BACKGROUND:

Premature infants are at high risk of early-onset sepsis (EOS) relative to term infants, and most are administered empirical antibiotics after birth. We aimed to determine if factors evident at birth could be used to identify premature infants at lower risk of EOS.

METHODS:

Study infants were born at 22 to 28 weeks’ gestation in Neonatal Research Network centers from 2006 to 2014. EOS was defined by isolation of pathogenic species from blood or cerebrospinal fluid culture at ≤72 hours age. Infants were hypothesized as “low risk” for EOS when delivered via cesarean delivery, with membrane rupture at delivery, and absence of clinical chorioamnionitis. Frequency of prolonged antibiotics (≥5 days) was compared between low-risk infants and all others. Risks of mortality, EOS, and other morbidities were assessed by using regression models adjusted for center, race, antenatal steroid use, multiple birth, sex, gestation, and birth weight.

RESULTS:

Of 15 433 infants, 5759 (37%) met low-risk criteria. EOS incidence among infants surviving >12 hours was 29 out of 5640 (0.5%) in the low-risk group versus 209 out of 8422 (2.5%) in the comparison group (adjusted relative risk = 0.24 [95% confidence interval, 0.16–0.36]). Low-risk infants also had significantly lower combined risk of EOS or death ≤12 hours. Prolonged antibiotics were administered to 34% of low-risk infants versus 47% of comparison infants without EOS.

CONCLUSIONS:

Delivery characteristics of extremely preterm infants can be used to identify those with significantly lower incidence of EOS. Recognition of differential risk may help guide decisions to limit early antibiotic use among approximately one-third of these infants.

What’s Known on This Subject:

Characteristics known at birth including gestational age identify infants at highest risk of early-onset sepsis (EOS). It is unknown if delivery characteristics can be used to identify premature infants at lowest risk of EOS to guide empirical antibiotic therapy.

What This Study Adds:

Specific criteria can identify premature infants with a 76% lower adjusted risk for EOS. A substantial proportion of these infants received prolonged antibiotics despite their lower a priori risk. Recognition of this differential risk may reduce early antibiotic exposures.

Accurately estimating individual risk of early-onset sepsis (EOS) among very premature infants is challenging because of the relatively high incidence of clinical instability among these infants and the associated concern for sepsis as a treatable cause of such instability.1,2 Antibiotics are initiated in the majority of extremely preterm infants and are frequently continued despite sterile blood cultures.3,5 Of concern, prolonged early antibiotic exposure has been associated with increased subsequent risk of late-onset sepsis (LOS), necrotizing enterocolitis (NEC), severe retinopathy of prematurity, and death.3,6,9 Identification of premature infants at low risk of EOS may help guide decisions for initiating and/or discontinuing empirical antibiotic treatments in the first days of life and would be a critical first step in promoting antibiotic stewardship among these infants.

The pathogenesis of EOS is predominantly that of ascending colonization of the uterine and fetal compartments with maternal recto-vaginal flora, progressing to inflammation and infection.10,11 Preterm onset of labor and preterm rupture of amniotic membranes are both significantly associated with risk of EOS.12,13 However, a proportion of preterm deliveries occur because of maternal medical indications (such as preeclampsia) or for chronic fetal conditions (such as growth restriction).14 In the absence of labor or rupture of membranes (ROM), the risk of EOS in preterm infants delivered via cesarean delivery (CD) should be substantially lower compared with preterm infants delivered because of preterm labor or premature ROM. We accessed data collected by the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network (NRN) and used delivery characteristics to define a group of extremely preterm infants we hypothesized would be at lower risk for EOS. Our objectives were to compare the incidence of culture-confirmed EOS and the prevalence of prolonged early antibiotic treatment among infants with and without low EOS risk criteria.

Infants were born at NRN centers between January 1, 2006, and December 31, 2014, and enrolled in a registry maintained by the NRN. Inclusion criteria for the cohort studied were: gestational age (GA) 22 weeks, 0 days to 28 weeks, 6 days; birth weight (BW) 401 to 1500 g; inborn; and no major congenital anomaly. Registry data included prospectively collected maternal pregnancy and delivery information; infant data from birth until death, hospital discharge or transfer, or 120 days of age; and if hospitalized at 120 days, final discharge or transfer date or death up to 1 year of age.15 Participation in the registry was approved at most sites by the local institutional review board under a waiver of consent; 3 sites required parental consent.

Preterm labor was not recorded in the registry during the study period. Delivery characteristics likely to capture infants delivered for maternal health indications without preterm onset of labor were used to define a hypothesized low-risk group for EOS. Low-risk infants had all of the following characteristics: delivery by CD, ROM at delivery, and no documentation of maternal clinical chorioamnionitis in the obstetrical record. Infants without low-risk criteria were denoted as the “comparison” group for description of study results. In a secondary analysis, the low-risk group was further defined to also exclude infants of mothers with histologic diagnosis of chorioamnionitis on the basis of placental pathologic examination.

Hospital morbidities and clinical interventions were recorded for infants who survived >12 hours. EOS (≤72 hours of age) and LOS (>72 hours of age) were defined by blood or cerebrospinal fluid (CSF) cultures positive for pathogenic bacteria or fungi and antibiotic therapy ≥5 days or intent to treat and death within 5 days. Micrococcus, Propionibacterium, Corynebacterium, Bacillus, and coagulase-negative staphylococci (CONS) grown on cultures of blood or CSF ≤72 hours of age were considered contaminants for the primary outcome of EOS incidence. Cultures growing ≥1 species of which at least 1 was considered a true pathogen were counted as EOS cases. For a secondary analysis, infants with CONS were included as EOS cases. CONS grown from LOS cultures were included as true pathogens. Other morbidities included respiratory distress syndrome defined by clinical features in the first 24 hours, pneumothorax, pulmonary hemorrhage, pulmonary interstitial emphysema, patent ductus arteriosus, NEC stage 2 to 3,16,17 spontaneous intestinal perforation (SIP), severe intracranial hemorrhage (ICH) grade 3 or 418 based on a cranial sonogram taken within 28 days of birth, periventricular leukomalacia (PVL), and bronchopulmonary dysplasia (BPD) defined as need for supplemental oxygen at 36 weeks postmenstrual age (PMA). Surviving infants discharged or transferred before 36 weeks PMA were classified on the basis of their status at 36 weeks, if known, or on the basis of oxygen use at discharge or transfer. Small for gestational age (SGA) was defined as BW <10th percentile for sex and GA.19 

Prolonged early antibiotic therapy was defined as receipt of antibiotics for ≥5 days starting <72 hours of age (or intent to treat for ≥5 days in an infant who died within 5 days and was receiving antibiotics). Those who did not receive prolonged antibiotics were either not started on antibiotics <72 hours, or had antibiotics initiated but discontinued before 5 days. Other interventions recorded included receipt of surfactant, nitric oxide, treatment of hypotension, and duration and type of respiratory support.

Characteristics and outcomes were compared between infants in the low-risk and comparison groups. Additionally, we compared outcomes for low-risk infants with and without early prolonged antibiotics, excluding infants with any of the following at <7 days of age: positive blood or CSF culture results (including isolates considered contaminants in this study) and/or a diagnosis of NEC or SIP. Statistical significance for unadjusted comparisons between groups was determined by χ2 test for categorical variables and Wilcoxon test for continuous variables. Poisson regression models with robust variance estimators20 were used to assess risk of mortality, the primary outcome EOS, and other binary outcomes in comparison groups while adjusting for study center, maternal race and/or ethnicity, antenatal steroid use, multiple birth, sex, GA (categorical by week), and BW (continuous). Adjusted relative risks (aRRs), 95% confidence intervals (CIs), and P values by the score or Wald χ2 test from these models were reported. Morbidity outcomes were compared between low-risk and comparison infants surviving >12 hours, accounting for the age at which assessment took place. Combined death or morbidity outcomes were also examined to account for the competing risk of death before evaluation. For analyses of low-risk infants with and without early prolonged antibiotics, morbidities were examined among infants who survived >7 days to allow for exposure to prolonged antibiotics to have occurred. For these analyses, combined death or morbidity outcomes included all deaths after 7 days and before discharge to examine risk of death in addition to morbidity before discharge. A P value < .05 was considered significant; no adjustment was made for multiple comparisons. Analyses were performed by using SAS version 9.4 (SAS Institute, Cary, NC).

Between January 1, 2006, and December 31, 2014, 16 185 infants with GA 22 to 28 weeks and BW 401 to 1500 g were born at NRN centers. Of these, 640 (4%) infants with a major congenital anomaly, 111 (0.7%) infants with insufficient information (maternal clinical chorioamnionitis, 54; ROM, 43; EOS or antibiotic information, 10; delivery type, 4), and 1 infant with congenital syphilis were excluded. Thus, 15 433 infants were studied, of whom 5759 (37.3%) were classified into the low-risk group and 9674 (62.7%) into the comparison group.

By definition, all infants in the low-risk group were delivered by CD, without ROM before delivery, and none had maternal clinical chorioamnionitis (Table 1). Maternal histologic chorioamnionitis was reported for a smaller proportion of infants in the low-risk group versus the comparison group (21% vs 66%). Mothers in the low-risk group were more likely to suffer complications such as hypertension, more likely to receive antenatal corticosteroids, and less likely to receive antenatal antibiotics. Median gestation was higher in the low-risk group, as was the proportion of female infants, multiple births, and SGA infants. In all measures except 1 (nitric oxide therapy), a significantly greater proportion of infants in the low-risk group had markers of illness and clinical intervention (Table 1).

