Clinical Decision Support for Improved Neonatal Care: The Development of a Machine Learning Model for the Prediction of Late-onset Sepsis and Necrotizing Enterocolitis.

OBJECTIVE: To develop an artificial intelligence-based software system for predicting late-onset sepsis (LOS) and necrotizing enterocolitis (NEC) in infants admitted to the neonatal intensive care unit (NICU). STUDY DESIGN: Single-center, retrospective cohort study, conducted in the NICU of the Antwerp University Hospital. Continuous monitoring data of 865 preterm infants born at <32 weeks gestational age, admitted to the NICU in the first week of life, were used to train an XGBoost machine learning (ML) algorithm for LOS and NEC prediction in a cross-validated setup. Afterward, the model's performance was assessed on an independent test set of 148 patients (internal validation). RESULTS: The ML model delivered hourly risk predictions with an overall sensitivity of 69% (142/206) for all LOS/NEC episodes and 81% (67/83) for severe LOS/NEC episodes. The model showed a median time gain of ≤10 hours (IQR, 3.1-21.0 hours), compared with historical clinical diagnosis. On the complete retrospective dataset, the ML model made 721 069 predictions, of which 9805 (1.3%) depicted a LOS/NEC probability of ≥0.15, resulting in a total alarm rate of <1 patient alarm-day per week. The model reached a similar performance on the internal validation set. CONCLUSIONS: Artificial intelligence technology can assist clinicians in the early detection of LOS and NEC in the NICU, which potentially can result in clinical and socioeconomic benefits. Additional studies are required to quantify further the effect of combining artificial and human intelligence on patient outcomes in the NICU.

Placing Ureaplasma within the Context of Bronchopulmonary Dysplasia Endotypes and Phenotypes.

Different pathophysiological pathways (endotypes), leading to very preterm birth may result in distinct clinical phenotypes of bronchopulmonary dysplasia (BPD). Ureaplasma is a unique player in the pathogenesis of BPD. The interaction between factors inherent to Ureaplasma (virulence, bacterial load, duration of exposure), and to the host (immune response, infection clearance, degree of prematurity, respiratory support, concomitant infections) may contribute to BPD development in a variable manner. The data reviewed herein support the hypothesis that Ureaplasma, as a representative of the infectious/inflammatory endotype, may produce pulmonary damage predominantly in parenchyma, interstitium, and small airways. In contrast, Ureaplasma may have a very limited role in the pathogenesis of the vascular phenotype of BPD. In addition, if Ureaplasma is a key factor in BPD pathogenesis, its eradication by macrolides should prevent BPD. However, various meta-analyses do not show consistent evidence that this is the case. The limitations of current definitions and classifications of BPD, based on respiratory support needs instead of pathophysiology and phenotypes, may explain this and other failures in strategies aimed to prevent BPD. The precise mechanisms through which Ureaplasma infection leads to altered lung development and how these pathways can result in different BPD phenotypes warrant further investigation.

Pertussis Immunization During Pregnancy: Assessment of the Role of Maternal Antibodies on Immune Responses in Term and Preterm-Born Infants.

BACKGROUND: Limited data exist on the impact of maternal tetanus, diphtheria, acellular pertussis (Tdap) vaccination for preterm born infants. We report its effect at birth and on antibody-mediated immune responses to a DTaP-IPV-HB-PRP~T vaccine in preterm compared with term infants. METHODS: Women delivering at term or prematurely were either vaccinated with a Tdap vaccine (Boostrix; GSK) during pregnancy or not vaccinated in the last 5 years. Cord and maternal blood were collected at delivery. Infants were vaccinated with DTaP-IPV-HB-PRP~T vaccine (Hexyon; Sanofi Pasteur) and blood collected before and 1 month after primary (8-12-16 weeks) and before and 1 month after booster vaccination (13 or 15 months for preterm and term, respectively). Immunoglobulin G antibodies against all antigens included in DTaP-IPV-HB-PRP~T vaccine were measured (NCT02511327). RESULTS: Cord blood geometric mean concentrations (GMCs) in preterm infants from Tdap-vaccinated women were significantly higher than in term and preterm infants from unvaccinated women. A longer time interval between maternal vaccination and delivery resulted in higher cord blood GMCs in preterm infants. Equal GMCs in term and preterm infants from Tdap-vaccinated women were observed after primary vaccination. After boosting, significantly lower GMCs were seen for pertussis toxin, filamentous hemagglutinin, and tetanus toxoid in preterm compared with term infants from Tdap-vaccinated women, yet still comparable to GMCs in both term and preterm infants from unvaccinated women. CONCLUSIONS: Preterm infants profit from maternal Tdap vaccination. Prematurity did not influence primary immune responses in the presence of maternal antibodies but was associated with a lower booster immune response.

