Boldly going Introducing conflict management training to Starship Children's Hospital.

Conflict between health professionals and whanau (families) in paediatric hospitals is common and leads to significant distress for families and staff. The likelihood of challenges and conflict around communication and critical medical decision-making is increased where there are cultural and social complexities. Training staff to recognise and manage conflict as early as possible improves patient/whanau outcomes and staff well-being. This article describes an ongoing collaboration between Starship Children's Hospital in Auckland New Zealand (NZ) and the UK Medical Mediation Foundation (MMF) focused on educating staff in the early recognition and management of conflict using mediation skills. An evidence-based training programme and structured ongoing supervision of a small group of champions has enabled this training to be embedded into Starship clinical practice. The collaboration has included careful consideration of the New Zealand setting, ensuring that the content of the programme specifically addresses our unique cultural and social context. Mediation skills are an important step in ensuring that our patients and whanau feel heard, acknowledged, and respected, and contribute to the Starship Child Health's strategic priority of eliminating inequity and addressing institutional bias.

Transforming training into practice with the conflict management framework: a mixed methods study.

Objective: To implement and evaluate the use of the conflict management framework (CMF) in four tertiary UK paediatric services. Design: Mixed methods multisite evaluation including prospective pre and post intervention collection of conflict data alongside semistructured interviews. Setting: Eight inpatient or day care wards across four tertiary UK paediatric services. Interventions: The two-stage CMF was used in daily huddles to prompt the recognition and management of conflict. Results: Conflicts were recorded for a total of 67 weeks before and 141 weeks after implementation of the CMF across the four sites. 1000 episodes of conflict involving 324 patients/families across the four sites were recorded. After implementation of the CMF, time spent managing episodes of conflict around the care of a patient was decreased by 24% (p<0.001) (from 73 min to 55 min) and the estimated cost of this staff time decreased by 20% (p<0.02) (from £26 to £21 sterling per episode of conflict). This reduction occurred despite conflict episodes after implementation of the CMF having similar severity to those before implementation. Semistructured interviews highlighted the importance of broad multidisciplinary leadership and training to embed a culture of proactive and collaborative conflict management. Conclusions: The CMF offers an effective adjunct to conflict management training, reducing time spent managing conflict and the associated staff costs.

Reactivity of Routine HIV Antibody Tests in Children With Perinatally Acquired HIV-1 in England: Cross-sectional Analysis.

We assessed HIV antibody prevalence in children with perinatally acquired HIV in England. Eighteen percent (10/55) of those starting combination antiretroviral therapy <6 months of age were seronegative at median age 9.1 years and had lower viral load at diagnosis and combination antiretroviral therapy start and fewer viral rebounds, than 45 of 55 seropositives. Implications for patient selection for HIV cure research, and interpretation of routine antibody testing, are discussed.

Using a simple point-prevalence survey to define appropriate antibiotic prescribing in hospitalised children across the UK.

BACKGROUND: The National Health Service England, Commissioning for Quality and Innovation for Antimicrobial Resistance (CQUIN AMR) aims to reduce the total antibiotic consumption and the use of certain broad-spectrum antibiotics in secondary care. However, robust baseline antibiotic use data are lacking for hospitalised children. In this study, we aim to describe, compare and explain the prescription patterns of antibiotics within and between paediatric units in the UK and to provide a baseline for antibiotic prescribing for future improvement using CQUIN AMR guidance. METHODS: We conducted a cross-sectional study using a point prevalence survey (PPS) in 61 paediatric units across the UK. The standardised study protocol from the Antibiotic Resistance and Prescribing in European Children (ARPEC) project was used. All inpatients under 18 years of age present in the participating hospital on the day of the study were included except neonates. RESULTS: A total of 1247 (40.9%) of 3047 children hospitalised on the day of the PPS were on antibiotics. The proportion of children receiving antibiotics showed a wide variation between both district general and tertiary hospitals, with 36.4% ( 95% CI 33.4% to 39.4%) and 43.0% (95% CI 40.9% to 45.1%) of children prescribed antibiotics, respectively. About a quarter of children on antibiotic therapy received either a medical or surgical prophylaxis with parenteral administration being the main prescribed route for antibiotics (>60% of the prescriptions for both types of hospitals). General paediatrics units were surprisingly high prescribers of critical broad-spectrum antibiotics, that is, carbapenems and piperacillin-tazobactam. CONCLUSIONS: We provide a robust baseline for antibiotic prescribing in hospitalised children in relation to current national stewardship efforts in the UK. Repeated PPS with further linkage to resistance data needs to be part of the antibiotic stewardship strategy to tackle the issue of suboptimal antibiotic use in hospitalised children.

