Predictive factors for perinatal bacterial transmission from colonized mothers to delivered very-low-birth-weight infants: a retrospective cohort study

Abstract

This study investigated the predictive factors for perinatal bacterial transmission in very-low-birth-weight infants (VLBWIs) born to mothers with a history of intrapartum colonization. We retrospectively reviewed the medical records of 173 VLBWIs, wherein pathogens were confirmed in maternal cultures obtained from the blood, urine, and vagina during the intrapartum period from 2013 to 2020. Newborns were categorized based on microbiological tests, including gastric aspirates, endotracheal aspirates, blood, and skin/nasal swab cultures collected immediately after birth. Infants whose cultures matched their maternal pathogens were categorized into the “transmission group” (n = 45), while those who tested negative were assigned to the “control group” (n = 128). The predominant maternal-colonizing pathogen observed was Escherichia coli (30.6%), which also emerged as the primary colonizing pathogen in neonates (35.6%). Transmission group had higher incidences of maternal leukocytosis, chorioamnionitis, and cervical cerclage. Regarding neonatal characteristics, the transmission group demonstrated lower initial base excesses (− 6.3 ± 3.9 vs. − 9.2 ± 4.9, P < 0.05) and higher C-reactive protein levels (0.1 ± 0.3 vs. 0.4 ± 0.8, P < 0.05). Notably, regarding major neonatal outcomes, transmission group had higher mortality rates and incidences of severe intraventricular hemorrhage. These findings may be useful for making decisions when considering antibiotic treatment for infants with a history of maternal colonization.

Introduction

Early-onset neonatal sepsis (EOS) is a significant global public health concern that increases the risks of morbidity and mortality among neonates, especially preterm infants. Advances in medical technology have improved the survival rates of very preterm infants in recent decades^1,2; however, sepsis remains the leading cause of death in infants with very low birth weight (VLBWIs; birth weight < 1500 g)³.

These infants are susceptible to infections transmitted from maternal sources, including the gastrointestinal and genitourinary tracts, amniotic fluid, chorioamnionitis, and maternal bacteremia, either before or during delivery^4,5. Preterm infants are at high risk of developing infections due to their underdeveloped immune systems. Certain colonizing bacteria may lead to serious infections, such as pneumonia and sepsis, in these infants^6,7. Symptoms of such infections are often present in vague and nonspecific ways, posing challenges for early diagnosis in VLBWIs. Consequently, clinicians often resort to empirically administering broad-spectrum antibiotics to prevent potential bacterial transmission. However, the indiscriminate use of broad-spectrum antibiotics yields numerous adverse outcomes, including necrotizing enterocolitis (NEC), bronchopulmonary dysplasia (BPD), and poor neurodevelopmental outcomes^8–10. Therefore, identifying high-risk groups for perinatal transmission and administering appropriate antibiotics are important. By stratifying patients into high- and low-risk groups for perinatal transmission, antibiotic strategies can be more precisely tailored. Specifically, antibiotic use should be limited to high-risk groups, thereby curtailing their overuse in the broader population. This approach serves to alleviate the practice of indiscriminate antibiotic prescription driven by fear of potential infections caused by colonizing pathogens.

Several recent studies have attempted to predict EOS by leveraging a combination of biochemical markers and innovative tools such as a neonatal early-onset sepsis calculator^11–14. Additionally, research has explored perinatal transmission, recognizing that even colonizing pathogens in infants can cause infections. However, the majority of these studies have targeted term or late preterm infants. Information regarding predictive factors for bacterial transmission stemming from maternal colonization in VLBWIs remains limited. Therefore, we investigated the predictive maternal and neonatal elements influencing perinatal bacterial transmission in VLBWIs with a history of maternal colonization or infection, such as bacteremia, urinary tract infection (UTI), and vaginitis.

Methods

Ethics statement

The data collection procedure was approved by the Institutional Review Board of the Samsung Medical Center (2023-02-154), which waived the requirement for informed consent for this retrospective chart review. All methods were performed in accordance with an IRB-approved protocol and in compliance with the relevant guidelines and regulations.

Study design

The medical records of 935 VLBWIs (birth weight < 1500 g) born and admitted to the Samsung Medical Center neonatal intensive care unit (NICU) between January 2013 and December 2020 were retrospectively reviewed. Among these, 193 VLBWIs were born to mothers with bacterial colonization. Maternal colonization was defined as the presence of any Gram-negative or Gram-positive bacteria or fungi, excluding normal flora, from cultures obtained from the endocervix and upper vagina, in addition to urine, blood, or cerclage knots. For cerclage knot culture, we specifically removed and cultured the knot immediately before delivery using semi-quantitative methods on agar plates to identify the presence of microbes. We excluded infants whose culture results within the first seven days of life indicated bacterial strains different from those detected in the mother (n = 20). Infants who tested positive for the same pathogen as their mother, as confirmed by blood, endotracheal aspirate, gastric aspirate, and skin tests within the first seven days of life, were assigned to the transmission group (n = 45). Transmission was confirmed only when the antibiotic susceptibility patterns of the colonized pathogen in mothers and neonates were identical. Infants who tested negative on these cultures were defined as the control group (n = 128). The transmission group was further subdivided into a blood culture-confirmed EOS group and a transmission without EOS group (Fig. 1).

Figure 1 [Images not available. See PDF.]

Flow chart of study population.

We compared maternal variables between the two groups to investigate the factors affecting perinatal transmission from mothers to VLBWIs. We also compared neonatal baseline characteristics and major outcomes between the two groups. Additionally, to assess whether bacterial transmission without EOS could affect neonatal outcomes, we analyzed the major neonatal clinical outcomes in the transmission group, excluding infants with EOS. EOS was defined as the presence of bacteria in blood cultures obtained within the first week of life.