Cranial ultrasound and MRI: complementary or not in the diagnostic assessment of children with congenital CMV infection?

Whether or not cranial ultrasound (crUS) and cerebral magnetic resonance imaging (MRI) have both a place in the assessment of children with congenital cytomegalovirus infection (cCMV) remains a topic of discussion between research groups. Literature suggests that MRI is indicated only in children with abnormal crUS.In Flanders, Belgium, combined crUS and MRI was performed on 639 children with cCMV, referred for diagnostic assessment. Cranial US was classified as abnormal in the presence of striatal vasculopathy, calcifications, cysts, cystic germinolysis, and/or ventriculomegaly. MRI findings were classified as abnormal in the presence of gyration disorders, cerebellar abnormalities, ventriculomegaly, cysts, or pathologic white matter lesions.One in five children (93/480) with normal crUS showed abnormal findings on MRI. Of them, 85 (91.4%) were classified as symptomatic. In 37 of those 93 children (39.8%), classification as severely symptomatic was made based on MRI lesions alone. MRI and crUS proved to be complementary in the assessment of CNS involvement in children with cCMV. Long-term studies are needed to evaluate the importance of this finding with respect to outcome and benefit of therapy in this particular subgroup of patients with cCMV infection.Conclusion: Our findings support an enhanced role of MRI in the diagnosis of CNS involvement in children with cCMV infection. The ideal assessment should include both imaging techniques, as the strengths of each test compensate for the other's weaknesses. What is Known: • Congenital CMV infection involves the central nervous system with direct injury to and possible disruption of brain development. • Experts suggest that MRI is indicated only in children with abnormal crUS. What is New: • In almost 20% of our children with a normal cranial ultrasound, abnormalities were detected on MRI. • Our results suggest that performing both MRI and cranial US is important to obtain a complete assessment of central nervous system involvement in children with cCMV.

Impact of Maternal Pertussis Antibodies on the Infants' Cellular Immune Responses.

INTRODUCTION: Maternal antibody interference of the infant's humoral immune responses raises some concern to the strategy of maternal Tdap (tetanus, diphtheria, acellular pertussis [aP]) vaccination. This study assessed the impact of maternal Tdap antibodies on the infant's pertussis-specific T lymphocyte responses following infant vaccination with an aP containing vaccine, in a term and preterm born cohort. METHODS: Heparin samples (±0.5 mL) were conveniently drawn from infants of a Belgian prospective cohort study (N = 79, NCT02511327), including Tdap vaccinated (Boostrix®) and nonvaccinated women (no Tdap vaccine in the last 5 years) that delivered at term or prematurely. Sampling was performed before and 1 month after primary (8-12-16 weeks) and booster vaccination (13 or 15 months) with DTaP-IPV-HB-PRP~T vaccine (Hexyon®). Pertussis toxin (PT)-specific CD3+, CD3+ CD4+ and CD3+ CD8+ lymphoblasts and their cytokine secretions were measured using a flow cytometric assay on whole blood (FASCIA) and multiplex technology (Meso Scale Discovery), respectively. RESULTS: In total, 57% of all infants were considered PT-specific CD3+ CD4+ lymphoblasts responders after primary and booster vaccination, whereas 17% were CD3+ CD8+ lymphoblast responders. Interferon (IFN)-γ, interleukin (IL)-13, IL-17A, and IL-5 cytokine secretions after primary and booster vaccination were indicative of a mixed T helper (Th) 1/Th2/Th17 cell profile. Lymphoblast and cytokine levels were comparable between term and preterm infants. Nonresponders for IL-13 after booster vaccination had higher maternal PT immunoglobulin G (IgG) levels at birth when compared to responders. CONCLUSIONS: Term and preterm born infants are capable of inducing Th1, Th2, and Th17 responses after aP vaccination, yet maternal vaccination modulate these responses. Evaluation of this effect in larger trials is needed.