Schedules for Pneumococcal Vaccination of Preterm Infants: An RCT.

BACKGROUND AND OBJECTIVE: Premature infants have a higher risk of invasive pneumococcal disease and are more likely to have lower vaccine responses compared with term infants. Increasingly, immunization schedules are including a reduced, 2-dose, pneumococcal conjugate vaccine priming schedule. Our goal was to assess the immunogenicity of 3 commonly used 13-valent pneumococcal conjugate vaccine (PCV13) priming schedules in premature infants and their response to a 12-month booster dose. METHODS: Premature infants (<35 weeks' gestation) were randomized to receive PCV13 at 2 and 4 months (reduced schedule); 2, 3, and 4 months (accelerated schedule); or 2, 4, and 6 months (extended schedule). All infants received a 12-month PCV13 booster. Serotype-specific pneumococcal immunoglobulin G (IgG) for PCV13 serotypes was measured by using enzyme-linked immunosorbent assay 1 month after the primary and booster vaccinations. RESULTS: A total of 210 infants (median birth gestation, 29(+6) weeks; range, 23(+2)-34(+6) weeks) were included. After the primary vaccination, 75% (95% confidence interval [CI], 62-85), 88% (95% CI, 76-95), and 97% (95% CI, 87-99) of participants had protective antibody concentrations for at least one-half the PCV13 serotypes for the reduced, accelerated, and extended schedules, respectively. After the booster vaccination, participants receiving the extended schedule had significantly lower (P < .05) geometric mean concentrations compared with reduced (for 9 of 13 serotypes) and accelerated (for 4 of 13 serotypes) schedules, but nearly all participations, regardless of schedule or serotype, had seroprotective IgG concentrations. CONCLUSIONS: A reduced priming schedule of PCV13 resulted in higher post-booster IgG concentrations but lower post-primary concentrations. The optimum vaccine schedule for preterm infants will therefore depend on when they are most at risk for invasive pneumococcal disease.

Lymphocyte subpopulations in premature infants: an observational study.

BACKGROUND AND OBJECTIVES: The infant's immune system evolves over the first months and years of life. Strong correlation exists between lymphocyte count, lymphocyte subpopulations and gestational age at birth. Associations with antenatal and postnatal steroid treatment, infection and chronic lung disease have also been described. Few published studies report the effect of increasing postnatal age (PNA) and comorbidities on lymphocyte subpopulations in premature infants beyond the first 4 months of life. This study aimed to describe changes in lymphocyte subpopulations in preterm infants up to 13 months PNA. METHODS: Premature infants (23-34 weeks completed gestation) from five centres had lymphocyte subpopulations measured at 2, 5 or 7, 12 and 13 months PNA alongside their vaccine responses in a vaccination trial. RESULTS: 393 blood samples from 151 babies were analysed. There was an increase in absolute numbers of total lymphocytes (median cell count 6.21×109/L at 13 months compared with 4.9×109/L at 2 months PNA) and CD3+, CD4+, CD8+, natural killer and B cells with increasing age. At 2 months PNA, there was a positive correlation between gestation and CD3+ and CD4+ counts (r=0.32 and 0.46, respectively) and proportions (r=0.22 and 0.41, respectively), and CD4+:CD8+ ratios (r=0.57), but a negative correlation with CD8+ proportions (r=-0.32). CONCLUSIONS: This longitudinal study describes the distribution of lymphocyte subpopulations in premature infants and provides reference ranges for the major lymphocyte subsets to help guide clinicians when assessing premature infants for immunodeficiency in the first year of life. TRIAL REGISTRATION NUMBER: EudraCT 2007-007535-23.

Conflict in a paediatric hospital: a prospective mixed-method study.