TABLE 1

Maternal, Delivery, and Neonatal Characteristics

N (Column %) or as ShownaLow Risk of EOS (N = 5759)Comparison Group (N = 9674)Pb
Study Criteria for Infants With Low Risk of EOS 
 CD 5759 (100.0) 3899 (40.3) ≤.001 
 ROM before delivery 0 (0.0) 7330 (76.2) <.001 
 Maternal clinical chorioamnionitis 0 (0.0) 2538 (26.2) <.001 
 Placental pathology performed 4898 (85.3) 8258 (85.6) .58 
 Histologic chorioamnionitis reported 1042/4884 (21.3) 5457/8221 (66.4) <.001 
Maternal and delivery characteristics 
 Maternal age in y, median (IQR) 28 (23–32) 27 (22–32) <.001 
 Gravida   .003 
  1 1876 (32.6) 3001 (31.0)  
  2 1378 (23.9) 2279 (23.6)  
  3 992 (17.2) 1589 (16.4)  
  4+ 1512 (26.3) 2802 (29.0)  
 Maternal race and/or ethnicityc   <.001 
  African American, non-Hispanic 2141 (37.3) 3846 (39.9)  
  White, non-Hispanic 2415 (42.1) 3648 (37.9)  
  Hispanic 878 (15.3) 1620 (16.8)  
  Other 304 (5.3) 523 (5.4)  
 At least 1 prenatal visit 5536 (96.2) 9141 (94.5) <.001 
 Maternal insulin-dependent diabetes 332 (5.8) 428 (4.4) <.001 
 Maternal hypertension 2677 (46.5) 1204 (12.4) <.001 
 Antepartum hemorrhage 917 (15.9) 2038 (21.1) <.001 
 Maternal antibiotics during delivery admission 3105 (54.1) 7767 (80.6) <.001 
 Antenatal steroids 5062 (88.1) 7995 (82.8) <.001 
 Magnesium sulfate during delivery admissiond 1858/2473 (75.1) 2977/4124 (72.2) .009 
 Multiple birth 1928 (33.5) 2175 (22.5) <.001 
Infant characteristics 
 GA, weeks, median (IQR) 27 (25–28) 26 (24–27) <.001 
 By GA week   <.001 
  22 16 (0.3) 601 (6.2)  
  23 181 (3.1) 1207 (12.5)  
  24 662 (11.5) 1445 (14.9)  
  25 916 (15.9) 1488 (15.4)  
  26 1089 (18.9) 1504 (15.5)  
  27 1353 (23.5) 1661 (17.2)  
  28 1542 (26.8) 1768 (18.3)  
 BW, g, median (IQR) 820 (660–1000) 820 (640–1028) .64 
 SGA 844 (14.7) 299 (3.1) <.001 
 Male 2836 (49.3) 5171 (53.5) <.001 
 Apgar <5 at 5 min 1030 (17.9) 2487 (26.0) <.001 
Infants surviving >12 h 5640 8422 — 
 Surfactant 4883 (86.6) 6549 (77.8) <.001 
 Nitric oxide 371 (6.6) 777 (9.2) <.001 
 Median (IQR) highest base deficit in first 24 hb 6 (4–9) 5 (3–8) <.001 
 Treated for hypotension in first 24 he 684/2429 (28.2) 948/3683 (25.7) .04 
 Median (IQR) DOL first enteral feed 4 (3–6) 4 (2–5) <.001 
Infants who survived >24 h 5608 8321 — 
 Respiratory support at 24 hf 5297 (94.7) 7606 (91.6) <.001 
 Mechanical ventilation at 24 h 3562 (63.7) 4773 (57.5) <.001 
Infants who survived >3 d 5499 8131 — 
 Any respiratory support day 1–3f 5440 (99.0) 7932 (97.6) <.001 
 Any mechanical ventilation day 1–3 4698 (85.5) 6371 (78.4) <.001 
N (Column %) or as ShownaLow Risk of EOS (N = 5759)Comparison Group (N = 9674)Pb
Study Criteria for Infants With Low Risk of EOS 
 CD 5759 (100.0) 3899 (40.3) ≤.001 
 ROM before delivery 0 (0.0) 7330 (76.2) <.001 
 Maternal clinical chorioamnionitis 0 (0.0) 2538 (26.2) <.001 
 Placental pathology performed 4898 (85.3) 8258 (85.6) .58 
 Histologic chorioamnionitis reported 1042/4884 (21.3) 5457/8221 (66.4) <.001 
Maternal and delivery characteristics 
 Maternal age in y, median (IQR) 28 (23–32) 27 (22–32) <.001 
 Gravida   .003 
  1 1876 (32.6) 3001 (31.0)  
  2 1378 (23.9) 2279 (23.6)  
  3 992 (17.2) 1589 (16.4)  
  4+ 1512 (26.3) 2802 (29.0)  
 Maternal race and/or ethnicityc   <.001 
  African American, non-Hispanic 2141 (37.3) 3846 (39.9)  
  White, non-Hispanic 2415 (42.1) 3648 (37.9)  
  Hispanic 878 (15.3) 1620 (16.8)  
  Other 304 (5.3) 523 (5.4)  
 At least 1 prenatal visit 5536 (96.2) 9141 (94.5) <.001 
 Maternal insulin-dependent diabetes 332 (5.8) 428 (4.4) <.001 
 Maternal hypertension 2677 (46.5) 1204 (12.4) <.001 
 Antepartum hemorrhage 917 (15.9) 2038 (21.1) <.001 
 Maternal antibiotics during delivery admission 3105 (54.1) 7767 (80.6) <.001 
 Antenatal steroids 5062 (88.1) 7995 (82.8) <.001 
 Magnesium sulfate during delivery admissiond 1858/2473 (75.1) 2977/4124 (72.2) .009 
 Multiple birth 1928 (33.5) 2175 (22.5) <.001 
Infant characteristics 
 GA, weeks, median (IQR) 27 (25–28) 26 (24–27) <.001 
 By GA week   <.001 
  22 16 (0.3) 601 (6.2)  
  23 181 (3.1) 1207 (12.5)  
  24 662 (11.5) 1445 (14.9)  
  25 916 (15.9) 1488 (15.4)  
  26 1089 (18.9) 1504 (15.5)  
  27 1353 (23.5) 1661 (17.2)  
  28 1542 (26.8) 1768 (18.3)  
 BW, g, median (IQR) 820 (660–1000) 820 (640–1028) .64 
 SGA 844 (14.7) 299 (3.1) <.001 
 Male 2836 (49.3) 5171 (53.5) <.001 
 Apgar <5 at 5 min 1030 (17.9) 2487 (26.0) <.001 
Infants surviving >12 h 5640 8422 — 
 Surfactant 4883 (86.6) 6549 (77.8) <.001 
 Nitric oxide 371 (6.6) 777 (9.2) <.001 
 Median (IQR) highest base deficit in first 24 hb 6 (4–9) 5 (3–8) <.001 
 Treated for hypotension in first 24 he 684/2429 (28.2) 948/3683 (25.7) .04 
 Median (IQR) DOL first enteral feed 4 (3–6) 4 (2–5) <.001 
Infants who survived >24 h 5608 8321 — 
 Respiratory support at 24 hf 5297 (94.7) 7606 (91.6) <.001 
 Mechanical ventilation at 24 h 3562 (63.7) 4773 (57.5) <.001 
Infants who survived >3 d 5499 8131 — 
 Any respiratory support day 1–3f 5440 (99.0) 7932 (97.6) <.001 
 Any mechanical ventilation day 1–3 4698 (85.5) 6371 (78.4) <.001 

DOL, day of life; IQR, interquartile range; —, not applicable.

a

Information was missing as follows: CD, 3; ROM before delivery, 60; placental pathology performed, 49; histologic chorioamnionitis, 51; maternal age, 1 infant; gravida, 4; maternal race and/or ethnicity, 58; prenatal care, 7; maternal insulin dependent diabetes, 6; maternal hypertension, 7; antepartum hemorrhage, 4; maternal antibiotics, 54; antenatal steroids, 29; SGA, 4; infant sex, 4; Apgar score, 116; first temperature within 60 min, 125; temperature if within 60 min, 10; surfactant, 1 infant; nitric oxide, 17 infants; timing of enteral feeds, 3 infants; respiratory support at 24 h, 30 infants; respiratory support day 1–3, 12 infants.

b

P value by the χ2 test (categorical variables) or the Wilcoxon test (continuous variables).

c

Maternal white or African American race with missing ethnicity information (4.6% of African Americans, 0.8% of whites) were classified as non-Hispanic. Other races included Asian American and/or Pacific Islander, American Indian and/or Alaskan native, >1 race, other not specified, with non-Hispanic ethnicity.

d

Maternal magnesium sulfate use was collected beginning April 1, 2011.

e

Infant hypotension treatment and highest base deficit were collected beginning April 1, 2011. In this group, hypotension treatment was missing for 2 infants and highest base deficit was missing for 263 infants.

f

Respiratory support was defined as any one of mechanical ventilation (high frequency or conventional), nasal synchronized intermittent mandatory ventilation, or continuous positive airway pressure received at 24 h or during any of the first 3 d of life.