Data collection

We analyzed various maternal clinical characteristics. These factors include maternal age, gestational diabetes mellitus, and oligohydramnios. We also considered instances of preterm premature rupture of membranes (PPROM) over 24 h and placental abruption. We also analyzed the use of antenatal antibiotics, history of amniocentesis, and chorioamnionitis. In this study, acute chorioamnionitis was defined histopathologically. Stage 0 indicates no inflammation; stage 1 indicates initial inflammation, such as acute subchorionitis or chorionitis; stage 2 indicates more extensive acute chorioamnionitis; and stage 3 indicates severe necrotizing inflammation. Stages 1–3 were classified as pathological chorioamnionitis. We accounted for instances of cervical cerclage, maternal fever (≥ 38 °C), maternal tachycardia (> 100/min), and fetal tachycardia (> 160/min). Laboratory results, such as C-reactive protein (CRP) levels and absolute neutrophil counts (ANC), were also considered. Finally, we included maternal culture outcomes of the endocervix, upper vagina, urine, blood, and cerclage knots during the entire intrapartum period in our analysis. According to our institutional protocols, vaginal culture is routinely performed in every mother at high risk of spontaneous birth at admission, limited to a gestational age (GA) of 14–34 weeks, using a sterile cotton swab with a sterile speculum. Cerclage knot cultures were performed immediately before delivery if cervical cerclage was performed during the intrapartum period. Blood culture was performed only when the mother exhibited symptoms indicating sepsis, such as fever. Our intrapartum antibiotic prophylaxis protocol was initiated upon confirmation of PPROM by administering intravenous cefazolin and oral clarithromycin until delivery. This practice extends beyond the standard recommendation of usage within one week, based on findings from our previous institutional research indicating improved long-term neonatal outcomes with this extended regimen¹⁵. Immediate delivery is pursued in cases in which chorioamnionitis is clinically suspected to mitigate infection risks and enhance neonatal health outcomes. Data on several baseline neonatal characteristics were collected. These included the GA, birth weight (BW), sex, and delivery mode. We recorded the Apgar score at 1 and 5 min, the use of surfactant, and the need for positive pressure ventilation or endotracheal intubation as initial resuscitation. The initial laboratory results of blood pH, base excess, serum CRP level, white blood cell (WBC) count, and ANC on complete blood count (CBC) after birth were included. Laboratory tests were conducted within 2 h after birth according to our institutional NICU admission protocol. We collected culture outcomes of blood and endotracheal aspirates within the first week of life. Every infant admitted to the NICU underwent routine blood culture on the first day after birth. Experienced, skilled medical professionals conducted the tests. Before each test, medical staff placed the infants on sterile gloves and thoroughly cleaned their skin using 0.5% chlorhexidine. In case of infants who underwent endotracheal intubation, aspirates were carefully collected immediately after intubation using a closed tracheal suction system. Additionally, we included the culture outcomes of gastric aspirates and samples from the skin and nose. Shortly after birth, the medical staff wore sterile gloves, placed an orogastric tube, and collected the prefeed gastric aspirate. Skin and nasal cultures were routinely performed immediately after birth using sterile cotton swabs to collect samples for these assessments. Empirical antibiotic therapy is guided by the outcomes of maternal culture and clinical presentation of the neonate. During the hospital stay, if symptoms or signs arise that indicate a potential infection, as evidenced by unstable vital signs, such as hypotension, hypo/hyperthermia, hypoxia, or abnormal laboratory findings, including hyperglycemia and acidosis, a comprehensive pan-culture is reinitiated. Subsequent evaluation included cultures of blood, endotracheal secretions, skin, and when applicable, wound cultures.

The Clinical Risk Index for Babies (CRIB) II score was calculated for each infant using the following variables: sex, GA, BW, temperature at admission, and base excess.

We also compared neonatal major outcomes between two groups including mortality rates and major morbidities including intraventricular hemorrhage (IVH) ≥ grade III, NEC ≥ stage IIb, BPD ≥ moderate, periventricular leukomalacia (PVL), retinopathy of prematurity (ROP) ≥ stage III^16,17. Additionally, we established a subgroup within the transmission group, termed the ‘transmission without sepsis group’, which excluded the sepsis group in which a pathogen was identified in blood culture within the first week of life. Furthermore, we conducted a comparative analysis between this subgroup and the control group.

Statistical analysis

Continuous variables are presented as means ± standard deviation and were compared using Student’s t-test or the Mann–Whitney U test. Categorical variables are presented as percentages and frequencies and were compared using the chi-square or Fisher’s exact test. Statistical significance was set at P < 0.05. Statistical analyses were performed using STATA version 16.0 [StataCorp, College Station, TX, USA].

Results

Of the 173 preterm infants born to colonized mothers, 45 were colonized by pathogens from their mothers. The highest number of positive culture results was observed for gastric aspirates, skin, and nose (31 cases each), followed by blood and endotracheal aspirates (10 and 9 cases, respectively).

Figure 2 shows the route-dependent incidence of perinatal bacterial transmission from mother to infant. The vagina, including the cerclage knots, was the most common site of colonization and transmission. Among the vaginal culture positive mothers (totaling 166), 41 pregnant women showed bacterial transmission to their infants, which includes twins (8) and singletons (37). In cases where a pathogen was identified solely in maternal blood or urine, transmission did not occur. However, transmission occurred in all cases in which either blood or urine cultures were positive, which coincided with the positive vaginal cultures.