Breast Milk Antibody Levels in Tdap-Vaccinated Women After Preterm Delivery.

BACKGROUND: Enrichment of breast milk (BM) with immunoglobulin (Ig) A and IgG through maternal vaccination could help infants combat targeted pathogens. However, evidence on this effect after preterm delivery is lacking. In this study, we investigated the total and anti-pertussis toxin (anti-PT)-specific IgA and IgG production in BM after term and preterm delivery in the presence of maternal Tdap (tetanus, diphtheria, acellular pertussis) vaccination. METHODS: Serum and BM samples of lactating women who delivered at term or prematurely and did or did not receive Tdap vaccine (Boostrix, GSK Biologicals) during pregnancy were collected as part of a clinical study (N = 234). Anti-PT IgA/IgG (IBL assay; Meso Scale Discovery assay) and total IgA/IgG (Thermofisher, on BM samples only) immunosorbent assays were performed on all samples collected at 72 hours and 4, 8, and 12 weeks postpartum. RESULTS: BM after preterm delivery contained anti-PT IgA and IgG geometric mean concentrations (GMCs) comparable to those after term delivery (eg, colostrum anti-PT IgA, 5.39 IU/mL vs 6.69 IU/mL, respectively). Maternal Tdap vaccination induced significantly higher anti-PT IgG GMCs in colostrum of vaccinated compared with unvaccinated women who delivered at term (0.110 IU/mL vs 0.027 IU/mL, P = .009). Anti-PT antibodies persisted up to 12 weeks postpartum. CONCLUSIONS: This study provides evidence that maternal Tdap vaccination induces high Ig levels in BM after both term and preterm delivery and that these antibodies remain abundantly present throughout lactation, possibly offering additional mucosal protection during the most vulnerable period in early life. CLINICAL TRIAL REGISTRATION: NCT02511327.

Association between maternal cervicovaginal swab positivity for Ureaplasma spp. or other microorganisms and neonatal respiratory outcome and mortality.

OBJECTIVE: We investigated the association between maternal cervicovaginal cultures, its antibiotic treatment, and neonatal outcome. STUDY DESIGN: This retrospective cohort study enrolled 480 neonates born prior to 32 weeks' gestation. They were divided into groups according to maternal cervicovaginal culture results. Multivariate logistic regression analysis was used to predict neonatal outcome based on maternal culture results, adjusted for perinatal risk factors and neonatal morbidities. RESULT: Maternal cervicovaginal Ureaplasma colonization was independently associated with bronchopulmonary dysplasia at 36 weeks (BPD) (OR 8.34; 95% CI 1.21-57.45). In neonates with and without maternal cervicovaginal Ureaplasma colonization BPD occurred in 12.3% and 3.8%, respectively. Maternal colonization with other microorganisms was associated with a higher neonatal mortality (p = 0.002), lower gestational age (p = 0.026), and birth weight (p = 0.036). CONCLUSIONS: This study underscores the role of the maternal cervicovaginal microbiome as a predictor of neonatal outcome. Cervicovaginal Ureaplasma colonization seems not to be an innocent bystander in the multifactorial etiology of BPD.

A new technique for difficult removal of a peripherally inserted central venous catheter (PICC) in a neonate.