BACKGROUND: Conflict in healthcare is a well-recognised but under-examined phenomenon. Little is known about the prevalence and causes of conflict across paediatric specialties. OBJECTIVE: To report the frequency and characteristics of conflict in a paediatric hospital. DESIGN AND SETTING: An explanatory sequential mixed-method approach was adopted. A bespoke questionnaire recorded frequency, severity, cause and staff involved in conflict prospectively. Data were recorded for the same two 12-week periods in 2013 and 2014, in one UK children's teaching hospital. Data were analysed using descriptive statistics and correlation, the findings of which informed the construction of a semistructured interview schedule. Qualitative interviews were conducted with six key informant healthcare professionals to aid data interpretation; interviews were analysed thematically. RESULTS: 136 individual episodes of conflict were reported. The three most common causes were 'communication breakdown', 'disagreements about treatment' and 'unrealistic expectations'. Over 448 h of healthcare professional time was taken up by these conflicts; most often staff nurses, consultants, doctors in training and matrons. The mean severity rating was 4.9 out of 10. Qualitative interviews revealed consensus regarding whether conflicts were ranked as low, medium or high severity, and explanations regarding why neurology recorded the highest number of conflicts in the observed period. CONCLUSIONS: Conflict is prevalent across paediatric specialties, and particularly in neurology, general paediatrics and neonatology. Considerable staff time is taken in managing conflict, indicating a need to focus resources on supporting staff to resolve conflict, notably managing communication breakdown.

Characteristics and outcomes of human parechovirus infection in infants (2008-2012).

UNLABELLED: Human parechoviruses (HPeVs) cause a spectrum of disease ranging from self-limiting illness to severe disease and, sometimes, death. We describe the clinical characteristics and outcomes of HPeV infection in infants. The study describes the clinical and laboratory characteristics and outcomes of infants with HPeV infection during 2008-2012, from three paediatric hospitals in London each with a paediatric intensive care unit. The infants were retrospectively identified through laboratory and patient discharge databases and diagnosed through HPeV PCR. Fifty infants were identified. Half required admission to PICU. Infants less than 3 months were more likely to require PICU (16/25: p < 0.01). Clinical signs at presentation were often indistinguishable from those of bacterial sepsis and meningitis, but inflammatory markers were nearly always (95 % of cases) within normal ranges. Brain MRI showed white matter changes in 10/12 infants. Three of 19 infants with follow-up data (16 %) had significant neurological sequelae. CONCLUSION: HPeV may cause severe disease and long-term neurological sequelae in young infants. HPeV should be considered in infants with clinical features of sepsis/meningitis with normal CSF microscopy. Prospective observational studies are warranted to better define the epidemiology of infection and thus inform future treatment trials.

Fifteen minute consultation: managing neonatal and childhood herpes encephalitis.

Herpes simplex encephalitis (HSE) is the most common single cause of viral encephalitis in infants and children. Treated or untreated, it can be associated with considerable morbidity and mortality, and its presentation is usually insidious and non-specific. Prompt and careful investigation is important in order to establish the diagnosis so that treatment can be optimised. We address some common questions arising when diagnosing and treating presumed HSE throughout childhood.

Long-term virological outcome in children on antiretroviral therapy in the UK and Ireland.

OBJECTIVE: To assess factors at the start of antiretroviral therapy (ART) associated with long-term virological response in children. DESIGN: Multicentre national cohort. METHODS: Factors associated with viral load below 400 copies/ml by 12 months and virologic failure among children starting 3/4-drug ART in the UK/Irish Collaborative HIV Paediatric Study were assessed using Poisson models. RESULTS: Nine hundred and ninety-seven children started ART at a median age of 7.7 years (inter-quartile range 2.9­11.7), 251 (25%) below 3 years: 411 (41%) with efavirenz and two nucleoside reverse transcriptase inhibitors (EFVþ2NRTIs), 264 (26%) with nevirapine and two NRTIs (NVPþ2NRTIs), 119 (12%; 106 NVP, 13 EFV) with non-nucleoside reverse transcriptase inhibitor and three NRTIs (NNRTIþ3NRTIs), and 203 (20%) with boosted protease inhibitor-based regimens. Median follow-up after ART initiation was 5.7 (3.0­8.8) years. Viral load was less than 400 copies/ml by 12 months in 92% [95% confidence interval (CI) 91­94%] of the children. Time to suppression was similar across regimens (P»0.10), but faster over calendar time, with older age and lower baseline viral load. Three hundred and thirtynine (34%) children experienced virological failure. Although progression to failure varied by regimen (P<0.001) and was fastest for NVPþ2NRTIs regimens, risk after 2 years on therapy was similar for EFVþ2NRTIs and NVPþ2NRTIs, and lowest for NNRTIþ3NRTIs regimens (P-interaction»0.03). Older age, earlier calendar periods and maternal ART exposure were associated with increased failure risk. Early treatment discontinuation for toxicity occurred more frequently for NVP-based regimens, but 5-year cumulative incidence was similar: 6.1% (95% CI 3.9­8.9%) NVP, 8.3% (95% CI 5.6­11.6) EFV, and 9.8% (95% CI 5.7­15.3%) protease inhibitor-based regimens (P»0.48). CONCLUSION: Viral load suppression by 12 months was high with all regimens. NVPþ3NRTIs regimens were particularly efficacious in the longer term and may be a good alternative to protease inhibitor-based ART in young children.