Figure 2 [Images not available. See PDF.]

The route dependent incidence of perinatal transmission. This diagram illustrates the sites of positive culture in the mother and how these correlates to the respective sites in the VLBWIs. The central silhouette represents the mother, with surrounding circles indicating the sites of colonization. Arrows extend from these circles to depict the number of cases where the same microorganism was identified in the neonates, and the specific sites of positive culture in the neonates are indicated at the arrowheads.

Our cohort comprised 107 individuals with singleton gestations and 66 with multifetal pregnancies. Specifically, the multifetal group included 34 twin gestations (representing 53 individuals) and 6 triplet gestations (encompassing 13 individuals).

Among the singletons, 25 neonates were confirmed to have acquired pathogens from their mothers. Within the twin subgroup, analysis indicated that 16 individuals were recipients of maternal pathogens. Notably, in the three pairs of twins, both siblings were colonized by identical pathogens identified as E. coli, Candida species, and Klebsiella species. Among these triplets, pathogen transmission was observed in four individuals, with two stemming from a single set of triplets, suggesting a common maternal source of E. coli infection.

Table 1 presents the distribution of colonizing pathogens in mothers and neonates. Among the 173 colonized mothers, 217 unique positive culture results were identified; among the 45 VLBWIs who acquired pathogens from their mothers, 47 unique positive culture results were identified. The culture results of the mothers revealed that E. coli were predominant (30.6%), followed by Candida spp. (29.5%). Similarly, the most frequently transmitted pathogen to neonates was E. coli (35.6%), followed by Candida spp. (20.0%). Of the 10 EOS cases in the transmission group, eight were caused by E. coli, and the other two were caused by Candida species.

Table 1. Distribution of colonized pathogens.

Organism	Numbers	Percent (%)
Maternal culture outcome (blood, vagina, cerclage knot, urine)
Escherichia coli	53	30.6
Candida species	51	29.5
Group B Stereptococcus	29	16.8
Klebsiella species	21	12.1
Enterococcus species	20	11.6
Neonatal culture outcome (blood, endotracheal/gastric aspirate, skin) among transmission group
Escherichia coli	16	35.6
Candida species	9	20.0
Group B Stereptococcus	7	15.6
Enterococcus species	7	15.6
Klebsiella species	4	8.9
Culture proven EOS among transmission group
Escherichia coli	8	17.8
Candida species	2	4.4

The percentages for mothers are calculated based on the total number of participants (n = 173), and for neonates, the percentages are based on the total number of transmission cases (n = 45). The overall total of unique colonization events identified in mothers is 217 and in neonates is 47, with a total of 45 transmission cases to neonates.

We analyzed pathogen distributions across different culture sites in mothers and VLBWIs. For mothers, E. coli was the predominant pathogen found in blood and vagina, accounting for 60% and 31.9% of cultures respectively. In urine samples, Enterococcus spp. led a diverse array of identified bacteria. For VLBWIs, E. coli was the most common pathogen across all sampled sites, representing the largest portion of pathogens detected. However, the small number of samples in each group may limit the generalizability of these findings. (Supplementary Table S1).

We also analyzed the transmission rates of pathogens identified from different maternal culture sites to neonates. A key observation is the high transmission rate of E. coli. Specifically, when E. coli was isolated from maternal blood, the transmission rate to neonates reached 100%, with 33.3% of these cases developing into neonatal sepsis. Additionally, E. coli had the highest transmission rate of 30.2% from the vagina, with 15.1% of these cases leading to EOS. Group B Streptococcus (GBS), when identified from the vagina, also showed a significant transmission rate of 24.1%, marking it as the second highest (Supplementary Table S2). The analysis is somewhat limited by the smaller number of cases identified from maternal blood and urine compared to the vagina. Nonetheless, the pronounced transmission rate of E. coli from both maternal blood and vagina is notable. This underscores the need for further large-scale, prospective studies to validate and expand upon these findings.

Colonized mothers were divided into control and transmission groups according to their transmission status. Table 2 shows the demographic and clinical characteristics of the patients. During the entire intrapartum period, mothers in the transmission group had significantly higher frequencies of cases with a WBC count of > 15,000/μL on CBC, pathologically-confirmed chorioamnionitis, and history of cervical cerclage compared to those in the control group.

Table 2. Baseline maternal characteristics.

Characteristics	Control group (n = 128)	Transmission group (n = 45)	OR (95% CI)
Maternal age, year	33.5 ± 3.7	33.8 ± 3.0	1.02 (0.93–1.13)
Gestational diabetes, n (%)	9/128 (7.0)	3/45 (6.7)	0.94 (0.24–3.65)
Maternal fever (> 38 °C), n (%)	7/128 (5.5)	5/45 (11.1)	2.16 (0.64–7.18)
Maternal tachycardia (> 100 per/min), n (%)	88/128 (68.8)	36/45 (80.0)	1.81 (0.80–4.13)
Fetal tachycardia (> 160 per/min), n (%)	28/127 (22.1)	16/45 (35.6)	1.95 (0.93–4.09)
Polyhydramnios, n (%)	1/98 (1.0)	0/36 (0.0)	1
Oligohydramnios, n (%)	30/128 (23.3)	9/45 (20.9)	0.87 (0.37–2.02)
PPROM ≥ 24 h, n (%)	41/128 (78.9)	23/31 (74.2)	0.77 (0.27–2.19)
Placental abruption, n (%)	16/128 (12.5)	4/45 (8.9)	0.68 (0.21–2.16)
Amniocentesis, n (%)	13/126 (10.3)	8/45 (17.8)	1.87 (0.72–4.88)
CRP (> 0.3 mg/dL), n (%)	86/110 (78.2)	38/43 (88.4)	2.12 (0.75–5.97)
WBC (> 15,000 /μL), n (%)	50/125 (40.0)	28/44 (63.6)	2.62 (1.28–5.34)*
Use of antibiotics, n (%)	120/128 (93.8)	42/45 (93.3)	0.93 (0.23–3.68)
Pathologic chorioamnionitis, n (%)	73/128 (57.0)	42/45 (93.3)	10.54 (3.11–35.81)*
Cervical cerclage, n (%)	24/126 (19.1)	23/45 (51.1)	4.44 (2.13–9.26)*

