Multiply robust causal inference of the restricted mean survival time difference.

The hazard ratio (HR) remains the most frequently employed metric in assessing treatment effects on survival times. However, the difference in restricted mean survival time (RMST) has become a popular alternative to the HR when the proportional hazards assumption is considered untenable. Moreover, independent of the proportional hazards assumption, many comparative effectiveness studies aim to base contrasts on survival probability rather than on the hazard function. Causal effects based on RMST are often estimated via inverse probability of treatment weighting (IPTW). However, this approach generally results in biased results when the assumed propensity score model is misspecified. Motivated by the need for more robust techniques, we propose an empirical likelihood-based weighting approach that allows for specifying a set of propensity score models. The resulting estimator is consistent when the postulated model set contains a correct model; this property has been termed multiple robustness. In this report, we derive and evaluate a multiply robust estimator of the causal between-treatment difference in RMST. Simulation results confirm its robustness. Compared with the IPTW estimator from a correct model, the proposed estimator tends to be less biased and more efficient in finite samples. Additional simulations reveal biased results from a direct application of machine learning estimation of propensity scores. Finally, we apply the proposed method to evaluate the impact of intrapartum group B streptococcus antibiotic prophylaxis on the risk of childhood allergic disorders using data derived from electronic medical records from the Children's Hospital of Philadelphia and census data from the American Community Survey.

Antibiotic Use Among Infants Admitted to Neonatal Intensive Care Units.

This cross-sectional study examines antibiotic exposure, days of therapy, types of antibiotics, and changes in use patterns among newborns in neonatal intensive care units (NICUs) across the US from 2009 to 2021.

Association of delivery risk phenotype with early-onset sepsis in preterm infants.

OBJECTIVE: To determine delivery risk phenotype-specific incidence of early-onset sepsis (EOS) among preterm infants. STUDY DESIGN: Retrospective cohort study of infants born <35 weeks' gestation at four perinatal centers during 2017-2021. Infants were classified into one of six delivery risk phenotypes incorporating delivery mode, presence of labor, and duration of rupture of membranes (ROM). The primary outcome was EOS incidence within the overall cohort and each risk phenotype. RESULTS: Among 2937 preterm infants, 21 had EOS (0.7%, or 7.1 cases/1000 preterm infants). The majority of EOS cases (13/21, 62%) occurred in the setting of prolonged ROM ≥ 18 h, with a phenotype incidence of 23.8 cases/1000 preterm infants. There were no EOS cases among infants born by cesarean section without ROM (with or without labor), nor via cesarean section with ROM < 18 h without labor. CONCLUSION: Delivery risk phenotyping may inform EOS risk stratification in preterm infants.

Early antibiotic exposure in very-low birth weight infants and infection risk at 3-7 days after birth.

OBJECTIVE: To determine rates of late-onset infection (LOI) during postnatal days 3-7 among preterm infants, based on antibiotic exposure during days 0-2. STUDY DESIGN: Retrospective cohort study of infants born <1500 grams and ≤30 weeks gestation, 2005-2018. We analyzed the incidence and microbiology of LOI at days 3-7 based on antibiotic exposure during postnatal days 0-2. RESULTS: The cohort included 88,574 infants, of whom 85% were antibiotic-exposed. Fewer antibiotic-exposed compared to unexposed infants developed LOI (1.5% vs. 2.1%; adjusted hazard ratio, 0.28, 95% CI 0.24-0.33). Among antibiotic-exposed compared to unexposed infants, Gram-negative (38% vs. 28%, p = 0.002) and fungal (11% vs. 1%, p < 0.001) species were more commonly isolated, and gram-positive organisms (49% vs. 70%, p < 0.001) were less commonly isolated. CONCLUSIONS: We observed low overall rates of LOI at days 3-7 after birth, but antibiotic exposure from birth was associated with lower rates, and with differing microbiology, compared to no exposure.

Higher body mass index after intrapartum antibiotic exposure in children persists over 10-years.