N-methyl-D-aspartate receptor antibodies in post-herpes simplex virus encephalitis neurological relapse.

Herpes simplex virus encephalitis (HSVE) is a devastating condition that relapses, often with a chorea in children, despite adequate antiviral treatment. At relapse, evidence of viral replication is frequently absent, suggesting that the relapse may be immune-mediated. Seven children who had a neurological relapse following their initial encephalitis, identified from 20 cases of pediatric HSVE, were studied. Serum and/or cerebrospinal fluid (CSF) were tested for N-methyl-D-aspartate receptor (NMDAR) and other antibodies previously reported in central nervous system autoimmunity. Five of the 7 relapsing children had choreoathetosis; 2 of these were NMDAR antibody-positive, 2 were negative (1 with HSV-positive CSF), and 1 was not available for testing. An additional patient, who relapsed with cognitive regression but with no movement disorder, was also NMDAR antibody-positive. In 2 of the NMDAR antibody-positive patients who were treated at relapse and in 1 who was treated only after 10 years of having a relapsing encephalopathy, a beneficial response was observed. Neurological relapses after HSVE may frequently be immune-mediated, particularly in children with chorea. NMDAR antibodies are common, and immunotherapy may be beneficial.

Fifteen-minute consultation: a structured approach to the management of recurrent oral ulceration in a child.

OBJECTIVE: To present a structured approach for an outpatient consultation of a child with recurrent mouth ulcers. METHOD: Review of literature and description of approach followed in our unit. CONCLUSIONS: The literature emphasises the need to consider local and systemic causes for oral ulceration in a child. Focused history and examination are key in establishing the cause and in order to ensure appropriate management.

Identification and control of a gentamicin resistant, meticillin susceptible Staphylococcus aureus outbreak on a neonatal unit.

We describe the identification and control of an outbreak of gentamicin resistant, meticillin susceptible Staphylococcus aureus (GR-MSSA) on a 36-bed neonatal unit (NNU) in London. Control measures included admission and weekly screening for GR-MSSA, cohorting affected babies, environmental and staff screening, hydrogen peroxide vapour (HPV) for terminal disinfection of cohort rooms, and reinforcement of hand hygiene. Seventeen babies were affected by the outbreak strain over ten months; seven were infected and ten were asymptomatic carriers. The outbreak strain was gentamicin resistant and all isolates were indistinguishable by pulsed-field gel electrophoresis. The outbreak strains spread rapidly and were associated with a high rate of bacteraemia (35% of 17 affected patients had bacteraemia vs. 10% of 284 patients with MSSA prior to the outbreak, p=0.007). None of 113 staff members tested were colonised with GR-MSSA. GR-MSSA was recovered from 11.5% of 87 environmental surfaces in cohort rooms, 7.1% of 28 communal surfaces and 4.1% of 74 surfaces after conventional terminal disinfection. None of 64 surfaces sampled after HPV decontamination yielded GR-MSSA. Recovery of GR-MSSA from two high level sites suggested that the organism could have been transmitted via air. Occasional breakdown in hand hygiene compliance and contaminated environmental surfaces probably contributed to transmission.

Lopinavir dosing in HIV-infected children in the United Kingdom and Ireland.