Data are expressed as numbers (%) or mean ± SD. OR odds ratio, CI confidence interval, PPROM preterm premature rupture of membranes, CRP C-reactive protein, WBC white blood cell, SD standard deviation. *P value < 0.05.

Table 3 presents the baseline characteristics of the VLBWIs born to colonized mothers. Infants in the transmission group had significantly lower GA, BW, Apgar score at 1 min and 5 min, initial blood pH, and base excess than those in the control group. Additionally, serum CRP levels tested shortly after birth, incidence of surfactant use, number of cases requiring positive pressure ventilation or endotracheal intubation as initial resuscitation, number of cases of hypotension requiring medication, and CRIB II score were significantly higher among infants in the transmission group. To enable a clear comparison of baseline characteristics between the two groups, we adjusted for GA, which is the most influential factor. After adjusting for GA, only base excess and CRP levels differed between the groups. Base excess was significantly lower, and initial serum CRP levels were significantly higher in the transmission group than in the control group.

Table 3. Baseline neonatal characteristics between the control and transmission group.

Characteristics	Control group (n = 128)	Transmission group (n = 45)	OR (95% CI)	Adjusted OR† (95% CI)
GA, weeks	26.9 ± 2.8	24.8 ± 2.4	0.72 (0.62–0.84)*
Birth weight, g	982.7 ± 299.4	788.9 ± 278.7	0.99 (0.98–0.99)*	1.00 (0.99–1.00)
Sex, male, n (%)	62/128 (48.4)	18/45 (40.0)	0.70 (0.35–1.41)	0.60 (0.28–1.26)
Cesarean section, n (%)	97/128 (75.8)	29/45 (64.4)	0.57 (0.27–1.20)	0.67 (0.31–1.48)
Apgar score, 1 min, median [IQR]	6 (4–7)	5 (4–6)	0.72 (0.59–0.88)*	0.87 (0.68–1.10)
Apgar score, 5 min, median [IQR]	8 (7–9)	7 (6–8)	0.72 (0.58–0.91)*	0.88 (0.68–1.13)
Initial pH	7.2 ± 0.1	7.2 ± 0.2	0.02 (0.00–0.35)*	0.08 (0.00–1.25)
Base excesses	− 6.3 ± 3.9	− 9.2 ± 4.9	1.15 (1.06–1.25)*	1.10 (1.01–1.20)*
Initial CRP§, mg/dL	0.1 ± 0.3	0.4 ± 0.8	4.23 (1.54–11.61)*	3.46 (1.31–9.11)*
Initial ANC§, number	6221.9 ± 6881.0	9506.3 ± 10167.8	1.00 (1.00–1.01)	1.00 (0.99–1.01)
Use of surfactant, n (%)	100/128 (78.1)	43/45 (95.6)	6.02 (1.37–26.40)*	2.12 (0.42–10.55)
PPV, n (%)	90/118 (77.6)	42/45 (93.3)	4.04 (1.15–14.11)*	1.99 (0.52–7.61)
Intubation, n (%)	76/118 (65.5)	39/45 (86.7)	3.42 (1.33–8.76)*	1.67 (0.57–4.86)
Hypotension‡, n (%)	19/128 (14.8)	13/45 (28.9)	2.33 (1.03–5.22)*	1.12 (0.45–2.80)
CRIB II score, median [IQR]	9 (6–12)	14 (10–16)	1.21 (1.11–1.32)*	1.11 (0.87–1.42)

Data are expressed as numbers (%), mean ± SD or median [IQR]. OR odds ratio, CI confidence interval, GA gestational weeks, CRP C-reactive protein, ANC absolute neutrophil count, PPV positive pressure ventilation, SD standard deviation, IQR interquartile range. *P value < 0.05. ^†Adjusted for GA. ^§CRP and ANC were tested right after birth. ^‡Hypotension requiring medication within a week.

Table 4 compares the neonatal outcomes according to perinatal bacterial transmission. Perinatal transmission is associated with high mortality rates, grade III or higher IVH, and stage III or higher ROP. After adjusting for GA, we consistently observed significantly higher mortality and grade III or higher IVH rates in the transmission group than in the control group (Table 4). Furthermore, we conducted analyses targeting 35 patients who experienced transmission only, resulting in colonization, and excluded cases of blood culture-proven sepsis. After adjusting for the GA, the incidence of grade III or higher IVH was significantly higher in the transmission group than in the control group (Supplementary Table S3).

Table 4. Neonatal major outcomes between the control and transmission group.