In general, the removal of peripherally inserted central venous catheters (PICC) in neonates by gentle traction is easy. In our case, the removal of a 28G PICC in a term neonate was impossible by manual traction even with force. Previously described non-invasive interventions using a stylet were not successful because it was not possible to pass the stylet along the catheter hub of the narrow 28G PICC. In the end, the catheter could be removed non-operatively by cutting the catheter just distal to the hub and inserting a stylet of a new PICC (same brand and size) into the patients' retained catheter. Subsequently, the force of manual traction on the catheter could be increased without increased risk of catheter stretching and breakage. After catheter removal, the surface of the remaining PICC was intact.Conclusion: By thinking outside the box, surgical intervention was prevented in this neonate. What is Known: • On general, peripherally inserted central venous catheters (PICC) can be removed easily by gentle traction. • There are no clear recommendations about what to do if standard interventions fail to remove a PICC. What is New: • Our technique is a non-invasive option for difficult PICC removal and can prevent surgery. • The retained PICC is cut distal to the hub, and after stylet reinsertion, sustained manual traction is performed.

Functional respiratory imaging provides novel insights into the long-term respiratory sequelae of bronchopulmonary dysplasia.

RATIONALE: Bronchopulmonary dysplasia (BPD) is a common complication of preterm birth. Lung function and imaging are classically used to assess BPD. Functional respiratory imaging (FRI) combines a structural and functional assessment of the airways and their vasculature. We aimed to assess BPD using FRI and to correlate these findings with the clinical presentation. METHODS: We included 37 adolescents with a history of preterm birth (22 BPD cases and 15 preterm controls). The study protocol included a detailed history, lung function testing and computed tomography (CT) (at total lung capacity (TLC) and functional residual capacity (FRC)) with FRI. CT images were also assessed using the Aukland scoring system. RESULTS: BPD patients had lower forced expiratory volume in 1 s to forced vital capacity ratio (p=0.02) and impaired diffusion capacity (p=0.02). Aukland CT scores were not different between the two groups. FRI analysis showed higher lobar volumes in BPD patients at FRC (p<0.01), but not at TLC. Airway resistance was significantly higher in the BPD group, especially in the distal airways. Additionally, FRI showed more air trapping in BPD patients, in contrast to findings on conventional CT images. CONCLUSION: This study is the first to use FRI in research for BPD. FRI analysis showed higher lobar volumes in BPD patients, indicating air trapping and reduced inspiratory capacity. In contrast to Aukland CT scores, FRI showed more air trapping in the BPD group, suggesting that FRI might be a more sensitive detection method. Importantly, we also showed increased distal airway resistance in BPD patients. By combining structural and functional assessment, FRI may help to better understand the long-term sequelae of BPD.

Machine Learning to Support Hemodynamic Intervention in the Neonatal Intensive Care Unit.

Hemodynamic support in neonatal intensive care is directed at maintaining cardiovascular wellbeing. At present, monitoring of vital signs plays an essential role in augmenting care in a reactive manner. By applying machine learning techniques, a model can be trained to learn patterns in time series data, allowing the detection of adverse outcomes before they become clinically apparent. In this review we provide an overview of the different machine learning techniques that have been used to develop models in hemodynamic care for newborn infants. We focus on their potential benefits, research pitfalls, and challenges related to their implementation in clinical care.

Prospective multicenter comparison of urine culture with PCR on dried blood spots using 2 different extraction and PCR methods in neonates suspected for congenital cytomegalovirus infection.

To evaluate the potential of dried blood spots (DBS) as a congenital cytomegalovirus (cCMV) testing specimen, the laboratory diagnostic accuracy of polymerase chain reaction (PCR) on DBS was compared to viral urine cultures from neonates suspected for cCMV. Two different extraction methods (EasyMAG, bioMérieux versus Qiagen) and 2 real-time PCR protocols (in-house versus Argene) were compared. We were able to collect both DBS and urine samples in 6 Belgian neonatal units from 276 neonates suspected for cCMV registered in CMVREG (an online neonatal registry system). Forty-eight neonates (17.4%) were positive by viral culture in urine. Laboratory diagnostic accuracy parameters of DBS-PCR were both extraction method and PCR protocol dependent. Not all DBS-CMV-PCR methods successfully detected urine-culture-positive neonates born after first-trimester seroconversions. Interestingly, however, all urine-culture-positive neonates having clinical signs of cCMV did consistently score positive.