The proportion of infants who died in the first 12 hours after birth was smaller in the low-risk group than in the comparison group (Table 2). Risk of the composite outcome of death within 12 hours or EOS was also reduced for infants in the low-risk group (2.6% vs 15.1%; aRR [95% CI]: 0.36 [0.30–0.43]). Among infants surviving >12 hours, 29 out of 5640 (0.5%) in the low-risk group and 209 out of 8422 (2.5%) in the comparison group had EOS (aRR [95% CI]: 0.24 [0.16–0.36]). The risk of EOS in the low-risk group was consistently lower than in the comparison group for infants born at each GA, as well as when CONS cases were included. Escherichia coli was the most frequent pathogen isolated in both the low-risk group (9 out of 29 cases) and the comparison group (103 out of 209 cases) (Supplemental Tables 8 and 9). The predictive performance of risk categorization by study criteria is shown in Supplemental Table 10. The likelihood of not having EOS was 3 times greater for low-risk infants than for infants in the comparison group.

TABLE 2

Mortality and EOS

N (Column %)Low Risk of EOSComparison GroupAdjusted RR for Outcome (95% CI): Low Risk of EOS Versus the Comparison Groupa
All infants 5759 9674 — 
 Died before discharge 938 (16.3) 2629 (27.2) 0.88 (0.82–0.94) 
 Died ≤12 h 119 (2.1) 1252 (12.9) 0.42 (0.35–0.52) 
 EOS or death ≤12 hb 148 (2.6) 1461 (15.1) 0.36 (0.30–0.43) 
 By GA, wk    
  22 7/16 (43.8) 497/601 (82.7) 0.61 (0.35–1.05) 
  23 31/181 (17.1) 480/1207 (39.8) 0.50 (0.36–0.68) 
  24 44/662 (6.6) 200/1445 (13.8) 0.46 (0.34–0.63) 
  25 21/916 (2.3) 107/1488 (7.2) 0.28 (0.18–0.45) 
  26 19/1089 (1.7) 73/1504 (4.9) 0.29 (0.18–0.48) 
  27 16/1353 (1.2) 50/1661 (3.0) 0.31 (0.17–0.53) 
  28 10/1542 (0.6) 54/1768 (3.1) 0.17 (0.08–0.33) 
 EOS (incl. CONS) or death ≤12 h 165 (2.9) 1484 (15.3) 0.39 (0.33–0.46) 
Infants who survived >12 h 5640 8422 — 
 EOSc 29 (0.5) 209 (2.5) 0.24 (0.16–0.36) 
 EOS (incl. CONS) 46 (0.8) 232 (2.8) 0.33 (0.24–0.46) 
N (Column %)Low Risk of EOSComparison GroupAdjusted RR for Outcome (95% CI): Low Risk of EOS Versus the Comparison Groupa
All infants 5759 9674 — 
 Died before discharge 938 (16.3) 2629 (27.2) 0.88 (0.82–0.94) 
 Died ≤12 h 119 (2.1) 1252 (12.9) 0.42 (0.35–0.52) 
 EOS or death ≤12 hb 148 (2.6) 1461 (15.1) 0.36 (0.30–0.43) 
 By GA, wk    
  22 7/16 (43.8) 497/601 (82.7) 0.61 (0.35–1.05) 
  23 31/181 (17.1) 480/1207 (39.8) 0.50 (0.36–0.68) 
  24 44/662 (6.6) 200/1445 (13.8) 0.46 (0.34–0.63) 
  25 21/916 (2.3) 107/1488 (7.2) 0.28 (0.18–0.45) 
  26 19/1089 (1.7) 73/1504 (4.9) 0.29 (0.18–0.48) 
  27 16/1353 (1.2) 50/1661 (3.0) 0.31 (0.17–0.53) 
  28 10/1542 (0.6) 54/1768 (3.1) 0.17 (0.08–0.33) 
 EOS (incl. CONS) or death ≤12 h 165 (2.9) 1484 (15.3) 0.39 (0.33–0.46) 
Infants who survived >12 h 5640 8422 — 
 EOSc 29 (0.5) 209 (2.5) 0.24 (0.16–0.36) 
 EOS (incl. CONS) 46 (0.8) 232 (2.8) 0.33 (0.24–0.46) 

RR, relative risk; —, not applicable.

a

RRs of each outcome for infants in the low risk of EOS group versus the comparison group were adjusted for center, maternal race and/or ethnicity, antenatal steroids, multiple birth, GA, BW, and sex.

b

The RRs for EOS or death within 12 h of delivery varied by GA (group GA interaction, P = .015).

c

The RRs for EOS did not vary significantly by GA (group GA interaction, P = .94). In the model assessing the interaction, GA 22 and 23 wk were combined.

Placental pathology was available for 4802 out of 5622 (85.4%) low-risk infants and 7202 out of 8394 (85.7%) comparison infants surviving >12 hours (excluding those with no information). The proportion of infants with EOS was similar for those with and without placental pathology (with versus without pathology, low-risk group: 0.5% vs 0.4%; comparison group: 2.5% vs 2.5%; P = .86 for overall difference). Among low-risk infants surviving >12 hours and born to mothers without a histologic diagnosis of chorioamnionitis on placental pathology, 10 out of 3771 (0.3%) infants had EOS (Table 3).

TABLE 3

Distribution of EOS Cases Among 5640 Infants in the Low Risk of EOS Group and 8422 Infants in the Comparison Group Who Survived >12 Hours Based on Maternal Clinical and Histopathology Diagnosis of Chorioamnionitis

Clinical ChorioamnionitisNo Clinical Chorioamnionitis
HCaNo HCNo PathologyHCNo HCNo Pathology
Low risk, N 1018 3771 851 
 EOS cases, N (%) 16 (1.6) 10 (0.3) 
Comparison, N 1632 289 252 3081 2166 1002 
 EOS cases, N (%) 83 (5.1) 4 (1.4) 15 70 (2.3) 20 (0.9) 17 
Clinical ChorioamnionitisNo Clinical Chorioamnionitis
HCaNo HCNo PathologyHCNo HCNo Pathology
Low risk, N 1018 3771 851 
 EOS cases, N (%) 16 (1.6) 10 (0.3) 
Comparison, N 1632 289 252 3081 2166 1002 
 EOS cases, N (%) 83 (5.1) 4 (1.4) 15 70 (2.3) 20 (0.9) 17 

HC, histological chorioamnionitis.

a

In the low-risk group, placental pathology was not performed for 820 infants; pathology was performed but results were missing for 13 infants; and information about whether pathology was done was missing for 18 infants. In the comparison group, placental pathology was not performed for 1192 infants; pathology was performed but results were missing for 34 infants; and information about whether pathology was done was missing for 28 infants.

Among infants who survived >12 hours, the adjusted risks of respiratory distress syndrome, pulmonary hemorrhage, and patent ductus arteriosus were higher in the low-risk group compared with the comparison group (Table 4). The composite risks of death within 12 hours and each of pneumothorax, pulmonary interstitial emphysema, NEC, and SIP, were lower in the low-risk group, as was the risk of LOS or death within 3 days, and the risk of intraventricular hemorrhage and/or PVL or death within 28 days without evaluation (data not shown).

TABLE 4

Morbidities Among Infants Surviving >12 Hours

N (Column %)aLow Risk of EOSComparison GroupAdjusted RR for Outcome (95% CI): Low Risk of EOS Versus Comparison Groupb
Infants who survived >12 h 5640 8422 — 
 RDS 5575 (98.8) 8219 (97.6) 1.01 (1.00–1.01) 
 Pneumothorax 319 (5.7) 483 (5.7) 1.00 (0.87–1.16) 
 Pulmonary hemorrhage 418 (7.4) 445 (5.3) 1.50 (1.31–1.72) 
 PIEc 191/2430 (7.9) 352/3684 (9.6) 0.97 (0.81–1.15) 
 PDA 2648 (47.0) 3612 (42.9) 1.15 (1.11–1.20) 
 NEC in the first 7 d 54 (1.0) 65 (0.8) 1.12 (0.75–1.68) 
 NEC before discharge 561 (9.9) 864 (10.3) 0.97 (0.87–1.08) 
 SIP in the first 7 d 93 (1.6) 139 (1.7) 1.06 (0.81–1.40) 
 SIP before discharge 195 (3.5) 321 (3.8) 1.04 (0.87–1.25) 
Infants who survived >3 d 5499 8131 — 
 LOS on DOL 4–7 135 (2.5) 270 (3.3) 0.95 (0.76–1.19) 
 LOS before discharge 1363 (24.8) 2130 (26.2) 1.00 (0.94–1.06) 
Infants evaluated with cranial imaging 5517 8233 — 
 Severe ICH or PVL 784/5502 (14.2) 1456/8199 (17.8) 0.97 (0.89–1.05) 
Infants who survived to 36 wk PMA 4938 7190 — 
 BPD 2286/4900 (46.7) 3115/7128 (43.7) 1.02 (0.98–1.06) 
N (Column %)aLow Risk of EOSComparison GroupAdjusted RR for Outcome (95% CI): Low Risk of EOS Versus Comparison Groupb
Infants who survived >12 h 5640 8422 — 
 RDS 5575 (98.8) 8219 (97.6) 1.01 (1.00–1.01) 
 Pneumothorax 319 (5.7) 483 (5.7) 1.00 (0.87–1.16) 
 Pulmonary hemorrhage 418 (7.4) 445 (5.3) 1.50 (1.31–1.72) 
 PIEc 191/2430 (7.9) 352/3684 (9.6) 0.97 (0.81–1.15) 
 PDA 2648 (47.0) 3612 (42.9) 1.15 (1.11–1.20) 
 NEC in the first 7 d 54 (1.0) 65 (0.8) 1.12 (0.75–1.68) 
 NEC before discharge 561 (9.9) 864 (10.3) 0.97 (0.87–1.08) 
 SIP in the first 7 d 93 (1.6) 139 (1.7) 1.06 (0.81–1.40) 
 SIP before discharge 195 (3.5) 321 (3.8) 1.04 (0.87–1.25) 
Infants who survived >3 d 5499 8131 — 
 LOS on DOL 4–7 135 (2.5) 270 (3.3) 0.95 (0.76–1.19) 
 LOS before discharge 1363 (24.8) 2130 (26.2) 1.00 (0.94–1.06) 
Infants evaluated with cranial imaging 5517 8233 — 
 Severe ICH or PVL 784/5502 (14.2) 1456/8199 (17.8) 0.97 (0.89–1.05) 
Infants who survived to 36 wk PMA 4938 7190 — 
 BPD 2286/4900 (46.7) 3115/7128 (43.7) 1.02 (0.98–1.06) 