Outcomes	Control group (n = 128)	Transmission group (n = 45)	OR (95% CI)	Adjusted OR† (95% CI)
Mortality, n (%)	14/128 (10.9)	18/45 (40.0)	5.42 (2.40–12.26)*	2.81 (1.11–7.05)*
IVH ≥ grade 3, n (%)	4/125 (3.2)	9/38 (23.7)	9.38 (2.70–32.62)*	6.06 (1.66–22.08)*
NEC ≥ stage 2b, n (%)	14/128 (10.9)	8/45 (17.8)	1.76 (0.68–4.52)	1.06 (0.39–2.84)
BPD ≥ moderate, n (%)	38/118 (32.2)	14/28 (50.0)	2.10 (0.91–4.85)	1.16 (0.45–3.01)
PVL, n (%)	6/125 (4.8)	2/38 (5.3)	1.10 (0.21–5.69)	0.75 (0.13–4.23)
ROP ≥ stage 3, n (%)	16/81 (19.8)	11/21 (52.4)	4.46 (1.61–12.34)*	2.18 (0.68–6.90)

Data are expressed as numbers (%). CI confidence interval, OR odds ratio, IVH intra-ventricular hemorrhage, NEC necrotizing enterocolitis, BPD bronchopulmonary dysplasia, PVL periventricular leukomalacia, ROP retinopathy of prematurity. *P value < 0.05. ^†Adjusted for GA.

Discussion

In this study, we aimed to identify the factors influencing perinatal transmission from colonized mothers to VLBWIs. Primarily, we found that transmission predominantly originates from the maternal vagina, with E. coli emerging as the most frequently transmitted pathogen to neonates. Furthermore, we noted that maternal leukocytosis, pathologically confirmed chorioamnionitis, and a history of cervical cerclage contributed to transmission. In VLBWIs, decreased base excess and increased serum CRP levels were associated with transmission. Regarding neonatal outcomes, transmission correlated with elevated mortality rates and the incidence of grade III or higher IVH. Notably, these outcomes persisted even after excluding patients with culture-proven EOS, and the occurrence of IVH was significantly higher in this group.

To the best of our knowledge, our study represents the first endeavor to explore the risk factors associated with perinatal bacterial transmission in VLBWIs through a comprehensive comparison of maternal and neonatal pathogens.

Although numerous studies have investigated the risk factors associated with EOS^13,14,18, research on the predictive factors for perinatal transmission and subsequent infection development is lacking.

Furthermore, despite the advancements in care practices leading to improved survival rate among very preterm infants and an increased utilization of maternal antibiotics in recent years, no significant change has been observed in the rates of EOS^2,19. This lack of change highlights the vulnerability of VLBWIs to infections caused by colonizing bacteria due to their underdeveloped immune systems. Given that their microflora is predominantly acquired from maternal sources before and during delivery, the transmission of abnormally colonizing pathogens from their mothers poses an increased risk of EOS. Therefore, establishing methods for assessing transmission risk in these infants is crucial.

Previous studies aimed to elucidate the correlation between the maternal vaginal microbiome and that detected in newborns. Chan et al. conducted a global meta-analysis, revealing that maternal infection and colonization amplified the risk of neonatal colonization and infection in the first week after birth. They demonstrated that maternal vaginal colonization correlated with an increased risk of neonatal colonization in various sites such as the ear canal, umbilical, axilla, or anal cultures by Staphylococcus aureus and E. coli⁴. In our study including VLBWIs born to 173 colonized mothers, we observed that 45 (26%) of them acquired the pathogen from their mother, with 10 (5.8%) cases resulting in culture-proven sepsis. However, our investigation was limited to mothers with culture-proven colonization, thus precluding comparisons of colonization or infection rates within the first seven days of life in neonates born to non-colonized mothers.

We have adopted the definition of EOS, which encompasses sepsis occurring within the first 7 days of life. While a significant body of research adheres to a 72 h timeframe for EOS, several reports also recognize the first week as a critical period²⁰. This approach is influenced partly by the practical challenges associated with frequent and accurate blood sampling in VLBWIs. The limited volume of blood available in such cases can lead to diagnostic inaccuracies, necessitating clinical flexibility. Furthermore, we have specifically delineated an EOS subgroup within the context of transmission. This subgroup is characterized by cases in which the antibiotic susceptibility profile of the pathogen in the mother matches that of the pathogens identified in the infant's cultures. This classification underscores the high likelihood of vertical transmission even in cases where pathogen identification occurs after 72 h.

Our study revealed that transmission predominantly occurred via an ascending route through the maternal reproductive tract, with no evidence of transmission through alternative pathways. This suggests that neonates may potentially acquire pathogens either before or after delivery via contact with infected amniotic fluid or the vaginal canal. PPROM is a significant risk factor for transmission as it compromises the natural barrier between the mother and fetus, thereby facilitating pathogen colonization of the vagina, entry into the uterus, and exposure to the fetus. Moreover, invasive procedures performed during pregnancy, such as amniocentesis, increase the risk of amniotic fluid infection. However, we did not observe significant differences in the incidence of PPROM or amniocentesis between the two groups. Interestingly, we identified a significant association between a history of cervical cerclage and an increased risk of transmission. Among mothers with cervical insufficiency, intra-amniotic fluid infection has been reported in 8–52% of cases^21–23, and cerclage in these cases has been associated with an increased risk of clinical chorioamnionitis and PPROM, leading to adverse neonatal outcomes^21,23. Furthermore, another study suggested that some stitches used in cerclage procedures could induce vaginal dysbiosis, which in turn might be correlated with inflammatory cytokines, leading to poor neonatal outcomes²⁴. In this context, our findings provide evidence supporting the association between cervical cerclage and perinatal transmission.