Blood culture indications in critically ill neonates: a multicenter prospective cohort study.

Due to potential lethality of healthcare-associated sepsis (HAS), a low threshold for blood culturing and antimicrobial therapy (ABT) initiation is accepted. We assessed variability in the trigger for blood culturing between three neonatal intensive care units. A multicenter prospective cohort study was conducted. In newborns with suspicion of HAS, 10 predefined clinical signs, nosocomial sepsis (NOSEP) score, C-reactive protein, ABT initiation, and risk factors were registered at time of culturing. Outcome was lab-confirmed HAS, defined according to the NeoKISS-criteria. Two hundred ninety-nine suspected HAS episodes were considered in 212 infants, of which 118 had birth-weight ≤ 1500 g; proportion of lab-confirmed HAS per suspected episode was 30/192 (center 1), 28/60 (center 2), and 8/47 (center 3) (p < 0.001). Median C-reactive protein and number of clinical signs at time of culturing differed between centers 1, 2, and 3 (respectively 11 vs. 5 vs. 3 mg/L, p = 0.001; 1 sign [IQR 0-2, center 1] vs. 3 signs [IQR 2-4, centers 2 and 3], p < 0.001). Median NOSEP score at time of culturing was 5 (IQR 3-8, center 1), 5 (IQR 3-9, center 2), and 8 (IQR 5-11, center 3) (p = 0.016). Difference in ABT initiation was noticed (82 vs. 93 vs. 74%, p = 0.05). CONCLUSION: Center heterogeneity in sampling practice is substantial. Optimizing sampling practice can be recommended. What is Known: • Blood culture test is a common diagnostic procedure in critically-ill newborns. • A low threshold for sampling and antimicrobial therapy initiation is accepted. What is New: • Variability in blood culture practice was assessed between 3 neonatal intensive care units by the registration of sampling frequencies, clinical indications, and antimicrobial therapy initiation.

Impact of healthcare-associated sepsis on mortality in critically ill infants.

UNLABELLED: Healthcare-associated sepsis (HAS) is a life-threatening complication in neonatal intensive care. Research into the impact of HAS on mortality adjusted for comorbidities is however limited. We conducted a historical cohort study to evaluate impact of HAS on mortality stratified by birth weight and risk factors for mortality in the HAS cohort. HAS was defined according to the National Institute of Child Health and Human Development criteria. Logistic regression was used to calculate adjusted odds of mortality. Of 5134 admissions, 342 infants developed HAS (6.7 %). Mortality in the total and HAS cohort was 5.6 and 10.5 %, respectively. The majority of HAS was caused by commensals (HAS-COM, 59.4 %) and 40.6 % by recognized pathogens (HAS-REC). Adjusted for comorbidities, "HAS-REC" is only a risk factor for mortality in newborns >1500 g (adjusted odds ratio [aOR] 2.3, confidence interval [CI] 1.1-4.9). Post-hoc analysis identified HAS-REC as an independent risk factor for mortality in infants with gastrointestinal disease (aOR 4.8, CI 2.1-10.8). "Renal insufficiency," "focal intestinal perforation," and "necrotizing enterocolitis" are independent risk factors for mortality in the HAS cohort (aOR 13.5, CI 4.9-36.6; aOR 7.7, CI 1.5-39.2; aOR 2.1, CI 1.0-4.7, respectively). CONCLUSION: For very low birth weight infants (≤1500 g), several comorbidities overrule the impact of HAS on mortality. After adjustment for comorbidities, HAS-REC independently predicts in-hospital mortality in heavier infants and in those with gastrointestinal disease. WHAT IS KNOWN: • The relationship between healthcare-associated sepsis and mortality is influenced by the causative pathogen and is confounded by comorbidities. • Research on impact of healthcare-associated sepsis on mortality adjusted for comorbidities is limited as well as research on independent risk factors for mortality in neonates with sepsis. What is New: • We included a large list of comorbidities and stratified risk by birth weight in order to assess the true effect of healthcare-associated sepsis on mortality. • Risk for mortality was calculated for commensal flora and for recognized pathogens as causative micro-organisms.