DOL, day of life; ICH, intracranial hemorrhage; PDA, patent ductus arteriosus; PIE, pulmonary interstitial emphysema; RDS, respiratory distress syndrome; RR, relative risk; —, not applicable.

a

Information was missing as follows: RDS, 1 infant; pulmonary hemorrhage, 1 infant; PDA, 11 infants; NEC in the first 7 d, 4 infants; NEC before discharge, 2 infants; SIP in the first 7 d, 12 infants; SIP before discharge, 6 infants; LOS on days 4–7, 13 infants; LOS before discharge, 5 infants; ICH and/or PVL, 49 infants; BPD, 100 infants.

b

RRs of each outcome for infants in the low risk of EOS group versus the comparison group were adjusted for center, maternal race and/or ethnicity, antenatal steroids, multiple birth, GA (categorical), BW (continuous), and sex, except as noted below. Because of the small number of infants with the outcome, center was not included in the model assessing risk of NEC in the first 7 d of life. Only 1 infant born at GA 22 wk had NEC or SIP in the first 7 d of life; GAs 22 and 23 wk were combined for the purpose of assessing these outcomes. All infants born at GA 22 wk had RDS; therefore, GA 22 and 23 wk were combined for the purpose of assessing the outcome RDS.

c

PIE was collected beginning April 1, 2011.

Among infants surviving >12 hours without EOS, 34.1% of infants in the low-risk group received prolonged early antibiotics compared with 47.4% in the comparison group, P < .001 (Table 5). This difference persisted when infants with standard indications for antibiotic use (positive culture results, NEC, and/or SIP ≤7 days of life) were excluded. For each EOS case, 66 low-risk infants received prolonged early antibiotics compared with 19 infants in the comparison group, P < .001. Among low-risk and comparison infants, the proportion of infants who received prolonged early antibiotics varied by center (Fig 1). In comparison with the incidence of EOS cases, the proportion of infants treated with prolonged antibiotics was many times higher at all centers in both groups.

TABLE 5

Antibiotic Use in Low-Risk and Comparison Infants Surviving >12 Hours

N (Column %) or as ShownLow Risk of EOS, N = 5640Comparison Group, N = 8422Pa
Antibiotics for ≥5 d starting within 72 h 1940 (34.4) 4106 (48.8) <.001 
 Antibiotics in the absence of EOS 1911/5611 (34.1) 3897/8213 (47.4) <.001 
 Antibiotics in the absence of positive EOS culture (cases and contaminants)b 1890/5590 (33.8) 3862/8177 (47.2) <.001 
 Antibiotics in the absence of a positive blood or CSF culture result, NEC, or SIP ≤7 dc 1771/5334 (33.2) 3649/7752 (47.1) <.001 
 No. infants given prolonged early antibiotics per EOS case 66 19 <.001 
N (Column %) or as ShownLow Risk of EOS, N = 5640Comparison Group, N = 8422Pa
Antibiotics for ≥5 d starting within 72 h 1940 (34.4) 4106 (48.8) <.001 
 Antibiotics in the absence of EOS 1911/5611 (34.1) 3897/8213 (47.4) <.001 
 Antibiotics in the absence of positive EOS culture (cases and contaminants)b 1890/5590 (33.8) 3862/8177 (47.2) <.001 
 Antibiotics in the absence of a positive blood or CSF culture result, NEC, or SIP ≤7 dc 1771/5334 (33.2) 3649/7752 (47.1) <.001 
 No. infants given prolonged early antibiotics per EOS case 66 19 <.001 
a

P value by the χ2 test.

b

Of the 5640 infants in the low-risk group who survived >12 h, 50 infants with a positive blood or CSF culture result within 72 h of age were excluded (29 counted as EOS cases and 21 with an organism considered a contaminant: 17 CONS, 3 Bacillus sp., 1 Micrococcus sp.). Of the 8422 infants in the other group, 245 were excluded (209 counted as EOS cases and 36 with an organism considered a contaminant: 23 CONS, 3 Bacillus sp., 5 Micrococcus sp., 4 Corynebacterium sp., 1 Lactobacillus sp.).

c

In the low-risk group, 256 infants with NEC, SIP, or LOS within the first 7 d of life were excluded in addition to the 50 infants previously noted (44 infants with NEC, 79 with SIP, 113 with LOS, and 20 with ≥2 of NEC, SIP, or LOS). In the comparison group, 425 infants with NEC, SIP, or LOS within 7 d were excluded in addition to the 245 previously noted (49 infants with NEC, 110 with SIP, 232 with LOS, and 34 with 2 or more of NEC, SIP, or LOS).

FIGURE 1

Center variation in proportion of infants treated with prolonged antibiotics. Shown are the proportion of infants treated with prolonged antibiotics (gray bars) and EOS incidence (black bars) by NRN center, expressed as percent of total infants at each center. Centers are sorted by increasing incidence of EOS. A, Low-risk infant cohort. The rate of EOS was 0% at centers 1, 2, 11, 14, 18, and 22 and 0.2% at centers 6 and 7. B, Comparison infant cohort. The rate of EOS was 0% at center 14. Two centers were excluded from the Figure: 1 center that left the NRN in 2006 and had no low-risk infants and another center that left the NRN in 2011 and had only 2 low-risk infants.

FIGURE 1

Center variation in proportion of infants treated with prolonged antibiotics. Shown are the proportion of infants treated with prolonged antibiotics (gray bars) and EOS incidence (black bars) by NRN center, expressed as percent of total infants at each center. Centers are sorted by increasing incidence of EOS. A, Low-risk infant cohort. The rate of EOS was 0% at centers 1, 2, 11, 14, 18, and 22 and 0.2% at centers 6 and 7. B, Comparison infant cohort. The rate of EOS was 0% at center 14. Two centers were excluded from the Figure: 1 center that left the NRN in 2006 and had no low-risk infants and another center that left the NRN in 2011 and had only 2 low-risk infants.

Close modal

Maternal and delivery characteristics did not differ significantly between the 1771 infants in the low-risk group who received prolonged early antibiotics without a positive culture result and/or a diagnosis of NEC or SIP ≤7 days of life and the 3563 who did not (Supplemental Table 11). However, the low-risk infants who received prolonged early antibiotics were significantly younger and smaller at birth, more likely to be boys, SGA, and have 5 minute Apgar score <5, and more likely to have received surfactant and other interventions (Table 6). After adjustment for GA and other covariates, the adjusted risk of death after 7 days and before discharge and the composite outcomes of death and/or either NEC, SIP, or LOS were increased for low-risk infants who received prolonged early antibiotics compared with low-risk infants who did not (Table 7). Risks of pulmonary outcomes were also higher among low-risk infants who received prolonged early antibiotics, including the risk of BPD among survivors to 36 weeks PMA (aRR [95% CI]: 1.28 [1.20–1.36]).

