A retrospective study of risk factors for early-onset neonatal sepsis with intrapartum maternal fever

Introduction

Early-onset neonatal sepsis (EOS) is defined as sepsis that manifests within the first 72 h of life and is usually caused by vertical transmission through contaminated amniotic fluid or during vaginal delivery (Stoll et al., 2020; Benitz & Achten, 2021). The most common and characteristic manifestation of neonatal sepsis is a change in feeding behavior in late-onset sepsis and respiratory distress in early-onset sepsis. The baby, who had been active and sucking well, gradually or suddenly becomes lethargic, inactive or unresponsive, and refuses to suckle. Hypothermia is a common manifestation of sepsis, whilst fever is infrequent. Diarrhea, vomiting and abdominal distension may occur. Episodes of apneic spells or gasping may be the only manifestation of septicemia. A critical neonate may develop shock, bleeding and renal failure (Shane, Sánchez & Stoll, 2017). Despite advances in obstetric care, early-onset neonatal sepsis remains a clinical challenge that is associated with high neonatal morbidity and mortality rates (Fleischmann-Struzek et al., 2018; Kim, Polin & Hooven, 2020). In 2018, neonatal sepsis was estimated at a population level of 2,202 cases per 100,000 live births, with a mortality rate of between 11% and 19%. Globally, this represents three million cases of sepsis each year (Fleischmann-Struzek et al., 2018). Given these alarming statistics, it is important to find methods of early diagnosis, investigate potential causes of EOS and develop better treatment options.

Clinical signs of neonatal sepsis range from sub-clinical infection to severe manifestations of focal or systemic disease (Shane, Sánchez & Stoll, 2017). Due to its broad spectrum of clinical manifestations, it is difficult to diagnose early neonatal sepsis. Several researchers have described the risk factors associated with neonatal sepsis (Stoll et al., 2011). These include: delivery at <37 weeks of gestation (Soman, Green & Daling, 1985), maternal Group B Streptococcal (GBS) infection (Stoll et al., 2011), prolonged rupture of membranes ≥ 18 h (Herbst & Källén, 2007), and maternal chorioamnionitis (Yancey et al., 1996). Current research has focused on developing a predictive model for late-onset neonatal sepsis or EOS in full-term and late preterm infants only (Kudawla, Dutta & Narang, 2008; Escobar et al., 2014).

Intrapartum fever is defined as a temperature ≥ 38 degree Celsius (38 °C) during labour or within 24 h after delivery (Ashwal et al., 2018). Previous studies suggest that the rate of intrapartum fever is 3.3%–7% (Wendel et al., 2002; Braun et al., 2016). Peripartum maternal fever of 38 °C or higher is a risk factor for EOS, and if present, requires prophylactic antibiotic treatment for infants, even for clinically healthy newborns (Towers et al., 2017). The US Centers for Disease Control and Prevention (CDC) guidelines state that “in an effort to avert neonatal infections, maternal fever alone in labor may be used as a sign of chorioamnionitis and hence indication for antibiotic treatment” (American College of Obstetricians and Gynecologists, 2011; Committee on Infectious Diseases et al., 2011). Facilities that choose to follow these guidelines need to identify these newborns and begin intravenous antibiotic therapy. Some facilities may be able to treat newborns with intravenous antibiotics in a general neonatal unit where the mother and baby are in the same room, while others may need to transfer the newborn to a neonatal intensive care unit for treatment, which may separate the mother from the newborn. In most hospitals in China, newborns stay with their mothers on the postnatal floor and there is no special nursery for healthy newborns; intravenous antibiotics are often administered in the NICU.

Given the variety of causes of intrapartum maternal fever, its association with neonatal sepsis has not been fully determined (Ashwal et al., 2018). Previous studies have investigated the role of maternal fever as an indicator of neonatal sepsis with inconsistent results (Towers et al., 2017; Ashwal et al., 2018). In 2017, a study by Towers et al. (2017) found that 0.24% of 412 newborns of mothers with intrapartum fever were diagnosed with neonatal sepsis, but less than 0.0007% of the remaining 5,697 newborns of mothers without fever received a neonatal sepsis diagnosis. In contrast, another study found that out of 13,887 live births, 18 full-term newborns developed infections, all of which occurred prior to maternal fever (Wendel et al., 2002). Maternal fever may suggest maternal-fetal transmission of infection leading to neonatal sepsis, but it may also be non-specific. Neonates born to mothers without fever can also develop sepsis, highlighting the uncertainty of the correlation between maternal fever and EOS (Gupta et al., 2021). Therefore, it may not be reasonable to routinely administer antibiotics to all newborns of mothers with fever at delivery. Treating healthy neonates with unneeded antibiotics can be detrimental (Strunk et al., 2018a; Strunk et al., 2018b). Longer hospital stays for tests and treatment may lead to higher medical costs. The infant microbiome is also very sensitive, and exposure to avoidable antibiotics may have long-term health outcomes; prescribing avoidable antibiotics also impacts overall antibiotic stewardship (Gupta et al., 2021).