A rare infectious cause of severe neonatal skin lesions.

We present a case of a preterm infant of 28 weeks' gestation with unique cutaneous lesions characteristic of a congenital herpes simplex virus (HSV) type 1 infection. The infant was prematurely delivered due to intractable labour. The mother had no history or clinical signs of genital infection before or during pregnancy. The infant's skin lesions were described as rough white-yellow plaques; a skin biopsy demonstrated calcified plaques and absent epidermis. HSV type 1 was later determined using PCR on the infant's skin biopsy and cerebral spinal fluid as well as the mother's vaginal swab and the placenta. Calcifications have already been described by Allee et al, alongside a diagnosis of HSV type 2. As is well known, the morbidity and mortality of congenital herpes infections are very high.

Prediction models for neonatal health care-associated sepsis: a meta-analysis.

BACKGROUND AND OBJECTIVES: Blood culture is the gold standard to diagnose bloodstream infection but is usually time-consuming. Prediction models aim to facilitate early preliminary diagnosis and treatment. We systematically reviewed prediction models for health care-associated bloodstream infection (HABSI) in neonates, identified superior models, and pooled clinical predictors. DATA SOURCES: LibHub, PubMed, and Web of Science. METHODS: The studies included designed prediction models for laboratory-confirmed HABSI or sepsis. The target population was a consecutive series of neonates with suspicion of sepsis hospitalized for ≥ 48 hours. Clinical predictors had to be recorded at time of or before culturing. Methodologic quality of the studies was assessed. Data extracted included population characteristics, total suspected and laboratory-confirmed episodes and definition, clinical parameter definitions and odds ratios, and diagnostic accuracy parameters. RESULTS: The systematic search revealed 9 articles with 12 prediction models representing 1295 suspected and 434 laboratory-confirmed sepsis episodes. Models exhibit moderate-good methodologic quality, large pretest probability range, and insufficient diagnostic accuracy. Random effects meta-analysis showed that lethargy, pallor/mottling, total parenteral nutrition, lipid infusion, and postnatal corticosteroids were predictive for HABSI. Post hoc analysis with low-gestational-age neonates demonstrated that apnea/bradycardia, lethargy, pallor/mottling, and poor peripheral perfusion were predictive for HABSI. Limitations include clinical and statistical heterogeneity. CONCLUSIONS: Prediction models should be considered as guidance rather than an absolute indicator because they all have limited diagnostic accuracy. Lethargy and pallor and/or mottling for all neonates as well as apnea and/or bradycardia and poor peripheral perfusion for very low birth weight neonates are the most powerful clinical signs. However, the clinical context of the neonate should always be considered.

Implementation of a multi-parameter Point-of-Care-blood test analyzer reduces central laboratory testing and need for blood transfusions in very low birth weight infants.

Blood sampling for laboratory testing is a major cause of iatrogenic blood loss and anemia in neonatal intensive care unit [NICU] patients. The objective of the study was to assess whether the implementation of a multi-parameter Point of Care Test [POCT] (Roche, Cobas b221) analyzer affected blood loss for central laboratory testing and need for red blood cell transfusion in our NICU. This was a retrospective observational cohort study in a NICU with comparison of two serial cohorts of 2 years each. Implementation of a multi-parameter POCT decreased central laboratory performed testing for bilirubin (-32% per patient) and electrolytes (-36% per patient). On average, the net blood volume taken per admitted patient for electrolyte testing decreased with 23.7% and 22.2% for bilirubin testing in the second cohort. After implementation of POCT, fewer very low birth weight infants [VLBWI] required blood transfusion (38.9% vs. 50%, p<.05) as the number of transfusion/infants decreased by 48% (1.57 vs. 2.53, p<0.01). The implementation of POCT was cost-efficient for the Belgian national health insurance, cost reduction -8.3% per neonate. We conclude that implementation of a bedside multi-parameter POCT analyzer decreases transfusions among VLBWI by reducing iatrogenic blood loss for central laboratory testing.