Exposure to intrapartum antibiotic prophylaxis to reduce perinatal group B streptococcal disease was associated with increased childhood body mass index (BMI) persisting to age 10 years compared to no exposure (Δ BMI at 10 years: vaginal delivery 0.14 kg/m2 , caesarean 0.40 kg/m2 ).

Maternal and Newborn Hospital Outcomes of Perinatal SARS-CoV-2 Infection: A National Registry.

OBJECTIVES: The American Academy of Pediatrics National Registry for the Surveillance and Epidemiology of Perinatal coronavirus disease 2019 (COVID-19) (NPC-19) was developed to provide information on the effects of perinatal severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. METHODS: National Registry for the Surveillance and Epidemiology of Perinatal COVID-19 participating centers entered maternal and newborn data for pregnant persons who tested positive for SARS-CoV-2 infection between 14 days before and 10 days after delivery. Incidence of and morbidities associated with maternal and newborn SARS-CoV-2 infection were assessed. RESULTS: From April 6, 2020 to March 19, 2021, 242 centers in the United States centers reported data for 7524 pregnant persons; at the time of delivery, 78.1% of these persons were asymptomatic, 18.2% were symptomatic but not hospitalized specifically for COVID-19, 3.4% were hospitalized for COVID-19 treatment, and 18 (0.2%) died in the hospital of COVID-related complications. Among 7648 newborns, 6486 (84.8%) were tested for SARS-CoV-2, and 144 (2.2%) were positive; the highest rate of newborn infection was observed when mothers first tested positive in the immediate postpartum period (17 of 125, 13.6%). No newborn deaths were attributable to SARS-CoV-2 infection. Overall, 15.6% of newborns were preterm: among tested newborns, 30.1% of polymerase chain reaction-positive and 16.2% of polymerase chain reaction-negative were born preterm (P < .001). Need for mechanical ventilation did not differ by newborn SARS-CoV-2 test result, but those with positive tests were more likely to be admitted to a NICU. CONCLUSIONS: Early in the pandemic, SARS-CoV-2 infection was acquired by newborns at variable rates and without apparent short-term effects. During a period that preceded widespread availability of vaccines, we observed higher than expected numbers of preterm births and maternal in-hospital deaths.

Diagnostic Performance and Patient Outcomes With C-Reactive Protein Use in Early-Onset Sepsis Evaluations.

OBJECTIVES: To determine performance of C-reactive protein (CRP) in the diagnosis of early-onset sepsis, and to assess patient outcomes with and without routine use of CRP. STUDY DESIGN: This was a retrospective cohort study of infants admitted to 2 neonatal intensive care units. CRP was used routinely in early-onset sepsis evaluations during 2009-2014; this period was used to determine CRP performance at a cut-off of ≥10 mg/L in diagnosis of culture-confirmed early-onset sepsis. Routine CRP use was discontinued during 2018-2020; outcomes among infants admitted during this period were compared with those in 2012-2014. RESULTS: From 2009 to 2014, 10 134 infants were admitted; 9103 (89.8%) had CRP and 7549 (74.5%) had blood culture obtained within 3 days of birth. CRP obtained ±4 hours from blood culture had a sensitivity of 41.7%, specificity 89.9%, and positive likelihood ratio 4.12 in diagnosis of early-onset sepsis. When obtained 24-72 hours after blood culture, sensitivity of CRP increased (89.5%), but specificity (55.7%) and positive likelihood ratio (2.02) decreased. Comparing the periods with (n = 4977) and without (n = 5135) routine use of CRP, we observed lower rates of early-onset sepsis evaluation (74.5% vs 50.5%), antibiotic initiation (65.0% vs 50.8%), and antibiotic prolongation in the absence of early-onset sepsis (17.3% vs 7.2%) in the later period. Rate and timing of early-onset sepsis detection, transfer to a greater level of care, and in-hospital mortality were not different between periods. CONCLUSIONS: CRP diagnostic performance was not sufficient to guide decision-making in early-onset sepsis. Discontinuation of routine CRP use was not associated with differences in patient outcomes despite lower rates of antibiotic administration.