China has a large population base and a large number of births each year with many newborns suffering from sepsis. Few studies have been reported on intrapartum maternal fever and neonatal sepsis. It is important to study the relationship between maternal fever and neonatal sepsis at the frontline of clinical practice in China. Therefore, the aim of this study was to identify the predictors of EOS. We retrospectively evaluated neonates with febrile mothers during delivery and explored the risk factors of neonatal sepsis to inform the early clinicalassessment and prediction of neonatal sepsis risk.

Materials and Methods

Patients and data selection

A retrospective cohort study involving neonates born to mothers with intrapartum fever in Hangzhou Women’s Hospital was performed from January 1st to December 31st, 2019. The medical records of the subjects between 37^0/7 and 41^0/7 gestational age who experienced active labor (induced or spontaneous) and experienced a systemic fever of ≥ 38.0 °C along with the records of their respective neonates were reviewed retrospectively. Subjects whose medical records were incomplete, experienced non-singleton gestations, stillbirths, scheduled cesarean sections, had infants delivered prior to 37^0/7 gestational age, or had congenital fetal anomalies were not included in this analysis.

The oral temperature of the parturients were measured routinely prior to entering the labor room, as well as every eight hours thereafter. When the oral temperature measured ≥ 37.5 °C, additional temperature measurements followed in the next hour. These procedures are summarized in the flow diagram in Fig. 1. Maternal baseline data, intrapartum manifestations, neonatal outcomes, and obstetric outcomes were acquired through internal reviews of the electronic medical records. This study was approved by the Medical Ethics Committee of Hangzhou Women’s Hospital (2020A10-24). The participants provided written informed consent to participate in this study.

Results

Discussion

In this study, the overall incidence of fever of more than 38 °C was 4.93% (592/12000), which is similar to previous studies (Towers et al., 2017). The overall incidence of fever in mothers of newborns with culture-proven sepsis was 9.86% (56/568), which is much lower than other published reports (Wortham et al., 2016; Gupta et al., 2021), likely due to non-infectious maternal fevers. This is supported by previous studies, which have reported that 40% of women with intrapartum fevers had non-infectious fevers (Kim et al., 2015). Maternal temperature increases as a result of inflammation and may also be influenced by various factors such as the activity and intensity of uterine contractions, as well as physiological changes that occur during labor such as an increase in basal metabolic rates (Anim-Somuah et al., 2018). Other causes of intrapartum fever include antepartum infections, such as urinary/respiratory tract infections, and side effects from medications taken during labor (Gupta et al., 2021). Antepartum and intrapartum characteristics provide important information regarding exposure to infectious diseases and inform neonatologists of obstetric risk factors for neonatal infections (American College of Obstetricians and Gynecologists’ Task Force on Neonatal Encephalopathy, 2014).