Time to positivity of neonatal blood cultures: fast and furious?

The aim of this study was to determine the time to positivity (TTP) of neonatal blood cultures, to investigate differences between early onset versus late-onset sepsis, and non-proven versus proven sepsis, and to examine differences in TTP by organism type using a retrospective observational study at the Neonatal Intensive Care Unit, Antwerp University Hospital, Belgium. The subjects were 1828 neonates with suspected sepsis who were treated with antimicrobials for at least 3 days. The TTP was recorded for all episodes of suspected sepsis in an approximately 6.5 year period. A total of 2916 blood cultures were collected, of which 437 (15%) became positive. The overall TTP was 21.33 h (Q1-Q3 13.17-32.46). The difference between the median TTP in early onset versus late-onset sepsis was 0.83 h (22.00 versus 21.17 h, P=0.75). The median TTP for Gram-negative organisms was 11.17 h (Q1-Q3 8.84-15.67), whereas the median TTP for Gram-positive organisms was 23.59 h (Q1-Q3 15.29-34.58, P<0.001). In Gram-positive isolates, the median TTP for coagulase-negative staphylococci (CNS) was 26.67 h (Q1-Q3 19.00-38.17), whereas the median TTP for non-CNS was 12.83 h (Q1-Q3 10.50-18.17, P<0.001). The median TTP in proven sepsis was 20.17 h (Q1-Q3 13.00-30.37), whereas it was 29.67 h (Q1-Q3 21.17-50.63, P<0.001) in non-proven sepsis. TTP of neonatal blood cultures was significantly shorter for Gram-negative organisms. We suggest shortening the total incubation time of neonatal blood cultures to a maximum of 3 days. However, blood cultures collected in infants<72 h of age might require a longer incubation time. According to our results, it may be safe to narrow the antimicrobial spectrum to solely target Gram-positive bacteria when the culture is still negative after 48 h, and to cease antimicrobial therapy when the culture is still negative after 72 h in clinically well infants.

Number of sites of perinatal Candida colonization and neutropenia are associated with nosocomial candidemia in the neonatal intensive care unit patient.

OBJECTIVES: To determine the role of perinatally acquired Candida colonization to invasive Candida infection (candidemia) and to assess risk factors associated with Candida colonization and candidemia in neonatal intensive care unit patients. DESIGN: Retrospective case-control study. SETTING: Neonatal intensive care unit of a teaching hospital. PATIENTS: A total of 39 of 3219 (1.2%) who were positive for Candida colonization at birth were compared with 117 noncolonized controls. INTERVENTIONS: Routine surveillance cultures for Candida of skin and meconium were performed at admission. All neonates with Candida colonization at birth during a 10-yr period were identified. Each case was matched to place of birth and date of admission with three noncolonized controls. MEASUREMENTS AND MAIN RESULTS: Perinatal and neonatal variables were collected. Blood or skin culture was obtained when signs of sepsis or dermatitis were present. Patients with Candida colonization were compared with their noncolonized controls, whereas in this cohort, patients with candidemia were compared with those without by multivariate analysis. Vaginal candidiasis (odds ratio [OR] 15.8, 95% confidence interval [CI] 2.63, 94.77), birth weight below 1000 g (OR 8.1, 95% CI 1.22, 52.26), and vaginal delivery (OR 7.08, 95% CI 1.17, 42.70) were associated with Candida colonization. An increased risk for nosocomial candidemia was independently associated with the number of sites of Candida colonization (OR 24.02, 95% CI 1.89, 304), early neonatal neutropenia (OR 7.15, 95% CI 0.98, 80.95) and illness severity (clinical risk index for babies [CRIB]) score at day 1 (OR 1.38, 95%CI 1.065, 1.811). CONCLUSIONS: Maternal vaginal candidiasis and vaginal birth are risk factors for neonatal colonization. When controlling for illness severity, the number of sites colonized with Candida at birth contributes to neonatal nosocomial candidemia. Early neutropenia increases the risk further. These findings offer opportunities for prevention of Candida infection in neonatal intensive care unit patients.