Knowledge, Attitudes, and Perceptions about Antibiotic Stewardship Programs among Neonatology Trainees.

OBJECTIVE: Antibiotic stewardship should be an essential component of neonatology training as neonatal intensive care units (NICU) have unique stewardship needs. Our aim was to assess neonatology fellowship trainees' knowledge, attitudes, and perceptions about antibiotic stewardship to inform sustainable curriculum development. STUDY DESIGN: We distributed an electronic survey to neonatology fellows in the United States over 4 months (January-April 2018) via Accreditation Council for Graduate Medical Education program directors. RESULTS: Of 99 programs in the United States with an estimated 700 fellows, 159 (23%) fellows from 52 training programs (53%) responded to the survey and 139 (87%) provided analyzed responses. Majority of respondents were training in southern (59; 42%) and northeastern (43; 31%) regions and were equally spread across all 3 years of training. One hundred (72%) respondents reported an antibiotic stewardship program (ASP) in their institution. While 86% (120/139) were able to identify the components of an ASP, 59% (82/139) either did not or were unsure if they had received antibiotic stewardship training during fellowship.Furthermore, while answering case studies, 124 (89%) respondents identified the optimal antibiotic for methicillin susceptible Staphylococcus aureus (MSSA) infection and 69 (50%) respondents chose appropriate empiric antibiotics for neonatal meningitis. Notably, fellowship training year was not significantly related to the proportion of incorrect knowledge responses (p = 0.40). Most survey respondents (81; 59%) identified small group sessions as the most useful teaching format, while others chose audit and feedback of individual prescribing behavior (52; 38%) and didactic lectures (52; 38%). Finally, ninety-five (69%) respondents preferred trainee-led ASP interventions targeting focal areas such as antifungal and surgical prophylaxis. CONCLUSION: Antibiotic stewardship is a critical part of neonatology training. Neonatology fellows report variation in access to ASP during their training. Fellows prefer dedicated trainee-led interventions and stewardship curriculum taught within small group settings to promote targeted NICU ASP. KEY POINTS: · Most neonatology programs expose trainees to internal or external antibiotic stewardship programs.. · Over half of fellow trainees are unsure about receiving targeted antibiotic stewardship training.. · Most neonatology fellows prefer a trainee-led antibiotic stewardship intervention..

Preterm infants at low risk for early-onset sepsis differ in early fecal microbiome assembly.

Antibiotics are administered near-universally to very low birth weight (VLBW) infants after birth for suspected early-onset sepsis (EOS). We previously identified a phenotypic group of VLBW infants, referred to as low-risk for EOS (LRE), whose risk of EOS is low enough to avoid routine antibiotic initiation. In this cohort study, we compared 18 such infants with 30 infants categorized as non-LRE to determine if the lower risk of pathogen transmission at birth is accompanied by differences in microbiome acquisition and development. We did shotgun metagenomic sequencing of 361 fecal samples obtained serially. LRE infants had a higher human-to-bacterial DNA ratio than non-LRE infants in fecal samples on days 1-3 after birth, confirming lower bacterial acquisition among LRE infants. The microbial diversity and composition in samples from days 4-7 differed between the groups with a predominance of Staphylococcus epidermidis in LRE infants and Enterobacteriaceae sp. in non-LRE infants. Compositional differences were congruent with the distribution of virulence factors and antibiotic resistant genes. After the first week, the overall composition was similar, but changes in relative abundance for several taxa with increasing age differed between groups. Of the nine late-onset bacteremia episodes, eight occurred in non-LRE infants. Species isolated from the blood culture was detected in the pre-antibiotic fecal samples of the infant for all episodes, though these species were also found in infants without bacteremia. In conclusion, LRE infants present a distinct pattern of microbiome development that is aligned with their low risk for EOS. Further investigation to determine the impact of these differences on later outcomes such as late-onset bacteremia is warranted.