TABLE 6

Neonatal Characteristics and Early Clinical Interventions for Low-Risk Infants Who Survived >12 Hours, Excluding Infants With a Positive Blood or CSF Culture Result and/or NEC or SIP ≤7 Days

N (Column %) or as ShownProlonged Early Antibiotics (N = 1771)No Prolonged Early Antibiotics (N = 3563)Pa
Infant characteristicsb 
 GA, wk, median (IQR) 26 (25–27) 27 (26–28) <.001 
 By GA week   <.001 
  22 5 (0.3) 4 (0.1)  
  23 67 (3.8) 67 (1.9)  
  24 251 (14.2) 311 (8.7)  
  25 327 (18.5) 492 (13.8)  
  26 350 (19.8) 670 (18.8)  
  27 394 (22.2) 904 (25.4)  
  28 377 (21.3) 1115 (31.3)  
 BW, g, median (IQR) 754 (620–930) 865 (710–1040) <.001 
 SGA 353 (19.9) 411 (11.5) <.001 
 Male 947 (53.5) 1683 (47.2) <.001 
 Apgar <5 at 5 min 340 (19.2) 527 (14.8) <.001 
 First temperature ≤60 min of birth 1644 (93.5) 3314 (93.5) .92 
 Temperature (°F) if ≤60 min, median (IQR) 97.5 (96.5–98.2) 97.7 (96.8–98.3) <.001 
Clinical interventionsc 
 Surfactant 1653 (93.3) 2941 (82.6) <.001 
 Nitric oxide 184 (10.4) 153 (4.3) <.001 
 Treated for hypotension in first 24 hd 296/689 (43.0) 357/1621 (22.0) <.001 
 Highest median (IQR) base deficit in 24 hd 7 (5–9) 6 (4–8) <.001 
 Enteral feeds 1648 (93.1) 3393 (95.2) .001 
 If yes, median (IQR) DOL first enteral feed 5 (3–8) 4 (2–5) <.001 
Infants who survived >24 h 1764 3538 — 
 Respiratory support at 24 he 1695 (96.1) 3306 (93.8) <.001 
 Mechanical ventilation at 24 h 1325 (75.2) 2014 (57.1) <.001 
Infants who survived >3 d 1742 3454 — 
 Any respiratory support day 1–3e 1736 (99.7) 3401 (98.6) <.001 
 Any mechanical ventilation day 1–3 1621 (93.1) 2795 (81.0) <.001 
N (Column %) or as ShownProlonged Early Antibiotics (N = 1771)No Prolonged Early Antibiotics (N = 3563)Pa
Infant characteristicsb 
 GA, wk, median (IQR) 26 (25–27) 27 (26–28) <.001 
 By GA week   <.001 
  22 5 (0.3) 4 (0.1)  
  23 67 (3.8) 67 (1.9)  
  24 251 (14.2) 311 (8.7)  
  25 327 (18.5) 492 (13.8)  
  26 350 (19.8) 670 (18.8)  
  27 394 (22.2) 904 (25.4)  
  28 377 (21.3) 1115 (31.3)  
 BW, g, median (IQR) 754 (620–930) 865 (710–1040) <.001 
 SGA 353 (19.9) 411 (11.5) <.001 
 Male 947 (53.5) 1683 (47.2) <.001 
 Apgar <5 at 5 min 340 (19.2) 527 (14.8) <.001 
 First temperature ≤60 min of birth 1644 (93.5) 3314 (93.5) .92 
 Temperature (°F) if ≤60 min, median (IQR) 97.5 (96.5–98.2) 97.7 (96.8–98.3) <.001 
Clinical interventionsc 
 Surfactant 1653 (93.3) 2941 (82.6) <.001 
 Nitric oxide 184 (10.4) 153 (4.3) <.001 
 Treated for hypotension in first 24 hd 296/689 (43.0) 357/1621 (22.0) <.001 
 Highest median (IQR) base deficit in 24 hd 7 (5–9) 6 (4–8) <.001 
 Enteral feeds 1648 (93.1) 3393 (95.2) .001 
 If yes, median (IQR) DOL first enteral feed 5 (3–8) 4 (2–5) <.001 
Infants who survived >24 h 1764 3538 — 
 Respiratory support at 24 he 1695 (96.1) 3306 (93.8) <.001 
 Mechanical ventilation at 24 h 1325 (75.2) 2014 (57.1) <.001 
Infants who survived >3 d 1742 3454 — 
 Any respiratory support day 1–3e 1736 (99.7) 3401 (98.6) <.001 
 Any mechanical ventilation day 1–3 1621 (93.1) 2795 (81.0) <.001 

Of the 5640 infants in the low risk of EOS group who survived >12 h, the following 306 infants were excluded: 50 with a positive blood or CSF culture result within 72 h of age (29 counted as EOS cases and 21 with an organism considered a contaminant: 17 CONS, 3 Bacillus sp., 1 Micrococcus sp.), 44 with NEC ≤7 d, 79 with SIP ≤7 d, 113 with LOS on days 4–7, and 20 with 2 or more of NEC, SIP, or LOS ≤7 d. DOL, day of life; IQR, interquartile range; —, not applicable.

a

P value by the χ2 test (categorical variables) or the Wilcoxon test (continuous variables).

b

Information was missing as follows: SGA, 2 infants; infant sex, 2; Apgar score, 4; first temperature within 60 min, 32; temperature if within 60 min, 2.

c

Information was missing as follows: surfactant, 1 infant; nitric oxide, 3; timing of enteral feeds, 2; respiratory support at 24 h, 13 infants; respiratory support day 1–3, 4 infants.

d

Infant hypotension treatment and highest base deficit were collected beginning April 1, 2011. In this group, hypotension treatment was missing for 1 infant and highest base deficit was missing for 92 infants.

e

Respiratory support was defined as any one of mechanical ventilation (high frequency or conventional), nasal synchronized intermittent mandatory ventilation, or continuous positive airway pressure received at 24 h or during any of the first 3 d of life.

TABLE 7

Mortality for Low-Risk Infants Who Survived >12 Hours and Morbidities in Survivors >7 Days, Excluding Infants With a Positive Blood or CSF Culture Result and/or With NEC or SIP ≤7 Days

N (Column %)aProlonged Early Antibiotics (N = 1771)No Prolonged Early Antibiotics (N = 3563)Adjusted RR for Outcome (95% CI): Prolonged Early Antibiotics Versus No Antibioticsb
Died >12 h and before discharge 315 (17.8) 429 (12.0) 1.16 (1.01–1.32) 
Died >7 d 236 (13.3) 258 (7.2) 1.52 (1.28–1.80) 
Infants who survived >7 d 1692 3392 — 
 RDS 1683 (99.5) 3339 (98.4) 1.01 (1.00–1.01) 
 Pneumothorax 117 (6.9) 123 (3.6) 1.83 (1.38–2.43) 
 Pulmonary hemorrhage 148 (8.7) 127 (3.7) 1.96 (1.52–2.52) 
 PIEc 83/664 (12.5) 76/1547 (4.9) 1.80 (1.31–2.45) 
 PDA 951 (56.2) 1477 (43.6) 1.15 (1.08–1.22) 
 NEC >7 d 173 (10.2) 315 (9.3) 0.97 (0.80–1.17) 
 NEC >7 d or death before discharge 346 (20.4) 469 (13.8) 1.22 (1.07–1.40) 
 SIP >7 d 55 (3.3) 40 (1.2) 1.86 (1.23–2.82) 
 SIP >7 d or death before discharge 263 (15.5) 286 (8.4) 1.46 (1.24–1.71) 
 LOS >7 d 436 (25.8) 746 (22.0) 1.00 (0.89–1.11) 
 LOS >7 d or death before discharge 583 (34.5) 878 (25.9) 1.12 (1.03–1.23) 
Infants evaluated with cranial imaging 1692 3365 — 
 Severe ICH or PVL 316/1687 (18.7) 321/3357 (9.6) 1.65 (1.40–1.94) 
Infants who survived to 36 wk PMA 1500 3194 — 
 BPD 919/1491 (61.6) 1220/3167 (38.5) 1.28 (1.20–1.36) 
N (Column %)aProlonged Early Antibiotics (N = 1771)No Prolonged Early Antibiotics (N = 3563)Adjusted RR for Outcome (95% CI): Prolonged Early Antibiotics Versus No Antibioticsb
Died >12 h and before discharge 315 (17.8) 429 (12.0) 1.16 (1.01–1.32) 
Died >7 d 236 (13.3) 258 (7.2) 1.52 (1.28–1.80) 
Infants who survived >7 d 1692 3392 — 
 RDS 1683 (99.5) 3339 (98.4) 1.01 (1.00–1.01) 
 Pneumothorax 117 (6.9) 123 (3.6) 1.83 (1.38–2.43) 
 Pulmonary hemorrhage 148 (8.7) 127 (3.7) 1.96 (1.52–2.52) 
 PIEc 83/664 (12.5) 76/1547 (4.9) 1.80 (1.31–2.45) 
 PDA 951 (56.2) 1477 (43.6) 1.15 (1.08–1.22) 
 NEC >7 d 173 (10.2) 315 (9.3) 0.97 (0.80–1.17) 
 NEC >7 d or death before discharge 346 (20.4) 469 (13.8) 1.22 (1.07–1.40) 
 SIP >7 d 55 (3.3) 40 (1.2) 1.86 (1.23–2.82) 
 SIP >7 d or death before discharge 263 (15.5) 286 (8.4) 1.46 (1.24–1.71) 
 LOS >7 d 436 (25.8) 746 (22.0) 1.00 (0.89–1.11) 
 LOS >7 d or death before discharge 583 (34.5) 878 (25.9) 1.12 (1.03–1.23) 
Infants evaluated with cranial imaging 1692 3365 — 
 Severe ICH or PVL 316/1687 (18.7) 321/3357 (9.6) 1.65 (1.40–1.94) 
Infants who survived to 36 wk PMA 1500 3194 — 
 BPD 919/1491 (61.6) 1220/3167 (38.5) 1.28 (1.20–1.36) 