Prenatal screening for infections and the use of intrapartum antibiotic prophylaxis have significantly reduced the transmission of GBS and subsequent sepsis^25,26. However, EOS caused by non-GBS pathogens, especially E. coli, has been increasing among VLBWIs^5,26–29. In this study, E. coli emerged as the most frequently transmitted pathogen, accounting for 16 of 45 cases (35.6%), followed by Candida and Enterococcus species. Remarkably, sepsis in our study was solely caused by E. coli and Candida. Of the 10 EOS cases, E. coli was responsible for eight cases, which accounted for half of the transmission events. Candida species transmission occurred in eight cases, although only two resulted in candidemia. These findings align with recent research indicating that E. coli has surpassed GBS as the predominant bacterium causing EOS. Intriguingly, all 10 cases exhibited pathogens matching those detected in the mothers’ vaginal cultures.

In this study, we focused on pathological chorioamnionitis because of its definitive histopathological characteristics that offer a reliable measure of inflammation relevant to perinatal transmission. The inherent limitations of retrospective chart reviews, particularly the scarce documentation of key clinical indicators such as uterine tenderness or characteristics of amniotic fluid, constrained our ability to comprehensively analyze clinical chorioamnionitis. Consequently, our analysis was confined to pathological manifestations to minimize information bias due to inconsistent clinical data reporting.

In our analysis, we observed that several signs and symptoms associated with clinical chorioamnionitis, such as maternal fever, leukocytosis, tachycardia, and fetal tachycardia, were more prevalent in the transmission group. However, only maternal leukocytosis exhibited a significant association with transmission. This limitation arises from the relatively small sample size from a single center, necessitating further large-scale studies.

Additionally, we observed a significantly higher incidence of pathologically confirmed chorioamnionitis in the transmission group. However, obtaining pathological results takes several days to weeks, rendering it impractical to rely solely on pathological confirmation as an indicator for the immediate use of antibiotics. Acknowledging this constraint, future research should aim for a prospective design that captures both the clinical and pathological aspects of chorioamnionitis.

The incidence rates of chorioamnionitis were 57.0% in the control group and 93.3% in the transmission group. Our institution serves as a tertiary referral hospital renowned for managing high-acuity cases, often receiving expectant mothers transferred in utero due to rapidly deteriorating conditions after initial hospitalization in other tertiary facilities. This frequent occurrence may contribute to the high incidence of chorioamnionitis.

Chorioamnionitis often results from bacterial invasion of the amniotic fluid after prolonged membrane rupture or due to sterile inflammation, and it exhibits an inverse relationship with GA. This condition leads to elevated levels of inflammatory markers in the amniotic fluid, such as interleukin-6 and other pro-inflammatory cytokines, signaling the presence of intra-amniotic inflammation. The maternal inflammatory response induced by chorioamnionitis can trigger a fetal inflammatory response, significantly increasing the risk of adverse neonatal outcomes. Infants exposed to chorioamnionitis are two to three times more likely to suffer from severe complications, including perinatal death, EOS, IVH, cerebral white matter damage, and long-term disabilities such as cerebral palsy, ROP, and NEC. These conditions are among the various morbidities associated with preterm birth^6,30–32.

In the transmission group, higher rates of neonatal mortality and severe IVH (grade III or higher) were the major adverse outcomes. These associations persisted even after excluding the cases of culture-proven sepsis from the analysis. In extremely preterm infants, obtaining an adequate blood sample volume for culture can be challenging, potentially compromising the accuracy of culture results and leading to false-negative results and the inclusion of false-negative culture results in the transmission group. Therefore, caution should be exercised while interpreting the association between transmission and adverse outcomes.

Our study has several limitations. Firstly, being a single-center study with a small cohort, the generalizability of our findings is limited. Additionally, the retrospective nature of the data introduces complexities in interpreting maternal and infant culture results due to various confounding factors. Notably, not all maternal cultures were obtained immediately before delivery; some were performed weeks to months prior, affecting our ability to accurately assess the maternal pathogens’ colonization status at delivery. This heterogeneity in the timing of maternal culture collection poses challenges to accurately evaluating the maternal colonization status during delivery. Moreover, the use of prenatal antibiotics can mask infections or colonization in infants who have acquired pathogens from their mothers. Detection of infections in newborns becomes challenging as antibiotics may prevent bacterial growth in blood cultures, potentially leading to inappropriate antibiotic usage and subsequent escalation of morbidity and mortality. It is also suggested that there might be a relationship between bacterial load and transmission rates; however, due to the retrospective nature of our study, we were unable to obtain data on colony counts, which are indicative of bacterial load. Our inability to access colony count data restricted our exploration of this potential relationship. Furthermore, although we compared antibiotic susceptibility to determine if the bacteria in mothers and infants were identical strains, the absence of genetic sequencing prevents confirmation of their identical nature, thereby introducing another layer of complexity to our analysis. Additionally, several reports suggest that prenatal antibiotic use may disrupt the normal development of the neonatal microbiome, potentially increasing the risk of various infections. However, this aspect was not investigated in this study. Further research is warranted to elucidate the relationship between chorioamnionitis and transmission, and to explore additional variables that could inform appropriate antibiotic use in these cases. Nevertheless, our study underscores the significant impact of perinatal transmission on neonatal outcomes, notably increased mortality and occurrence of severe IVH. These findings emphasize the importance of effectively managing perinatal transmission to mitigate the risk of adverse outcomes in newborns.

In conclusion, our study identifies the predictive factors that may increase the risk of perinatal transmission. These findings may prove valuable in guiding the timely initiation of antibiotic treatment for infants with a history of maternal colonization.

Acknowledgements

This study was supported by the Grants SMC-Ottogi Research Fund (#SMX1210881) and Future Medicine 2030 Project of the Samsung Medical Center (#SMX1240621).