Comparison of Maternal and Neonatal Antibody Levels After COVID-19 Vaccination vs SARS-CoV-2 Infection.

Importance: Pregnant persons are at an increased risk of severe COVID-19 from SARS-CoV-2 infection, and COVID-19 vaccination is currently recommended during pregnancy. Objective: To ascertain the association of vaccine type, time from vaccination, gestational age at delivery, and pregnancy complications with placental transfer of antibodies to SARS-CoV-2. Design, Setting, and Participants: This cohort study was conducted in Pennsylvania Hospital in Philadelphia, Pennsylvania, and included births at the study site between August 9, 2020, and April 25, 2021. Maternal and cord blood serum samples were available for antibody level measurements for maternal-neonatal dyads. Exposures: SARS-CoV-2 infection vs COVID-19 vaccination. Main Outcomes and Measures: IgG antibodies to the receptor-binding domain of the SARS-CoV-2 spike protein were measured by quantitative enzyme-linked immunosorbent assay. Antibody concentrations and transplacental transfer ratios were measured after SARS-CoV-2 infection or receipt of COVID-19 vaccines. Results: A total of 585 maternal-newborn dyads (median [IQR] maternal age, 31 [26-35] years; median [IQR] gestational age, 39 [38-40] weeks) with maternal IgG antibodies to SARS-CoV-2 detected at the time of delivery were included. IgG was detected in cord blood from 557 of 585 newborns (95.2%). Among 169 vaccinated persons without SARS-CoV-2 infection, the interval from first dose of vaccine to delivery ranged from 12 to 122 days. The geometric mean IgG level among 169 vaccine recipients was significantly higher than that measured in 408 persons after infection (33.88 [95% CI, 27.64-41.53] arbitrary U/mL vs 2.80 [95% CI, 2.50-3.13] arbitrary U/mL). Geometric mean IgG levels were higher after vaccination with the mRNA-1273 (Moderna) vaccine compared with the BNT162b2 (Pfizer/BioNTech) vaccine (53.74 [95% CI, 40.49-71.33] arbitrary U/mL vs 25.45 [95% CI, 19.17-33.79] arbitrary U/mL; P < .001). Placental transfer ratios were lower after vaccination compared with after infection (0.80 [95% CI, 0.68-0.93] vs 1.06 [95% CI, 0.98-1.14]; P < .001) but were similar between the mRNA vaccines (mRNA-1273: 0.70 [95% CI, 0.55-0.90]; BNT162b2: 0.85 [95% CI, 0.69-1.06]; P = .25). Time from infection or vaccination to delivery was associated with transfer ratio in models that included gestational age at delivery and maternal hypertensive disorders, diabetes, and obesity. Placental antibody transfer was detectable as early as 26 weeks' gestation. Transfer ratio that was higher than 1.0 was present for 48 of 51 (94.1%) births at 36 weeks' gestation or later by 8 weeks after vaccination. Conclusions and Relevance: This study found that maternal and cord blood IgG antibody levels were higher after COVID-19 vaccination compared with after SARS-CoV-2 infection, with slightly lower placental transfer ratios after vaccination than after infection. The findings suggest that time from infection or vaccination to delivery was the most important factor in transfer efficiency.

Breast Milk and Saliva for Postnatal Cyto†megalovirus Screening among Very Low Birth Weight Infants.