Diagnosis of catheter-related bloodstream infection in neonates: a study on the value of differential time to positivity of paired blood cultures.

OBJECTIVE: Diagnosis of neonatal catheter-related bloodstream infection (CRBSI) is currently based on isolation of identical bacterial species from bloodstream and catheter tip cultures. This requires removal of the catheter followed by the insertion of a new catheter. The objective of this study was to investigate whether differential time to positivity (DTP) of blood cultures drawn from paired peripheral vein and central vascular catheter is useful for diagnosing neonatal CRBSI, avoiding removal of the catheter. DESIGN: Retrospective observational study. SETTING: Neonatal intensive care unit, University Hospital of Antwerp, Belgium. PATIENTS: Neonates with probable and definite nosocomial bloodstream infection. INTERVENTIONS: All episodes of nosocomial bloodstream infection (NBSI) in an approximately 7.5-yr period were identified retrospectively. Definite NBSI episodes in which paired blood cultures were obtained were retained to calculate DTP, to determine the optimal DTP cutoff for the diagnosis of CRBSI, and to assess the validity of DTP for the diagnosis of CRBSI. MEASUREMENTS AND MAIN RESULTS: Of 32 NBSI episodes included in the study, 16 were CRBSI, seven were non-CRBSI, and nine were classified as "diagnosis uncertain." In CRBSI, blood cultures drawn from a central vascular catheter were positive earlier than those drawn from a peripheral vein (median 9.67 hrs vs. 21.58 hrs, p < .01). Median DTP was 10.42 hrs in CRBSI and -0.33 hrs in non-CRBSI (p = .01). The optimal DTP cutoff for the diagnosis of CRBSI was > or =1 hr (area under the receiver operating characteristic curve = 0.84 +/- 0.11), with a sensitivity of 94%, a specificity of 71%, a positive predictive value of 88%, and a negative predictive value of 83%. CONCLUSIONS: Differential time to positivity of paired blood cultures may have some potential in the diagnosis of catheter-related infections in neonatal intensive care unit patients and should be subjected to a prospective study.

High levels of CXCL8 in tracheal aspirate samples taken at birth are associated with adverse respiratory outcome only in preterm infants younger than 28 weeks gestation.

We investigated the relation between perinatal endotracheal colonization, the associated cytokine response and respiratory outcome in ventilated preterm neonates. Between September 1999 and March 2002, a cohort of 141 neonates with a gestational age <31 weeks requiring ventilation directly after birth, were followed prospectively. All were admitted to the Neonatal Intensive Care Unit, University Hospital of Antwerp, Belgium. A tracheal aspirate (TA) sample was collected soon after birth and was processed for microbiological examination, leukocyte count, and cytokine analysis (interleukins [IL] IL-1beta, IL-6, CXCL8 (formerly called IL-8), IL-10, IL-12p70 and tumor necrosis factor alpha [TNF-alpha]). Together with the prospectively registered patient's comorbidities and severity of disease, these inflammatory parameters were analyzed in a multivariate Cox proportional hazards model with time of extubation and duration of oxygen therapy as main outcome measures. Of the 141 patients included, 31 (22%) died before discharge from the unit and 37 (26%) had a positive TA culture. Independent predictors of duration of mechanical ventilation were: gestational age <28 weeks, degree of respiratory distress syndrome (RDS) at birth, significant patent ductus arteriosus (PDA), the SNAP-score, and high levels of CXCL8 (>4,153 pg/ml) in TA only in neonates with a gestational age <28 weeks. Variables associated with extended duration of oxygen therapy were gestational age <28 weeks, birth weight <1,000 g, degree of RDS at birth, and duration of mechanical ventilation.