Author contributions

Conceptualization: S.Y.A. and Y.S.C.; Methodology: J.H., S.K.; formal analysis: M.Y. and S.I.S.; data curation: S.I.S. The STATA version 16.0 [StataCorp, College Station, TX, USA]. Validation: C.K. and H.K.; investigation: S.H.K.; writing–original draft preparation: J.H.; writing–review, and editing: S.Y.A. and Y.S.C. All authors have approved the final manuscript for submission.

Data availability

Data supporting the findings of this study are available from the corresponding author ([email protected]) upon request.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-024-67674-7.

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Hahn, WH; Chang, JY; Chang, YS; Shim, KS; Bae, CW. Recent trends in neonatal mortality in very low birth weight Korean infants: In comparison with Japan and the USA. J. Korean Med. Sci.; 2011; 26, pp. 467-473. [DOI: https://dx.doi.org/10.3346/jkms.2011.26.4.467] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21468252][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3069564]

2. Stoll, BJ et al. Trends in care practices, morbidity, and mortality of extremely preterm neonates, 1993–2012. JAMA; 2015; 314, pp. 1039-1051.[COI: 1:CAS:528:DC%2BC28XkslKjsA%3D%3D] [DOI: https://dx.doi.org/10.1001/jama.2015.10244] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26348753][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4787615]

3. Hornik, CP et al. Early and late onset sepsis in very-low-birth-weight infants from a large group of neonatal intensive care units. Early Hum. Dev.; 2012; 88, Suppl 2 pp. S69-S74. [DOI: https://dx.doi.org/10.1016/s0378-3782(12)70019-1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22633519][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3513766]

4. Chan, GJ; Lee, AC; Baqui, AH; Tan, J; Black, RE. Risk of early-onset neonatal infection with maternal infection or colonization: A global systematic review and meta-analysis. PLoS Med.; 2013; 10, e1001502. [DOI: https://dx.doi.org/10.1371/journal.pmed.1001502] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23976885][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3747995]

5. Folgori, L et al. The relationship between Gram-negative colonization and bloodstream infections in neonates: A systematic review and meta-analysis. Clin. Microbiol. Infect.; 2018; 24, pp. 251-257.[COI: 1:STN:280:DC%2BC1cbgs1aktQ%3D%3D] [DOI: https://dx.doi.org/10.1016/j.cmi.2017.08.008] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28830807]

6. Claud, EC; Walker, WA. Bacterial colonization, probiotics, and necrotizing enterocolitis. J. Clin. Gastroenterol.; 2008; 42, Suppl 2 pp. S46-S52. [DOI: https://dx.doi.org/10.1097/MCG.0b013e31815a57a8] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18520617]

7. Sadeghi, K et al. Immaturity of infection control in preterm and term newborns is associated with impaired toll-like receptor signaling. J. Infect. Dis.; 2007; 195, pp. 296-302.[COI: 1:CAS:528:DC%2BD2sXhs1amt7w%3D] [DOI: https://dx.doi.org/10.1086/509892] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17191175]

8. Alexander, VN; Northrup, V; Bizzarro, MJ. Antibiotic exposure in the newborn intensive care unit and the risk of necrotizing enterocolitis. J. Pediatr.; 2011; 159, pp. 392-397. [DOI: https://dx.doi.org/10.1016/j.jpeds.2011.02.035] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21489560][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3137655]

9. Cantey, JB et al. Antibiotic exposure and risk for death or bronchopulmonary dysplasia in very low birth weight infants. J. Pediatr.; 2017; 181, pp. 289-293. [DOI: https://dx.doi.org/10.1016/j.jpeds.2016.11.002] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27908652]

10. Ting, JY et al. Association of antibiotic utilization and neurodevelopmental outcomes among extremely low gestational age neonates without proven sepsis or necrotizing enterocolitis. Am. J. Perinatol.; 2018; 35, pp. 972-978. [DOI: https://dx.doi.org/10.1055/s-0038-1632390] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29475201]

11. Coggins, SA et al. Use of a computerized C-reactive protein (CRP) based sepsis evaluation in very low birth weight (VLBW) infants: A five-year experience. PLoS ONE; 2013; 8, e78602.2013PLoSO..878602C[COI: 1:CAS:528:DC%2BC3sXhslyls7rP] [DOI: https://dx.doi.org/10.1371/journal.pone.0078602] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24244325][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3823853]

12. Steinberger, E; Hofer, N; Resch, B. Cord blood procalcitonin and Interleukin-6 are highly sensitive and specific in the prediction of early-onset sepsis in preterm infants. Scand. J. Clin. Lab. Invest.; 2014; 74, pp. 432-436.[COI: 1:CAS:528:DC%2BC2cXhtl2nt7%2FM] [DOI: https://dx.doi.org/10.3109/00365513.2014.900696] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24738573]

13. Northern California Kaiser-Permanente. Neonatal Early-Onset Sepsis Calculator. 2018. https://neonatalsepsiscalculator.kaiserpermanente.org. Accessed 1 July 2023.

14. Puopolo, K. M., Benitz, W. E., Zaoutis, T. E., Committee on Fetus Newborn & Committee On Infectious Diseases. Management of neonates born at >/=35 0/7 weeks' gestation with suspected or proven early-onset bacterial sepsis. Pediatrics. https://doi.org/10.1542/peds.2018-2894 (2018).