Intrapartum fever is monitored to prevent the transfer of infections to newborns with immature immune systems (Gupta et al., 2021). We evaluated the differences in maternal characteristics before and after delivery as well as the obstetrical and neonatal outcomes between the EOS and non-sepsis groups in this study. The results showed that neonates with EOS had mothers with higher peak fever temperatures, higher prenatal WBC levels, and higher CRP levels before delivery. Neonates with EOS born to mothers with intrapartum fevers were also more likely have higher rates of NICU admission than neonates with EOS born to nonfebrile mothers, and mothers with intrapartum fevers whose neonates had EOS had a higher incidence of HCA. This finding indicates that maternal infection plays a role in neonatal EOS. We investigated the risk factors for EOS in neonates born to febrile mothers and evaluated their associations. The results indicated that prenatal maternal WBC elevation and a maternal pathological diagnosis of chorioamnionitis were found to be independent risk factors for neonatal EOS. Our results differ from previous studies that found that the duration of ruptured membranes ≥ 18 h and epidural analgesia were risk factors for maternal fever and neonatal sepsis (Wassen et al., 2014). Randis et al. (2018) reported that babies born to mothers with amnionitis were almost three times more likely to be diagnosed with suspected sepsis compared to those born to mothers who had not been exposed. Our findings are in line with this conclusion. Globally, clinical chorioamnionitis is the most common infection-associated complication in delivery units, affecting 1–6% of pregnancies in the United States (Perry, Rossi & De Franco, 2019; Venkatesh et al., 2019a; Venkatesh et al., 2019b). This syndrome is a risk factor for adverse maternal outcomes, including postpartum hemorrhage followed by uterine atony and sepsis (Chebbo et al., 2016). Newborns born to mothers diagnosed with clinical chorioamnionitis have a higher risk of sepsis (Gowda et al., 2017; Randis et al., 2018), admission to the neonatal intensive care unit (Venkatesh et al., 2019a; Venkatesh et al., 2019b), and neonatal death compared to neonates born to women without this syndrome (Conde-Agudelo et al., 2020; Racusin et al., 2021).

A multivariate logistic regression analysis revealed that maternal WBC levels before delivery are independently associated with EOS risk, suggesting that elevated prenatal white blood cell counts in mothers indicates an increased risk of EOS. WBC count is widely used in clinical practice. Clinically, it is very important to determine the appropriate maternal WBC count cut-off point for early identification of neonatal EOS. Further WBC investigations using an ROC curve analysis in the neonatal EOS group and non-sepsis group revealed a cut-off value of maternal WBC to be 16.75 × 10*⁹/L with a 71.40% sensitivity and an 82.20%. Maternal leukocytosis (WBC ≥ 15.00 × 10*⁹/L) is one of the diagnostic criteria for clinical chorioamnionitis (Gibbs et al., 1982; Yoder et al., 1983). Our results suggest that a maternal WBC ≥ 16.75 × 10*⁹/L is not only indicative of chorioamnionitis but also increases the possibility of neonatal EOS. Using objective clinical data available at birth, Ryan et al. (2019) found that maternal CRP >10 mg/L was associated with a fetal inflammatory response and increased risk of EOS. Puopolo et al. designed a predictive model for EOS for infants born at ≥ 34 weeks gestation, they found that the two most important predictors were gestational age and maternal fever (American College of Obstetricians and Gynecologists’ Task Force on Neonatal Encephalopathy, 2014). Their model performed well with an AUC of 0.807, which was similar to the AUC of our study.

Our study has several strengths. First, this is one of the few studies in China to evaluate risk factors for early-onset neonatal sepsis with intrapartum maternal fever. Secondly, this was a single-centre cohort study in which all women and newborns received uniform diagnostic criteria and clinical management strategies and all patients with fever underwent placental pathology, which ensured the integrity of the retrospective cohort.

This study has its limitation. As the data collection is retrospective, there is a potential risk of information bias. In order to curtail this potential bias, cases were meticulously screened and reviewed. In addition, this study relied on blood cultures to demonstrate sepsis. However, although blood cultures are the gold standard, they are less sensitive due to the use of antibiotics at delivery and the limitations of blood volume per culture. Therefore, some children with negative blood cultures were still clinically diagnosed with neonatal sepsis based on clinical presentation and other laboratory findings. Since all cases were selected from Hangzhou Women’s Hospital, there is a possibility of results having confounding variables that are inevitable in all single-centered studies. Therefore, large-scale multicenter prospective clinical studies are recommended to confirm our findings. In addition, the practical and clinical significance of the findings is not entirely clear and needs to be further confirmed in future multicentre studies.

Conclusions

In conclusion, we found that high maternal WBC counts before delivery and maternal HCA diagnosis are both independently risk factors for EOS and that maternal WBC levels before delivery was independently associated with EOS risk. ROC curve analysis showed that a maternal WBC level of 16.75 × 10*⁹/L before delivery, with a sensitivity and specificity of 71.40% and 82.20%, respectively, was a good predictor for EOS. Prenatal maternal WBC counts elevation can be used as a sensitive indicator for the early prediction of EOS, providing an important reference for neonatal pediatricians.

Supplemental Information