus faecalis and three for GBS.
The E. faecalis case involved a clinically stable preterm infant (36 weeks GA) with no signs of infection. The finding was considered likely contaminant. Both the EOS calculator and Neosafe protocol assigned a risk score of 0.06, indicating routine observation.
Case 1 GBS involved a term newborn who developed hematochezia at 24 h. Despite a low risk score (0.02) under both tools, early clinical alert led to prompt evaluation and treatment. Neosafe suggested advanced monitoring sheets for a gastrointestinal sign.
Case 2 GBS presented with meconium-stained amniotic fluid and elevated CRP (33 mg/L) without symptoms. Broad-spectrum antibiotics were started after positive culture. Both tools gave a risk score of 0.02, suggesting no initial intervention.
Case 3 GBS involved a term infant who developed neurological symptoms and meningitis at 40 h of life. Risk scores reached 1.25, prompting appropriate NICU-level care and treatment under both protocols.
Discussion
Early-onset neonatal sepsis is a complex and potentially severe condition that requires prompt diagnosis and appropriate decision-making for treatment. Identifying the appropriate diagnostic and therapeutic approach remains a challenge for neonatologists managing patients at risk of early-onset sepsis on a daily basis [3].
In this context, we propose the Neosafe protocol, designed for neonates ≥ 35 weeks’ gestation, as an integrative approach combining the strengths of the EOS risk calculator and serial physical examinations (SPE), particularly valuable in ambiguous cases such as those previously described.
In our simulated cohort, the Neosafe protocol would have led to significantly fewer diagnostic evaluations (2.4% vs. 4.1%) and antibiotic treatments (2.2% vs. 4.0%) compared to the EOS calculator, with statistically significant reductions in both outcomes.
A key divergence between the two approaches lies in the interpretation of respiratory symptoms, the most common yet non-specific clinical sign in the early neonatal period. 44 neonates categorized as “clinical illness” by the EOS calculator were reclassified as “equivocal” by Neosafe, avoiding unnecessary blood tests and antibiotics. This reclassification was grounded in the understanding that respiratory distress, particularly TTN, often mimics sepsis and leads to overtreatment, as previously noted by Cavigioli et al. [37]. By introducing a tiered classification of respiratory distress (mild, moderate, severe), the Neosafe protocol supports clinician to led observation before initiating a full diagnostic-therapeutic pathway.
Thus, if a patient exhibits only mild or moderate respiratory symptoms, the Neosafe protocol classifies the presentation as “equivocal,” leaving it to the clinician’s judgment to monitor for the appearance or progression of other clinical signs before initiating antibiotic therapy.
The protocol also refines the use of laboratory biomarkers, reserving CBC, CRP, and blood culture for neonates with concerning clinical signs rather than routine application. This targeted strategy may increases the positive predictive value of blood investigations and aligns with current concerns regarding the low specificity of inflammatory markers in the immediate neonatal period.
Importantly, Neosafe proposes a substantial revision of clinical classifications, recognizing presentations overlooked by the EOS calculator. In our cohort, seven neonates would have been classified as “well-appearing” or “equivocal” by the EOS calculator but were reclassified as “clinical illness” due to persistent multi-system abnormalities (e.g., respiratory distress plus altered tone or perfusion). Although all these neonates had negative blood cultures; they, in our view, represent a subgroup at increased risk of infection, in whom early intervention may be life-saving. The primary aim of any protocol designed to prevent early-onset sepsis (EOS) is the prompt identification of truly infected neonates. In this context, the Neosafe protocol appears to be a useful tool, as it reduces the number of neonates considered for treatment; however, its ability to accurately target therapy exclusively to cases of true infection could be further refined.
The Neosafe protocol’s emphasis on structured clinical observation, implemented via standard and advanced monitoring sheet, could provides a practical framework for continuous assessment. While this intensified clinical surveillance (91.5% standard monitoring sheet and 8.5% advanced monitoring sheet) may modestly increase the workload, it represents a targeted, high-yield investment in care quality. Importantly, it may also help prevent unnecessary NICU admissions and support rooming-in practices, which are essential for early bonding and breastfeeding. In this study, a comparison with clinical observation alone was not feasible. In our context, and considering the need for adequate training of all healthcare staff, a framework offering two levels of observation based on maternal-neonatal risk factors and clearly defined clinical criteria may represent a supportive and pragmatic approach.