Of the 5640 infants in the low risk of EOS group who survived >12 h, the following 306 infants were excluded: 50 with a positive blood or CSF culture result within 72 h of age (29 counted as EOS cases and 21 with an organism considered a contaminant: 17 CONS, 3 Bacillus sp., 1 Micrococcus sp.), 44 with NEC ≤7 d, 79 with SIP ≤7 d, 113 with LOS on days 4–7, and 20 with ≥2 of NEC, SIP, or LOS ≤7 d. ICH, intracranial hemorrhage; PDA, patent ductus arteriosus; PIE, pulmonary interstitial emphysema; RDS, respiratory distress syndrome; RR, relative risk; —, not applicable.

a

Information was missing for survivors >7 d as follows: PDA, 2 infants; SIP, 2 infants; LOS, 2 infants; ICH and/or PVL, 13 infants; BPD, 36 infants.

b

RRs of each outcome for infants who received antibiotics for ≥5 days started within 72 h of birth versus those who either did not receive antibiotics within 72 h or received antibiotics for <5 d were adjusted for center, maternal race and/or ethnicity, antenatal steroids, multiple birth, GA (categorical), BW (continuous), and sex except as noted below. Because of the number of centers with no infants who had the outcome (or, for RDS, centers with no infants without the outcome), center was not included in models assessing risk of RDS and SIP. All infants born at GA 22–24 wk had RDS; therefore, GA 22–25 wk were combined for the purpose of assessing RDS. GAs 22 and 23 wk were combined in the models assessing SIP and BPD (only 1 of the 9 infants included who were born at GA 22 wk and survived >7 d had SIP before discharge; all 5 infants born at GA 22 wk who survived to 36 wk PMA had BPD).

c

PIE was collected beginning April 1, 2011.

In this NRN cohort of extremely low gestation infants, characteristics evident at birth were able to be used to identify a group of infants with significantly lower risk of EOS. A third of the infants in the study cohort fulfilled the defined low-risk delivery criteria. The lower incidence of EOS in this group was, however, not associated with a proportional reduction in prolonged early antibiotic therapy, presenting an opportunity to revisit current antibiotic prescribing practices.

The lower incidence of EOS in the low-risk cohort confirmed our hypothesis that the absence of factors that might reflect (eg, preterm labor) or promote (eg, membrane rupture) the pathogenesis of ascending infection would be associated with a reduced occurrence of infection. However, 29 cases of EOS were identified in the low-risk group. These cases may be due to imprecision in our definition of these risk factors. The NRN database does not record presence of preterm labor, unexplained fetal distress, or criteria leading to a diagnosis of clinical chorioamnionitis. Therefore, we used CD with ROM at delivery and absence of clinical chorioamnionitis as indirect estimators. It is likely that we included some mothers who labored before being taken for a CD delivery before ROM, providing opportunity for ascending infection. Prospective collection of delivery criteria in real-time practice could improve capture of EOS risk factors and subsequently the precision of identifying infants at low risk for EOS. Hematogenous spread of infection across the placenta and infection after invasive intrauterine procedures can also rarely cause fetal infection.21 Emerging evidence also points to a placental and amniotic fluid microbiome whose disturbance may be associated with preterm labor.22 

Despite these limitations, the 76% lower risk of EOS among low-risk infants compared with the comparison group is substantial and warrants consideration in making management decisions. We recently reported a similar single-center study that included detailed maternal chart review of EOS cases among very low birth weight (VLBW) infants (BW <1500 g). In that study, we found a 92% lower unadjusted risk for EOS among VLBW infants delivered by CD to mothers with a diagnosis of preeclampsia and without diagnoses of preterm ROM or chorioamnionitis, compared with VLBW infants born without these characteristics (1 out of 605 [0.17%] vs 45 out of 2143 [2.1%], P = .001).23 The incidence of EOS in the current study was even lower (0.3% vs 0.5%) among low-risk infants when the definition was refined to exclude infants born in the setting of histologic as well as clinical chorioamnionitis. Placental histopathology results obtained shortly after delivery may further aid in decisions to continue early empirical antibiotic therapy beyond the first few days of age.

Our findings suggest that clinicians did not recognize the differential incidence of EOS among low-risk infants or failed to account for it in their antibiotic management decisions. Although prolonged empirical antibiotic use was lower among low-risk infants than among the comparison infants, the ratio of prolonged antibiotics to number of EOS cases was 3 times higher in the low-risk group. Fewer low-risk infants died at ≤12 hours, but those who survived were as sick as or sicker than comparison infants. Maternal morbidity, SGA status, and a stressed in utero environment in infants delivered for maternal indications are associated with greater respiratory morbidity and may explain the greater initial morbidity observed in the low-risk infants.24,26 Administration of antibiotics for increasing clinical instability and attribution of such instability to culture negative sepsis is well-documented in neonatal care.4,27,28 A previous study revealed that >90% of all infants in an earlier NRN cohort were treated with antibiotics at birth.3 We speculate that the greater need for intensive care interventions in the low-risk cohort and persistence of illness beyond 48 hours led to a reluctance to stop antibiotic treatment. These decisions are likely driven by clinical uncertainties and provider preference more than biology because we observed marked variation around prolonged antibiotic administration across study centers that did not correlate with variation in the incidence of EOS at these centers.

Clinicians make antibiotic and other treatment decisions to protect sick newborns, yet multiple reports now suggest harm from early and prolonged empirical antibiotics.3,6,9 We addressed this issue with an analysis restricted to the low-risk group. To focus on potentially modifiable use of antibiotics, we limited this analysis to low-risk infants who received prolonged early antibiotics despite negative blood culture results and/or NEC or SIP in the first week of life. After adjusting for baseline characteristics associated with increased risk of morbidity and mortality, we still found higher subsequent mortality and increased incidence of BPD and other pulmonary outcomes among the low-risk infants who received prolonged early antibiotics. The decision to extend early empirical therapy may reflect a severity of initial illness that strongly influences subsequent morbidity and mortality in a manner for which we could not adequately account in our risk-adjusted analyses. However, there is evidence for the critical role of the early life microbiome in the development of neonatal immune responses and for the significant disruption of that microbiome induced by antibiotic exposure.29,33 Although with this study we cannot determine why it may be, our findings reveal that among infants with a low previous probability of EOS, extended empirical administration of early antibiotics for critical illness may not be beneficial and could potentially be harmful.

Our study is limited by its retrospective design. During the study period, the registry lacked information on the indication for preterm delivery. We have no information on the presence of preterm labor, attempts to induce labor, or unexplained fetal distress. Information about antibiotic therapy was limited to whether antibiotics were started in the first 72 hours after birth and continued for ≥5 days. Types of antibiotics, total duration of antibiotic administration during the NICU admission, and indication for therapy were not recorded.

Delivery characteristics of infants born at 22 to 28 weeks GA were useful in identifying those with significantly lower risk of EOS. Prolonged early antibiotics were administered to a large proportion of these infants despite their lower a priori risk, and this was associated with higher adjusted incidence of death and pulmonary morbidity. Recognition of differential EOS risk may help guide early empirical antibiotic use among approximately one-third of extremely preterm infants.

     
  • aRR

    adjusted relative risk

  •  
  • BPD

    bronchopulmonary dysplasia

  •  
  • BW

    birth weight

  •  
  • CD

    cesarean delivery

  •  
  • CONS

    coagulase-negative staphylococci

  •  
  • CSF

    cerebrospinal fluid

  •  
  • EOS

    early-onset sepsis

  •  
  • GA

    gestational age

  •  
  • LOS

    late-onset sepsis

  •  
  • NEC

    necrotizing enterocolitis

  •  
  • NRN

    Neonatal Research Network

     
  • PMA

    postmenstrual age

  •  
  • PVL

    periventricular leukomalacia

  •  
  • ROM

    rupture of membranes

  •  
  • SGA

    small for gestational age

  •  
  • SIP

    spontaneous intestinal perforation

  •  
  • VLBW

    very low birth weight

Dr Puopolo conceptualized and designed the study and reviewed and revised the manuscript; Dr Mukhopadhyay designed the study and drafted the initial manuscript; Ms Hansen conducted the statistical analyses and critically reviewed the manuscript; Drs Stoll, Cotten, and Sanchez contributed to the study concept and reviewed and revised the manuscript; Ms Hensman and Drs Bell, Das, Van Meurs, and Wyckoff reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.

FUNDING: The National Institutes of Health: the Eunice Kennedy Shriver National Institute of Child Health and Human Development, the National Center for Research Resources, and the National Center for Advancing Translational Sciences. Funded by the National Institutes of Health (NIH).

The National Institutes of Health, the Eunice Kennedy Shriver National Institute of Child Health and Human Development, the National Center for Research Resources, and the National Center for Advancing Translational Sciences provided grant support for the NRN Generic Database through cooperative agreements. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Data collected at participating sites of the Eunice Kennedy Shriver National Institute of Child Health and Human Development NRN were transmitted to RTI International, the data coordinating center (DCC) for the network, which stored, managed, and analyzed the data for this study. On behalf of the NRN, Dr Abhik Das (DCC Principal Investigator) and Ms Nellie Hansen (DCC Statistician) had full access to all the data in the study and take responsibility for the integrity of the data and accuracy of the data analysis.

We are indebted to our medical and nursing colleagues and the infants and their parents who agreed to take part in this study. The following investigators, in addition to those listed as authors, participated in this study:

NRN Steering Committee Chair: Michael S. Caplan, MD, University of Chicago, Pritzker School of Medicine.