BACKGROUND: The optimal approach to managing postnatal cytomegalovirus disease (pCMV) among very low birth weight (VLBW) infants remains unknown. Methods to facilitate screening are needed. OBJECTIVE: Determine whether mother's milk and infant saliva can be used to reliably identify maternal cytomegalovirus (CMV) serostatus and detect infant pCMV acquisition. METHODS: This was a single-center, prospective cohort study of VLBW infants, and their mothers, born between 2017 and 2020. Maternal milk samples were tested for CMV immunoglobulin G (IgG) using a CMV glycoprotein B binding enzyme-linked immunosorbent assay and the results were compared with maternal serum CMV IgG results. Biweekly paired saliva and urine samples were collected from infants born to mothers with positive or unknown CMV serostatus. Saliva samples were tested for CMV DNA by quantitative real-time polymerase chain reaction (PCR) and compared with urine CMV qualitative PCR results obtained from a clinical laboratory. RESULTS: Among 108 infants without congenital CMV included in the study, 10 (9%) acquired pCMV. Both milk and blood CMV serology results were available for 70 mothers. Maternal milk antibody testing had a sensitivity of 97.2% (95% CI: 85.5-99.9%) and specificity of 91.2% (95% CI: 76.3-98.1%) in establishing CMV serostatus. Paired serially collected saliva and urine samples (n = 203) were available for 66 infants. Saliva PCR had a sensitivity of 30.0% (95% CI: 6.7-65.2%) and specificity of 92.7% (95% CI: 88.1-96.0%) in detecting pCMV acquisition. CONCLUSIONS: Maternal breast milk is a reliable alternative sample to determine CMV serostatus. Serial testing of infant saliva was not adequately sensitive for identifying pCMV acquisition in preterm infants.

Respiratory indices during sleep in healthy infants: A prospective longitudinal study and meta-analysis.

STUDY OBJECTIVES: Healthy infants may have a greater apnea hypopnea index (AHI) than older children during the newborn period, but the trajectory of these sleep-related events beyond the first month of life is poorly understood. In this study, we evaluated the longitudinal changes in respiratory indices during sleep in healthy infants during the first six months of life. METHODS: Single-center prospective cohort study. Thirty healthy infants underwent overnight in-lab polysomnography at one and five months of age and findings were compared between assessments. Systematic review of studies evaluating infant polysomnography and meta-analysis was conducted. RESULTS: At one month of age, total AHI, obstructive AHI, and central AHI model-adjusted means (95% confidence interval) were 16.9 events/hour (12.2, 21.5), 10.2 events/hour (7.4, 13.1), and 6.6 events/hour (4.2, 9.0), respectively. 16.8% of events were obstructive apneas and 36.1% central apneas. By five months of age, there were significant reductions in each index to 4.1 events/hour (3.2, 5.0), 1.9 events/hour (1.4, 2.4), and 2.2 events/hour (1.6, 2.9), respectively (p < 0.001 for each), and a lower proportion of events were obstructive apneas (8.6%, p = 0.007) and a greater proportion central apneas (52.3%, p = 0.002). Meta-analysis found high AHI in infants with significant heterogeneity. CONCLUSIONS: Central AHI and obstructive AHI are greater in healthy newborns than older children. There is a significant spontaneous reduction in events and change in type of events in the first six months of life in this low-risk population. These findings may serve as a reference for clinicians evaluating for obstructive sleep apnea in infants.

Perinatal COVID-19 maternal and neonatal outcomes at two academic birth hospitals.

OBJECTIVE: Describe 1-month outcomes among newborns of persons with perinatal COVID-19. STUDY DESIGN: Prospective observational study of pregnant persons who tested positive for SARS-CoV-2 between 14 days before and 3 days after delivery and their newborns, from 3/2020 to 3/2021 at two urban high-risk academic hospitals. Phone interviews were conducted to determine 1-month newborn outcomes. RESULTS: Among 9748 pregnant persons, 209 (2.1%) tested positive for perinatal SARS-CoV-2. Symptomatically infected persons were more likely to have a preterm delivery due to worsening maternal condition and their newborns were more likely to test positive for SARS-CoV-2 compared with asymptomatic persons. Six of 191 (3.1%) infants tested were positive for SARS-CoV-2; none had attributable illness before discharge. Of 169 eligible families, 132 (78.1%) participated in post-discharge interviews; none reported their newborn tested positive for SARS-CoV-2 by 1 month of age. CONCLUSION: Symptomatic perinatal COVID-19 had a substantial effect on maternal health but no apparent short-term effect on newborns.

Neighborhood Characteristics and Racial Disparities in Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Seropositivity in Pregnancy.