Regarding proven EOS, two out of four neonates with positive blood cultures would have been promptly identified by Neosafe protocol due to clinical signs. The remaining two cases, one with GBS and one with Enterococcus faecalis, raise questions about the clinical relevance of positive cultures in well-appearing infants. These findings may indicate a sample contamination (in the case of Enterococcus faecalis) or transient bacteremia (in the case of GBS). Certainly, identifying 2 of 4 cases (or 2 of 3 if the first case is considered contaminated) may give the impression Neosafe protocol is insufficiently safe. However, this interpretation should be weighed against the very small sample size and the highly atypical presentation of the reported cases, with the exception of the last case.
A final point of divergence lies in the management of neonates born to febrile mothers. The Neosafe protocol, incorporating Higgins’ criteria for Triple I, adopts a cautious but proportionate approach. In fact it recommends advanced SPE monitoring for this subgroup, enabling early detection detection of any clinical signs without unnecessary overtreatment.
It is conceivable that an even more conservative approach could be achieved by extending advanced SPE monitoring to neonates presenting with equivocal clinical signs and this is supported by current evidence and expert consensus [35].
Conclusions
In this pilot study, our simulation suggests that implementing a more structured protocol for the management of EOS risk in newborns ≥ 35 weeks’ gestation, based on risk stratification and through clinical observation, may represent a promising strategy. The primary aim is (1) to reduce unnecessary blood tests, which can often lead to misleading interpretations and consequently (2) to limit the inappropriate use of antibiotic therapy. Such an approach aligns with the principles of diagnostic and antimicrobial stewardship, aiming to optimize clinical decision-making while minimizing unwarranted interventions. The safety of the newborn and the early identification of potentially infected infants remain the highest priorities. Therefore, we underscore the importance of meticulous collection of obstetric and neonatal history, careful clinical monitoring and the involvement of experienced and qualified healthcare professionals. Although no optimal protocol has yet been established, as several authors have noted, we believe that any new approach should be grounded in a balanced combination of careful clinical evaluation, the enduring value of neonatal semeiotics and the enhancement provided by technological and algorithmic innovations.
Acknowledgements
Not applicable.
Abbreviations
- EOS
Early-onset sepsis
- GBS
Group B Streptococcus
- PROM
Prolonged rupture of membranes
- IAP
Intrapartum antibiotic prophylaxis
- TTN
Transient tachypnea of the newborn
- CBC
Complete blood count
- CRP
C-reactive protein
- PPV
Positive predictive value
- AAP
American academy of pediatrics
- CDC
Centers for disease control and prevention
- SPE
Serial physical examination
- NICU
Neonatal intensive care unit
- Triple I
Intrauterine inflammation, infection, or both
- SUPC
Sudden unexpected postnatal collapse
- AGA
Appropriate for gestational age
- SGA
Small for gestational age
- LGA
Large for gestational age
Author contributions
SA (Medical Fellow/Medical Doctor) drew the first draft with the assistance and contribution of SP and AB (Medical Doctor) under the supervision of MA (Director of Maternal and Child Department), SP and MA revised the final draft. All authors contributed to the intellectual contents and approved the final version.
Funding
This study was not supported by any sponsor or funder.
Data availability
Available upon request to the corresponding author.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.American Academy of Pediatrics. Red Book: report of the Committee on Infectious Diseases. 32nd ed. 2021–2024.