BACKGROUND: Uncertainty surrounds the correct dosing of lopinavir/r (LPV/r) in HIV-infected children not receiving non-nucleoside reverse transcriptase inhibitors. The licensed total daily dose is 460 mg/m², whereas the original study, reporting excellent viral load (VL) suppression, used a higher 600 mg/m² dose. METHODS: We calculated LPV/r daily doses prescribed from 2000 to 2009 within the UK/Irish national Collaborative HIV Paediatric Study (CHIPS) cohort. Logistic and binomial mixed models were used to explore whether higher LPV/r doses affected VL suppression. RESULTS: Four hundred forty-four of 1201 (37%) children on antiretroviral therapy in CHIPS had taken lopinavir/r without non-nucleoside reverse transcriptase inhibitors. Of 1065 recorded doses, 48% were syrup, 27% capsules and 25% tablets. Ten percent of doses were >10% below 460 mg/m² per day, and 12% were >10% above 600 mg/m². In multivariable models, predictors of lower doses were: once versus twice daily dosing (32 mg/m² lower); syrup versus tablets/capsules (33 mg/m² lower); higher weight-for-age and height-for-age (24 mg/m² and 13 mg/m² lower per unit higher, respectively); and older age (13 mg/m lower per year older for those aged >10 years, P < 0.05). Dosing varied widely by hospital (P = 0.0004), with some targeting higher and others lower doses. For those receiving lopinavir/r for ≥6 months, there was a greater chance of VL <400 copies/mL with higher doses (odds ratio = 1.15 [95% confidence interval: 1.06-1.25 per 50 mg/m² higher], P = 0.001). CONCLUSIONS: Findings suggest substantial variation and large hospital-level effects in the LPV/r dose prescribed to HIV-infected children in the United Kingdom/Ireland. Higher doses appeared to improve long-term VL suppression, which may be critical in children who need life-long therapy. Results highlight the importance of optimizing dosing in HIV-infected children of all ages.

Fatal laryngeal diphtheria in a UK child.

Over the last century, the infectious causes of acute upper airway obstruction have changed dramatically. Toxigenic Corynebacterium diphtheriae has become rare in the UK due to national immunisation programmes. Since 1986, eight sporadic cases of C diphtheriae were reported, all of whom had recently returned from endemic areas. We describe a case of fatal laryngeal diphtheria in an unimmunised child. Although appropriate antimicrobial cover was provided, antitoxin was not administered due to a low index of suspicion. This case represents the first UK death from C diphtheriae in 14 years and where travel to an endemic country or contact with a known case of diphtheria was not identified. We highlight the need to maintain a high index of suspicion in children for whom completion of the immunisation schedule is not confirmed regardless of travel history. Prompt recognition and timely administration of antitoxin may be life-saving.

Comparing neonatal and paediatric antibiotic prescribing between hospitals: a new algorithm to help international benchmarking.

OBJECTIVES: The WHO anatomical therapeutic chemical (ATC)/defined daily dose (DDD) methodology is a standardized method of comparing antimicrobial use. The ATC/DDD is defined as the average maintenance daily dose of a drug used in a 70 kg adult, ignoring the considerable differences in body weight of neonates and children. The aim of this study was to develop a new standardized way of comparing rates of antimicrobial prescribing between European children's hospitals. METHODS: This pilot study at four European children's hospitals (in the UK, Greece and Italy) collected data including demographics, antibiotic use, dosing and indication in children and neonates over a 14 day period. RESULTS: A total of 1217 antibiotic prescriptions were issued with 47 different antibiotics used. Approximately half of all children and a third of all neonates received antibiotics, with wide variation between centres in the type and dose of antibiotic used. We propose a new pragmatic three-step algorithm. The first step includes a simple comparison of the proportion of hospitalized children on antibiotics by weight bands and the number of antimicrobials that account for 90% of total DDD drug usage (DU90%). The second step is a comparison of the dosing used (mg/kg/day). The third step is to compare overall drug exposure using DDD/100 bed days for standardized weight bands between centres. CONCLUSIONS: This novel method has the potential to be a useful tool to provide antibiotic use comparator data and requires validation in a large prospective point prevalence study.

The prevalence of darunavir-associated mutations in HIV-1-infected children in the UK.