Alpert Medical School of Brown University and Women & Infants Hospital of Rhode Island (U10 HD27904): Abbott R. Laptook, MD; Martin Keszler, MD; Andrea M. Knoll; Emilee Little, RN, BSN; Elisa Vieira, RN, BSN; Kristin M. Basso, RN, MaT; Jennifer A. Keller, RN, BSN.

Case Western Reserve University, Rainbow Babies & Children’s Hospital (U10 HD21364, M01 RR80): Michele C. Walsh, MD, MS; Anna Maria Hibbs, MD; Avroy A. Fanaroff, MD; Nancy S. Newman, BA, RN; Allison H. Payne, MD, MS.

Children’s Mercy Hospital, University of Missouri Kansas City School of Medicine (U10 HD68284): William E. Truog, MD; Eugenia K. Pallotto, MD, MSCE; Howard W. Kilbride, MD; Cheri Gauldin, RN, BS, CCRC; Anne Holmes, RN, MSN, MBA-HCM, CCRC; Kathy Johnson, RN, CCRC; Allison Knutson, BSN, RNC-NIC.

Cincinnati Children’s Hospital Medical Center, University Hospital, and Good Samaritan Hospital (U10 HD27853, M01 RR8084): Kurt Schibler, MD; Edward F. Donovan, MD; Cathy Grisby, BSN, CCRC; Kate Bridges, MD; Barbara Alexander, RN; Estelle E. Fischer, MHSA, MBA; Holly L. Mincey, RN, BSN; Jody Hessling, RN; Lenora Jackson, CRC; Kristin Kirker, CRC; Greg Muthig, BS; Stacey Tepe, BS.

Duke University School of Medicine, University Hospital, University of North Carolina, and Duke Regional Hospital (U10 HD40492, UL1 TR1117, M01 RR30, UL1 TR1111): Ronald N. Goldberg, MD; Kathy J. Auten, MSHS; Kimberley A. Fisher, PhD, FNP-BC, IBCLC; Joanne Finkle, RN, JD; Matthew M. Laughon, MD, MPH; Carl L. Bose, MD; Janice Bernhardt, MS, RN; Gennie Bose, RN.

Emory University, Children’s Healthcare of Atlanta, Grady Memorial Hospital, and Emory University Hospital Midtown (U10 HD27851, M01 RR39): David P. Carlton, MD; Ellen C. Hale, RN, BS, CCRC; Yvonne Loggins, RN, BSN; Diane I. Bottcher, RN, MSN; Colleen Mackie, BS, RT.

Eunice Kennedy Shriver National Institute of Child Health and Human Development: Rosemary D. Higgins, MD; Stephanie Wilson Archer, MA.

Indiana University, University Hospital, Methodist Hospital, Riley Hospital for Children, and Wishard Health Services (U10 HD27856, M01 RR750): Brenda B. Poindexter, MD, MS; Gregory M. Sokol, MD; Dianne E. Herron, RN, CCRC; Lucy Miller, BSN, CCRC; Leslie Dawn Wilson, BSN, CCRC.

McGovern Medical School at The University of Texas Health Science Center at Houston, Children’s Memorial Hermann Hospital, Memorial Hermann Southwest Hospital, and Lyndon Baines Johnson General Hospital/Harris County Hospital District (U10 HD21373): Kathleen A. Kennedy, MD, MPH; Jon E. Tyson, MD, MPH; Georgia E. McDavid, RN; Julie Arldt-McAlister, RN, BSN; Katrina Burson, RN, BSN; Carmen Garcia, RN, CCRP; Beverly Foley Harris, RN, BSN; Anna E. Lis, RN, BSN; Karen Martin, RN; Sara C. Martin, RN, BSN; Shawna Rodgers, RN; Maegan C. Simmons, RN; Patti L. Pierce Tate, RCP.

Nationwide Children’s Hospital and The Ohio State University Wexner Medical Center (U10 HD68278): Leif D. Nelin, MD; Sudarshan R. Jadcherla, MD; Patricia Luzader, RN; Christine A. Fortney, PhD, RN; Nehal A. Parikh, MD.

RTI International (U10 HD36790): Dennis Wallace, PhD; Marie G. Gantz, PhD; W. Kenneth Poole, PhD (deceased); Jeanette O’Donnell Auman, BS; Margaret M. Crawford, BS, CCRP; Carolyn M. Petrie Huitema, MS, CCRP; Kristin M. Zaterka-Baxter, RN, BSN, CCRP.

Stanford University, Dominican Hospital, El Camino Hospital, and Lucile Packard Children’s Hospital (U10 HD27880, M01 RR70): David K. Stevenson, MD; Marian M. Adams, MD; M. Bethany Ball, BS, CCRC; Magdy Ismail, MD, MPH; Andrew W. Palmquist, RN; Melinda S. Proud, RCP.

Tufts Medical Center, Floating Hospital for Children (U10 HD53119, M01 RR54): Ivan D. Frantz III, MD; John M. Fiascone, MD; Brenda L. MacKinnon, RNC; Anne Furey, MPH; Ellen Nylen, RN, BSN.

University of Alabama at Birmingham Health System and Children’s Hospital of Alabama (U10 HD34216, M01 RR32): Waldemar A. Carlo, MD; Namasivayam Ambalavanan, MD; Monica V. Collins, RN, BSN, MaEd; Shirley S. Cosby, RN, BSN.

University of California Los Angeles, Mattel Children’s Hospital, Santa Monica Hospital, Los Robles Hospital and Medical Center, and Olive View Medical Center (U10 HD68270): Uday Devaskar, MD; Meena Garg, MD; Teresa Chanlaw, MPH; Rachel Geller, RN, BSN.

University of California San Diego Medical Center and Sharp Mary Birch Hospital for Women and Newborns (U10 HD40461): Neil N. Finer, MD; David Kaegi, MD; Maynard R. Rasmussen, MD; Kathy Arnell, RNC; Clarence Demetrio, RN; Wade Rich, BSHS, RRT.

University of Iowa and Mercy Medical Center (U10 HD53109, M01 RR59): Tarah T. Colaizy, MD, MPH; Dan L. Ellsbury, MD; John A. Widness, MD; Karen J. Johnson, RN, BSN; Donia B. Campbell, RNC-NC; Jacky R. Walker, RN; Tracy L. Tud, RN.

University of Miami, Holtz Children’s Hospital (U10 HD21397, M01 RR16587): Shahnaz Duara, MD; Charles R. Bauer, MD; Ruth Everett-Thomas, RN, MSN.

University of New Mexico Health Sciences Center (U10 HD53089, M01 RR997): Kristi L. Watterberg, MD; Robin K. Ohls, MD; Conra Backstrom Lacy, RN; Carol H. Hartenberger, MPH, RN; Sandra Sundquist Beauman, MSN, RNC-NIC, Mary Ruffaner Hanson, RN, BSN.

University of Pennsylvania, Hospital of the University of Pennsylvania, Pennsylvania Hospital, and Children’s Hospital of Philadelphia (U10 HD68244): Barbara Schmidt, MD, MSc; Haresh Kirpalani, MB, MSc; Sara B. DeMauro, MD, MSCE; Aasma S. Chaudhary, BS, RRT; Soraya Abbasi, MD; Toni Mancini, RN, BSN, CCRC; Dara M. Cucinotta, RN.

University of Rochester Medical Center, Golisano Children’s Hospital, and the University of Buffalo Women’s and Children’s Hospital of Buffalo (U10 HD68263, U10 HD40521, M01 RR44, UL1 TR42): Carl T. D’Angio, MD; Dale L. Phelps, MD; Ronnie Guillet, MD, PhD; Satyan Lakshminrusimha, MD; Linda J. Reubens, RN, CCRC; Cassandra A. Horihan, MS; Mary Rowan, RN; Holly I.M. Wadkins, MA; Rosemary Jensen; Melissa Bowman, MSN; Julianne Hunn, BS; Stephanie Guilford, BS; Deanna Maffett, RN; Diane Prinzing; Anne Marie Reynolds, MD, MPH; Ashley Williams, MSEd; Karen Wynn, RN; Erica Burnell, RN; Michael G. Sacilowski, MAT, CCRC.

University of Texas Southwestern Medical Center at Dallas, Parkland Health & Hospital System, and Children’s Medical Center Dallas (U10 HD40689, M01 RR633): Luc P. Brion, MD; Walid A. Salhab, MD; Charles R. Rosenfeld, MD; Diana M. Vasil, MSN, BSN, RNC-NIC; Lijun Chen, PhD, RN; Alicia Guzman; Gaynelle Hensley, RN; Lizette E. Lee, RN; Melissa H. Leps, RN; Nancy A. Miller, RN; Janet S. Morgan, RN; Lara Pavageau, MD.

University of Utah Medical Center, Intermountain Medical Center, LDS Hospital, and Primary Children’s Medical Center (U10 HD53124, M01 RR64, UL1 RR25764): Roger G. Faix, MD; Bradley A. Yoder, MD; Karen A. Osborne, RN, BSN, CCRC; Cynthia Spencer, RNC, BSN; Kimberlee Weaver-Lewis, RN, MS; Karie Bird, RN, BSN; Jill Burnett, RNC, BSN; Jennifer J. Jensen, RN, BSN; Karen Zanetti, RN.