- 2.Jefferies AL. Management of term infants at increased risk for early-onset bacterial sepsis. Paediatr Child Health. 2017;22(4):223–8. 10.1093/pch/pxx023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Puopolo KM, Benitz WE, Zaoutis TE, Committee on Fetus and Newborn; Committee on Infectious Diseases. Management of neonates born at ≥ 35 0/7 weeks’ gestation with suspected or proven early-onset bacterial sepsis. Pediatrics. 2018;142(6):e20182894. 10.1542/peds.2018-2894. [DOI] [PubMed] [Google Scholar]
- 4.Puopolo KM, Draper D, Wi S, Newman TB, Zupancic J, Lieberman E, et al. Estimating the probability of neonatal early-onset infection on the basis of maternal risk factors. Pediatrics. 2011;128(5):e1155–63. 10.1542/peds.2010-3464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Steinberg CB, Roth P. Early-onset sepsis in newborns. Pediatr Rev. 2023;44(1):14–22. [DOI] [PubMed] [Google Scholar]
- 6.Wolkowiez MC, Moran C, Benjamin DK, Cotten CM, Clark RH, Benjamin DK Jr, Smith PB. Early and late onset sepsis in late preterm infants. Pediatr Infect Dis J. 2009;28(12):1052–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Shane AL, Sánchez PJ, Stoll BJ. Neonatal sepsis. Lancet. 2017;390(10104):1770–80. 10.1016/S0140-6736(17)31002-4. [DOI] [PubMed] [Google Scholar]
- 8.Polin RA, Committee on Fetus and Newborn. Management of neonates with suspected or proven early-onset bacterial sepsis. Pediatrics. 2012;129(5):1006–15. 10.1542/peds.2012-0541. [DOI] [PubMed] [Google Scholar]
- 9.Weintraub AS, Cadet CT, Perez R, et al. Antibiotic use in newborns with transient tachypnea of the newborn. Neonatology. 2013;103:235–40. [DOI] [PubMed] [Google Scholar]
- 10.Morrison JJ, Rennie JM, Milton PJ. Neonatal respiratory morbidity and mode of delivery at term: Influence of timing of elective caesarean section. Br J Obstet Gynaecol. 1995;102:101–6. [DOI] [PubMed] [Google Scholar]
- 11.Wynn JL, Wong HR, Shanley TP, et al. Time for a neonatal-specific consensus definition for sepsis. Pediatr Crit Care Med. 2014;15:523–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cantey JB, Patel SJ. Antimicrobial stewardship in the NICU. Infect Dis Clin North Am. 2014;28:247–61. [DOI] [PubMed] [Google Scholar]
- 13.Klingenberg C, Kornelisse RF, Buonocore G, Maier RF, Stocker M. Culture negative early-onset neonatal sepsis—at the crossroad between efficient sepsis care and antimicrobial stewardship. Front Pediatr. 2018;6:285. 10.3389/fped.2018.00285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cantey JB, Lee JH. Biomarkers for the diagnosis of neonatal sepsis. Clin Perinatol. 2021;48(2):215–27. 10.1016/j.clp.2021.03.012. [DOI] [PubMed] [Google Scholar]
- 15.Schelonka RL, Chai MK, Yoder BA, et al. Volume of blood required to detect common neonatal pathogens. J Pediatr. 1996;129:275–8. [DOI] [PubMed] [Google Scholar]
- 16.Connell TG, Rele M, Cowley D, Buttery JP, Curtis N. How reliable is a negative blood culture result? Volume of blood submitted for culture in routine practice in a children’s hospital. Pediatrics. 2007;119(5):891–6. 10.1542/peds.2006-0440. [DOI] [PubMed] [Google Scholar]
- 17.Howman RA, Charles AK, Jacques A, et al. Inflammatory and haematological markers in the maternal, umbilical cord and infant circulation in histological chorioamnionitis. PLoS ONE. 2012;7:e51836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Marins LR, Anizelli LB, Romanowski MD, et al. How does preeclampsia affect neonates? Highlights in the disease’s immunity. J Matern Fetal Neonatal Med. 2019;32:1205–12. [DOI] [PubMed] [Google Scholar]
- 19.Tiozzo C, Mukhopadhyay S. Noninfectious influencers of early-onset sepsis biomarkers. Pediatr Res. 2022;91(2):425–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Stocker M, Fontana M, El Helou S, Wegscheider K, Berger TM. Use of procalcitonin guided decision-making to shorten antibiotic therapy in suspected neonatal early onset sepsis: A prospective randomized intervention trial. Neonatology. 2010;97(2):165–74. [DOI] [PubMed] [Google Scholar]