15. Sung, JH et al. A randomized clinical trial of antibiotic treatment duration in preterm pre-labor rupture of membranes: 7 days vs until delivery. Am. J. Obstet. Gynecol. MFM; 2023; 5, 100886.[COI: 1:CAS:528:DC%2BB3sXls1Oltbg%3D] [DOI: https://dx.doi.org/10.1016/j.ajogmf.2023.100886] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36746334]

16. Stoll, BJ et al. Neonatal outcomes of extremely preterm infants from the NICHD Neonatal Research Network. Pediatrics; 2010; 126, pp. 443-456. [DOI: https://dx.doi.org/10.1542/peds.2009-2959] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20732945]

17. Steurer, MA et al. Mortality and major neonatal morbidity in preterm infants with serious congenital heart disease. J. Pediatr.; 2021; 239, pp. 110-116. [DOI: https://dx.doi.org/10.1016/j.jpeds.2021.08.039] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34454949][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10866139]

18. Benitz, WE; Wynn, JL; Polin, RA. Reappraisal of guidelines for management of neonates with suspected early-onset sepsis. J. Pediatr.; 2015; 166, pp. 1070-1074. [DOI: https://dx.doi.org/10.1016/j.jpeds.2014.12.023] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25641240][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4767008]

19. Glaser, MA; Hughes, LM; Jnah, A; Newberry, D. Neonatal sepsis: A review of pathophysiology and current management strategies. Adv. Neonatal Care; 2021; 21, pp. 49-60. [DOI: https://dx.doi.org/10.1097/anc.0000000000000769] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32956076]

20. Briggs-Steinberg, C; Roth, P. Early-onset sepsis in newborns. Pediatr. Rev.; 2023; 44, pp. 14-22. [DOI: https://dx.doi.org/10.1542/pir.2020-001164] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36587021]

21. Romero, R et al. Infection and labor. VIII. Microbial invasion of the amniotic cavity in patients with suspected cervical incompetence: Prevalence and clinical significance. Am. J. Obstet. Gynecol.; 1992; 167, pp. 1086-1091.[COI: 1:STN:280:DyaK3s%2FjsFGntg%3D%3D] [DOI: https://dx.doi.org/10.1016/s0002-9378(12)80043-3] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/1415396]

22. Lee, SE; Romero, R; Park, CW; Jun, JK; Yoon, BH. The frequency and significance of intraamniotic inflammation in patients with cervical insufficiency. Am. J. Obstet. Gynecol.; 2008; 198, 633 pp. e631-638. [DOI: https://dx.doi.org/10.1016/j.ajog.2007.11.047]

23. Oh, KJ et al. Evidence that antibiotic administration is effective in the treatment of a subset of patients with intra-amniotic infection/inflammation presenting with cervical insufficiency. Am. J. Obstet. Gynecol.; 2019; 221, pp. 140.e141-140.e118.[COI: 1:CAS:528:DC%2BC1MXhtFGjt7vM] [DOI: https://dx.doi.org/10.1016/j.ajog.2019.03.017]

24. Kindinger, LM et al. Relationship between vaginal microbial dysbiosis, inflammation, and pregnancy outcomes in cervical cerclage. Sci. Transl. Med.; 2016; 8, 350ra102. [DOI: https://dx.doi.org/10.1126/scitranslmed.aag1026] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27488896]

25. Bauserman, MS et al. Group B Streptococcus and Escherichia coli infections in the intensive care nursery in the era of intrapartum antibiotic prophylaxis. Pediatr. Infect. Dis. J.; 2013; 32, pp. 208-212. [DOI: https://dx.doi.org/10.1097/INF.0b013e318275058a] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23011013][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3572304]

26. Schrag, SJ et al. Epidemiology of invasive early-onset neonatal sepsis, 2005 to 2014. Pediatrics; 2016; 138, e20162013. [DOI: https://dx.doi.org/10.1542/peds.2016-2013] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27940705]

27. Mukhopadhyay, S; Puopolo, KM. Clinical and microbiologic characteristics of early-onset sepsis among very low birth weight infants: Opportunities for antibiotic stewardship. Pediatr. Infect. Dis. J.; 2017; 36, pp. 477-481. [DOI: https://dx.doi.org/10.1097/inf.0000000000001473] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28403049][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6009981]

28. Stoll, BJ et al. Early-onset neonatal sepsis 2015 to 2017, the rise of Escherichia coli, and the need for novel prevention strategies. JAMA Pediatr; 2020; 174, e200593. [DOI: https://dx.doi.org/10.1001/jamapediatrics.2020.0593] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32364598][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7199167]

29. Miselli, F et al. Escherichia coli is overtaking Group B Streptococcus in early-onset neonatal sepsis. Microorganisms; 2022; 10, 1878.[COI: 1:CAS:528:DC%2BB38XivVGrsb%2FE] [DOI: https://dx.doi.org/10.3390/microorganisms10101878] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36296155][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9607315]

30. Pugni, L et al. Chorioamnionitis and neonatal outcome in preterm infants: A clinical overview. J. Matern. Fetal Neonatal Med.; 2016; 29, pp. 1525-1529. [DOI: https://dx.doi.org/10.3109/14767058.2015.1053862] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26135227]

31. Shane, AL; Sánchez, PJ; Stoll, BJ. Neonatal sepsis. Lancet; 2017; 390, pp. 1770-1780. [DOI: https://dx.doi.org/10.1016/s0140-6736(17)31002-4] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28434651]

32. Bastek, JA; Weber, AL; McShea, MA; Ryan, ME; Elovitz, MA. Prenatal inflammation is associated with adverse neonatal outcomes. Am. J. Obstet. Gynecol.; 2014; 210, 450.e451-410. [DOI: https://dx.doi.org/10.1016/j.ajog.2013.12.024]