OBJECTIVE: To quantify the extent to which neighborhood characteristics contribute to racial and ethnic disparities in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) seropositivity in pregnancy. METHODS: This cohort study included pregnant patients who presented for childbirth at two hospitals in Philadelphia, Pennsylvania from April 13 to December 31, 2020. Seropositivity for SARS-CoV-2 was determined by measuring immunoglobulin G and immunoglobulin M antibodies by enzyme-linked immunosorbent assay in discarded maternal serum samples obtained for clinical purposes. Race and ethnicity were self-reported and abstracted from medical records. Patients' residential addresses were geocoded to obtain three Census tract variables: community deprivation, racial segregation (Index of Concentration at the Extremes), and crowding. Multivariable mixed effects logistic regression models and causal mediation analyses were used to quantify the extent to which neighborhood variables may explain racial and ethnic disparities in seropositivity. RESULTS: Among 5,991 pregnant patients, 562 (9.4%) were seropositive for SARS-CoV-2. Higher seropositivity rates were observed among Hispanic (19.3%, 104/538) and Black (14.0%, 373/2,658) patients, compared with Asian (3.2%, 13/406) patients, White (2.7%, 57/2,133) patients, and patients of another race or ethnicity (5.9%, 15/256) (P<.001). In adjusted models, per SD increase, deprivation (adjusted odds ratio [aOR] 1.16, 95% CI 1.02-1.32) and crowding (aOR 1.15, 95% CI 1.05-1.26) were associated with seropositivity, but segregation was not (aOR 0.90, 95% CI 0.78-1.04). Mediation analyses revealed that crowded housing may explain 6.7% (95% CI 2.0-14.7%) of the Hispanic-White disparity and that neighborhood deprivation may explain 10.2% (95% CI 0.5-21.1%) of the Black-White disparity. CONCLUSION: Neighborhood deprivation and crowding were associated with SARS-CoV-2 seropositivity in pregnancy in the prevaccination era and may partially explain high rates of SARS-CoV-2 seropositivity among Black and Hispanic patients. Investing in structural neighborhood improvements may reduce inequities in viral transmission.

Supplementation with a probiotic mixture accelerates gut microbiome maturation and reduces intestinal inflammation in extremely preterm infants.

Probiotics are increasingly administered to premature infants to prevent necrotizing enterocolitis and neonatal sepsis. However, their effects on gut microbiome assembly and immunity are poorly understood. Using a randomized intervention trial in extremely premature infants, we tested the effects of a probiotic product containing four strains of Bifidobacterium species autochthonous to the infant gut and one Lacticaseibacillus strain on the compositional and functional trajectory of microbiome. Daily administration of the mixture accelerated the transition into a mature, term-like microbiome with higher stability and species interconnectivity. Besides infant age, Bifidobacterium strains and stool metabolites were the best predictors of microbiome maturation, and structural equation modeling confirmed probiotics as a major determinant for the trajectory of microbiome assembly. Bifidobacterium-driven microbiome maturation was also linked to an anti-inflammatory intestinal immune milieu. This demonstrates that Bifidobacterium strains are ecosystem engineers that lead to an acceleration of microbiome maturation and immunological consequences in extremely premature infants.

Early childhood antibiotic utilization for infants discharged from the neonatal intensive care unit.

OBJECTIVE: To determine antibiotic utilization for NICU infants, as compared to non-NICU infants, in the first 3 years after birth hospital discharge. STUDY DESIGN: Retrospective observational study using data from Medicaid Analytic Extract including 667 541 newborns discharged from 2007-2011. Associations between NICU admission and antibiotic prescription were assessed using regression models, adjusting for confounders, and stratified by gestational age and birth weight. RESULTS: 596 999 infants (89.4%) received ≥1 antibiotic, with a median of 4 prescriptions per 3 person-years (IQR 2-8). Prescribed antibiotics and associated indication were similar between groups. Compared to non-NICU infants (N = 586 227), NICU infants (N = 81 314) received more antibiotic prescriptions (adjusted incidence rate ratio 1.08, 95% confidence interval [CI] (1.08,1.08)). Similar results were observed in all NICU subgroups. CONCLUSIONS: Antibiotic utilization in early childhood was higher among infants discharged from NICUs compared to non-NICU infants.