- 21.Eschborn S, Weitkamp JH. Procalcitonin versus C-reactive protein: Review of kinetics and performance for diagnosis of neonatal sepsis. J Perinatol. 2019;39(7):893–903. [DOI] [PubMed] [Google Scholar]
- 22.Hincu M-A, Zonda G-I, Stanciu GD, Nemescu D, Păduraru L. Relevance of biomarkers currently in use or research for practical diagnosis approach of neonatal early-onset sepsis. Child (Basel). 2020;7(12):309. 10.3390/children7120309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Van Herk W, el Helou S, Janota J, Hagmann C, Klingenberg C, Staub E, et al. Variation in current management of term and late-preterm neonates at risk for early-onset sepsis: An international survey and review of guidelines. Pediatr Infect Dis J. 2016;35(5):494–500. [DOI] [PubMed] [Google Scholar]
- 24.Giannoni E, Dimopoulou V, Klingenberg C, Navér L, Nordberg V, Berardi A, et al. Analysis of antibiotic exposure and early onset neonatal sepsis in Europe, North America, and Australia. JAMA Netw Open. 2022;5(11):e2243691. 10.1001/jamanetworkopen.2022.43691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Prevention of Group B Streptococcal. Early-Onset Disease in Newborns: ACOG Committee Opinion, Number 797. Obstet Gynecol. 2020;135(2):e51–72. [DOI] [PubMed] [Google Scholar]
- 26.Verani JR, McGee L, Schrag SJ. Prevention of perinatal group B streptococcal disease: revised guidelines from CDC. MMWR Recomm Rep. 2010;59(RR-10). [PubMed]
- 27.Kuzniewicz MW, Puopolo KM, Fischer A, Walsh EM, Li S, Newman TB. A quantitative, risk-based approach to the management of neonatal early-onset sepsis. JAMA Pediatr. 2017;171(4):365–71. 10.1001/jamapediatrics.2016.4678. [DOI] [PubMed] [Google Scholar]
- 28.Mukhopadhyay S, Eichenwald EC, Puopolo KM. Neonatal early onset sepsis evaluations among well appearing infants: Projected impact of changes in CDC GBS guidelines. J Perinatol. 2013;33(3):198–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Escobar GJ, Puopolo KM, Wi S, et al. Stratification of risk of early-onset sepsis in newborns ≥ 34 weeks’ gestation. Pediatrics. 2014;133(1):30–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Northern California Kaiser Permanente. Neonatal Early Onset Sepsis Calculator [Internet]. 2018 Apr 5. Available from: https://neonatalsepsiscalculator.kaiserpermanente.org.
- 31.Berardi A, Fornaciari S, Rossi C, et al. Safety of physical examination alone for managing well appearing neonates ≥ 35 weeks’ gestation at risk for early-onset sepsis. J Matern Fetal Neonatal Med. 2015;28(10):1123–7. [DOI] [PubMed] [Google Scholar]
- 32.Joshi NS, Gupta A, Allan JM, et al. Clinical monitoring of well appearing infants born to mothers with chorioamnionitis. Pediatrics. 2018;141(4):e20172056. [DOI] [PubMed] [Google Scholar]
- 33.Laccetta G, Ciantelli M, Tuoni C, Sigali E, Miccoli M, Cuttano A. Early-onset sepsis risk calculator: A review of its effectiveness and comparative study with our evidence-based local guidelines. Ital J Pediatr. 2021;47(1):73. 10.1186/s13052-021-01028-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Barbini MC, Perniciaro S, Bresesti I, Folgori L, Barcellini L, Bossi A, Agosti M. The management of neonates ≥ 34 weeks’ gestation at risk of early-onset sepsis: A pilot study. Antibiotics. 2023;12(8):1306. 10.3390/antibiotics12081306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Higgins RD, Saade G, Polin RA, et al. Evaluation and management of women and newborns with a maternal diagnosis of chorioamnionitis: Summary of a workshop. Obstet Gynecol. 2016;127(3):426–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Coscia A, Davanzo R, Ilardi L, et al. Guidelines on the prevention and management of sudden unexpected postnatal collapse (SUPC). Italian Society of Neonatology; 2023.
- 37.Cavigioli F, Viaroli F, Daniele I, et al. Neonatal early onset sepsis (EOS) calculator plus universal serial physical examination (SPE): A prospective two step implementation of a neonatal EOS prevention protocol for reduction of sepsis workup and antibiotic treatment. Antibiotics. 2022;11(8):1089. 10.3390/antibiotics11081089. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Available upon request to the corresponding author.