BACKGROUND: We examined the prevalence of ritonavir-boosted darunavir (DRV) resistance-associated mutations (RAMs) in HIV-infected children in the UK to determine the drug's potential clinical utility as a first-line or second-line protease inhibitor (PI). METHODS: The prevalence of DRV RAMs, identified from IAS 2010 and Stanford, and the Stanford susceptibility score, were estimated in PI-naive and PI-experienced children in the Collaborative HIV Paediatric Study and the UK HIV Drug Resistance Database 1998-2008. Associations between type/duration of PI exposure and area under the viraemia curve on PI with the number of RAMs were investigated using multivariate Poisson regression. RESULTS: A total of 17/417 (4%) children with a resistance test when PI-naive had one IAS DRV RAM, and 1 had a Stanford mutation; none had multiple DRV RAMs. A total of 177 PI-experienced children had a test after a median 2.7 years (IQR 1.1-5.2) on PIs; 19 (11%) had one IAS DRV RAM, 7 (4%) had two RAMs, 1 (0.6%) had three RAMs and 1 (0.6%) had four RAMs. DRV RAMs were independently associated with increased years on a PI, a larger area under the viraemia curve since starting PIs, and any exposure to PIs other than lopinavir (all P≤0.05). Only 6 (3%) PI-experienced children had intermediate-level DRV/ritonavir resistance; none had high-level resistance. CONCLUSIONS: DRV resistance was negligible in PI-naive children and those with lopinavir PI exposure alone. However resistance increased with increasing time, and with higher levels of viraemia, on PIs. Once-daily DRV/ritonavir would be valuable as a second PI or an alternative first PI, particularly if coformulated with a booster in an appropriate formulation for children.

Factors influencing T cell activation and programmed death 1 expression in HIV-infected children.

Immune activation is the best marker of HIV disease progression in both adults and children. However, the factors that drive immune activation in HIV-infected children remain incompletely understood and may differ from those in adults. Immune activation was investigated in a cohort of 93 untreated HIV-infected children, of median age 10.8 years, and 37 HIV-uninfected children. CD8(+) T cell activation, which was higher in HIV-infected than HIV-uninfected children (p<0.001), did not correlate with viral load (R=-0.03, p=0.838). Similarly, programmed death 1 (PD-1) expression on CD8(+) T cells, which was higher in HIV-infected children than HIV-uninfected children (p<0.001), was not associated with viral load (R=0.11, p=0.40), but correlated with CD8 activation (R=0.41, p=0.002). Both CD8 activation and PD-1 expression were partially driven by the magnitude of the HIV-specific CD8(+) T cell response. CD3(+)CD4(+)CD25(hi)FoxP3(+) regulatory T cells (Tregs) were depleted in HIV-infected, compared to HIV-uninfected, children [median 1.0% (IQR 0.6, 1.9) vs. 2.6% (IQR 1.7, 3.2) CD3 cells; p<0.001]. Depletion was associated with increased CD8 activation (R=-0.27, p=0.068), suggesting that the decline in Tregs may allow immune activation to increase. Taken together, immune activation and PD-1 upregulation in children are not directly driven by viral load but may be influenced by the magnitude of the immune response to HIV itself, and to the depletion of Tregs that occurs during HIV infection. Further understanding of the factors that drive immune activation in children is critical to developing future therapeutic strategies in this population.

Characteristics of Invasive Staphylococcus aureus in United Kingdom Neonatal Units.

BACKGROUND: In industrialized countries, Staphylococcus aureus (SA) is a leading cause of late-onset neonatal sepsis. METHODS: Culture-proven episodes were identified prospectively from neonatal units participating in the neonatal infection surveillance network. Demographic, risk factor, and outcome data were collected. RESULTS: Between 2004 and 2009, there were 117 episodes of SA infections (including 8 methicillin-resistant SA) in 116 infants from 13 units. The median gestational age and birth-weight were 27 weeks (90% ≤ 37 weeks, 85% ≤ 32 weeks) and 850 g (90% ≤ 2500 g), respectively. The overall incidence was 0.6 per 1000 live births and 23/1000 in infants <1500 g. Most episodes (94%) occurred more than 48 hours after birth (late onset). There were 7 early-onset episodes (< 48 hours) (median gestational age, 38.5 weeks), all due to methicillin-susceptible SA. At the time of culture, 67 of 95 (71%) infants were receiving respiratory support and 47 of 94 (50%) had a central line in situ. The majority of infants had nonspecific clinical features although evidence of focal infection (skin, soft tissue, bone, joint, or pneumonia) was ultimately seen in 41 of 91 (45%). There were 18 deaths, 4 (all late onset) directly due to methicillin-susceptible SA sepsis (4.4%). CONCLUSIONS: SA is the second most common pathogen causing late-onset neonatal infections in this neonatal network. Infants who weigh < 1500 g in intensive care settings are the most vulnerable group. Clinical signs are not sufficiently distinctive to allow targeted therapy, suggesting that an antistaphylococcal agent should be part of empiric therapy for late-onset sepsis in premature infants.