Group B Streptococcus Infection in Extremely Preterm Neonates and Neurodevelopmental Outcomes at 2 Years.

BACKGROUND: This study was performed to determine the incidence of group B Streptococcus (GBS) disease among extremely preterm infants and assess to risk of death or neurodevelopmental impairment (NDI) at a corrected age of 18-26 months. METHODS: In this observational cohort study of infants enrolled in a multicenter registry, the incidence of GBS disease was assessed in infants born in 1998-2016 at 22-28 weeks' gestation and surviving for >12 hours. The composite outcome, death or NDI, was assessed in infants born in 1998-2014 at 22-26 weeks' gestation. Infection was defined as GBS isolation in blood or cerebrospinal fluid culture at ≤72 hours (early-onset disease [EOD]) or >72 hours (late-onset disease [LOD]) after birth. Using Poisson regression models, the outcome was compared in infants with GBS disease, infants infected with other pathogens, and uninfected infants. RESULTS: The incidence of GBS EOD (2.70/1000 births [95% confidence interval (CI), 2.15-3.36]) and LOD (8.47/1000 infants [7.45-9.59]) did not change significantly over time. The adjusted relative risk of death/NDI was higher among infants with GBS EOD than in those with other infections (adjusted relative risk, 1.22 [95% CI, 1.02-1.45]) and uninfected infants (1.44 [1.23-1.69]). Risk of death/NDI did not differ between infants with GBS LOD and comparator groups. GBS LOD occurred at a significantly later age than non-GBS late-onset infection. Among infants surviving >30 days, the risk of death was higher with GBS LOD (adjusted relative risk, 1.90 [95% CI, 1.36-2.67]), compared with uninfected infants. CONCLUSIONS: In a cohort of extremely preterm infants, the incidence of GBS disease did not change during the study period. The increased risk of death or NDI with GBS EOD, and of death among some infants with GBS LOD, supports the need for novel preventive strategies for disease reduction. CLINICAL TRIALS REGISTRATION: NCT00063063.

Time to positivity of blood cultures in neonatal late-onset bacteraemia.

OBJECTIVE: To determine the time to positivity (TTP) of blood cultures among infants with late-onset bacteraemia and predictors of TTP >36 hours. DESIGN: Retrospective cohort study. SETTING: 16 birth centres in two healthcare systems. PATIENTS: Infants with positive blood cultures obtained >72 hours after birth. OUTCOME: The main outcome was TTP, defined as the time interval from specimen collection to when a neonatal provider was notified of culture growth. TTP analysis was restricted to the first positive culture per infant. Patient-specific and infection-specific factors were analysed for association with TTP >36 hours. RESULTS: Of 10 235 blood cultures obtained from 3808 infants, 1082 (10.6%) were positive. Restricting to bacterial pathogens and the first positive culture, the median TTP (25th-75th percentile) for 428 cultures was 23.5 hours (18.4-29.9); 364 (85.0%) resulted in 36 hours. Excluding coagulase-negative staphylococci (CoNS), 275 of 294 (93.5%) cultures were flagged positive by 36 hours. In a multivariable model, CoNS isolation and antibiotic pretreatment were significantly associated with increased odds of TTP >36 hours. Projecting a 36-hour empiric duration at one site and assuming that all negative evaluations were associated with an empiric course of antibiotics, we estimated that 1164 doses of antibiotics would be avoided in 629 infants over 10 years, while delaying a subsequent antibiotic dose in 13 infants with bacteraemia. CONCLUSIONS: Empiric antibiotic administration in late-onset infection evaluations (not targeting CoNS) can be stopped at 36 hours. Longer durations (48 hours) should be considered when there is pretreatment or antibiotic therapy is directed at CoNS.