Wake Forest University, Baptist Medical Center, Forsyth Medical Center, and Brenner Children’s Hospital (U10 HD40498, M01 RR7122): T. Michael O’Shea, MD, MPH; Nancy Peters, RN.

Wayne State University, Hutzel Women’s Hospital, and Children’s Hospital of Michigan (U10 HD21385): Seetha Shankaran, MD; Athina Pappas, MD; John Barks, MD; Rebecca Bara, RN, BSN; Girija Natarajan, MD; Mary Christensen, RT; Stephanie A. Wiggins, MS; Diane White, RT.

Yale University, Yale-New Haven Children’s Hospital, and Bridgeport Hospital (U10 HD27871, ULTR142, M01 RR125): Richard A. Ehrenkranz, MD; Harris Jacobs, MD; Patricia Cervone, RN; Monica Konstantino, RN, BSN; JoAnn Poulsen, RN; Janet Taft, RN, BSN.

1
Schrag
SJ
,
Farley
MM
,
Petit
S
, et al
.
Epidemiology of invasive early-onset neonatal sepsis, 2005 to 2014.
Pediatrics
.
2016
;
138
(
6
):
e20162013
[PubMed]
2
Stoll
BJ
,
Hansen
NI
,
Sánchez
PJ
, et al;
Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network
.
Early onset neonatal sepsis: the burden of group B Streptococcal and E. coli disease continues.
Pediatrics
.
2011
;
127
(
5
):
817
826
[PubMed]
3
Cotten
CM
,
Taylor
S
,
Stoll
B
, et al;
NICHD Neonatal Research Network
.
Prolonged duration of initial empirical antibiotic treatment is associated with increased rates of necrotizing enterocolitis and death for extremely low birth weight infants.
Pediatrics
.
2009
;
123
(
1
):
58
66
[PubMed]
4
Cordero
L
,
Ayers
LW
.
Duration of empiric antibiotics for suspected early-onset sepsis in extremely low birth weight infants.
Infect Control Hosp Epidemiol
.
2003
;
24
(
9
):
662
666
[PubMed]
5
Schulman
J
,
Dimand
RJ
,
Lee
HC
,
Duenas
GV
,
Bennett
MV
,
Gould
JB
.
Neonatal intensive care unit antibiotic use.
Pediatrics
.
2015
;
135
(
5
):
826
833
[PubMed]
6
Alexander
VN
,
Northrup
V
,
Bizzarro
MJ
.
Antibiotic exposure in the newborn intensive care unit and the risk of necrotizing enterocolitis.
J Pediatr
.
2011
;
159
(
3
):
392
397
[PubMed]
7
Kuppala
VS
,
Meinzen-Derr
J
,
Morrow
AL
,
Schibler
KR
.
Prolonged initial empirical antibiotic treatment is associated with adverse outcomes in premature infants.
J Pediatr
.
2011
;
159
(
5
):
720
725
[PubMed]
8
Benjamin
DK
 Jr
,
Stoll
BJ
,
Gantz
MG
, et al;
Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network
.
Neonatal candidiasis: epidemiology, risk factors, and clinical judgment.
Pediatrics
.
2010
;
126
(
4
). Available at: www.pediatrics.org/cgi/content/full/126/4/e865
[PubMed]
9
Ting
JY
,
Synnes
A
,
Roberts
A
, et al;
Canadian Neonatal Network Investigators
.
Association between antibiotic use and neonatal mortality and morbidities in very low-birth-weight infants without culture-proven sepsis or necrotizing enterocolitis.
JAMA Pediatr
.
2016
;
170
(
12
):
1181
1187
[PubMed]
10
Benirschke
K
.
Routes and types of infection in the fetus and the newborn.
AMA J Dis Child
.
1960
;
99
(
6
):
714
721
[PubMed]
11
Baker
CJ
,
Barrett
FF
.
Transmission of group B streptococci among parturient women and their neonates.
J Pediatr
.
1973
;
83
(
6
):
919
925
[PubMed]
12
Benitz
WE
,
Gould
JB
,
Druzin
ML
.
Risk factors for early-onset group B streptococcal sepsis: estimation of odds ratios by critical literature review.
Pediatrics
.
1999
;
103
(
6
). Available at: www.pediatrics.org/cgi/content/full/103/6/e77
[PubMed]
13
Schrag
SJ
,
Zell
ER
,
Lynfield
R
, et al;
Active Bacterial Core Surveillance Team
.
A population-based comparison of strategies to prevent early-onset group B streptococcal disease in neonates.
N Engl J Med
.
2002
;
347
(
4
):
233
239
[PubMed]
14
Henderson
JJ
,
McWilliam
OA
,
Newnham
JP
,
Pennell
CE
.
Preterm birth aetiology 2004-2008. Maternal factors associated with three phenotypes: spontaneous preterm labour, preterm pre-labour rupture of membranes and medically indicated preterm birth.
J Matern Fetal Neonatal Med
.
2012
;
25
(
6
):
642
647
[PubMed]
15
Stoll
BJ
,
Hansen
NI
,
Bell
EF
, et al;
Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network
.
Trends in care practices, morbidity, and mortality of extremely preterm neonates, 1993-2012.
JAMA
.
2015
;
314
(
10
):
1039
1051
[PubMed]
16
Walsh
MC
,
Kliegman
RM
.
Necrotizing enterocolitis: treatment based on staging criteria.
Pediatr Clin North Am
.
1986
;
33
(
1
):
179
201
[PubMed]
17
Bell
MJ
,
Ternberg
JL
,
Feigin
RD
, et al
.
Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging.
Ann Surg
.
1978
;
187
(
1
):
1
7
[PubMed]
18
Papile
LA
,
Burstein
J
,
Burstein
R
,
Koffler
H
.
Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm.
J Pediatr
.
1978
;
92
(
4
):
529
534
[PubMed]
19
Alexander
GR
,
Himes
JH
,
Kaufman
RB
,
Mor
J
,
Kogan
M
.
A United States national reference for fetal growth.
Obstet Gynecol
.
1996
;
87
(
2
):
163
168
[PubMed]
20
Zou
G
.
A modified poisson regression approach to prospective studies with binary data.
Am J Epidemiol
.
2004
;
159
(
7
):
702
706
[PubMed]
21
Goldenberg
RL
,
Culhane
JF
,
Iams
JD
,
Romero
R
.
Epidemiology and causes of preterm birth.
Lancet
.
2008
;
371
(
9606
):
75
84
[PubMed]
22
DiGiulio
DB
,
Romero
R
,
Amogan
HP
, et al
.
Microbial prevalence, diversity and abundance in amniotic fluid during preterm labor: a molecular and culture-based investigation.
PLoS One
.
2008
;
3
(
8
):
e3056
[PubMed]
23
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
(
5
):
477
481
[PubMed]
24
Chen
A
,
Feresu
SA
,
Barsoom
MJ
.
Heterogeneity of preterm birth subtypes in relation to neonatal death.
Obstet Gynecol
.
2009
;
114
(
3
):
516
522
[PubMed]
25
Mendola
P
,
Mumford
SL
,
Männistö
TI
,
Holston
A
,
Reddy
UM
,
Laughon
SK
.
Controlled direct effects of preeclampsia on neonatal health after accounting for mediation by preterm birth.
Epidemiology
.
2015
;
26
(
1
):
17
26
[PubMed]
26
Bernstein
IM
,
Horbar
JD
,
Badger
GJ
,
Ohlsson
A
,
Golan
A
;
The Vermont Oxford Network
.
Morbidity and mortality among very-low-birth-weight neonates with intrauterine growth restriction.
Am J Obstet Gynecol
.
2000
;
182
(
1, pt 1
):
198
206
[PubMed]
27
Escobar
GJ
.
What have we learned from observational studies on neonatal sepsis?
Pediatr Crit Care Med
.
2005
;
6
(
suppl 3
):
S138
S145
[PubMed]
28
Piantino
JH
,
Schreiber
MD
,
Alexander
K
,
Hageman
J
.
Culture negative sepsis and systemic inflammatory response syndrome in neonates.
Neoreviews
.
2013
;
14
(
6
):
e294
29
Arboleya
S
,
Sánchez
B
,
Milani
C
, et al
.
Intestinal microbiota development in preterm neonates and effect of perinatal antibiotics.
J Pediatr
.
2015
;
166
(
3
):
538
544
[PubMed]
30
Azad
MB
,
Konya
T
,
Persaud
RR
, et al;
CHILD Study Investigators
.
Impact of maternal intrapartum antibiotics, method of birth and breastfeeding on gut microbiota during the first year of life: a prospective cohort study.
BJOG
.
2016
;
123
(
6
):
983
993
[PubMed]
31
Deshmukh
HS
,
Liu
Y
,
Menkiti
OR
, et al
.
The microbiota regulates neutrophil homeostasis and host resistance to Escherichia coli K1 sepsis in neonatal mice.
Nat Med
.
2014
;
20
(
5
):
524
530
[PubMed]
32
Olszak
T
,
An
D
,
Zeissig
S
, et al
.
Microbial exposure during early life has persistent effects on natural killer T cell function.
Science
.
2012
;
336
(
6080
):
489
493
[PubMed]
33
Vangay
P
,
Ward
T
,
Gerber
JS
,
Knights
D
.
Antibiotics, pediatric dysbiosis, and disease.
Cell Host Microbe
.
2015
;
17
(
5
):
553
564
[PubMed]

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data