Background: Data on Candida bloodstream infections in pediatric patients in Europe are limited. We performed a retrospective multicenter European study of the epidemiology and outcome of neonatal and pediatric candidemia. Material and Methods: All first positive blood cultures from patients ≤ 18 years of age with candidemia were registered. Patients' demographic and clinical characteristics and causative Candida species were collected and analyzed. Regression analysis was used to identify factors independently associated with mortality. Results: One thousand three hundred ninety-five episodes of candidemia (57.8% male) were reported from 23 hospitals in 10 European countries. Of the 1395 episodes, 36.4% occurred in neonates (≤ 44 weeks postmenstrual age), 13.8% in infants (> 44 weeks postmenstrual age to 1 year) and 49.8% in children and adolescents. Candida albicans (52.5%) and Candida parapsilosis (28%) were the predominant species. A higher proportion of candidemia caused by C. albicans was observed among neonatal patients (60.2%) with highest rates of C. parapsilosis seen among infants (42%). Children admitted to hematology-oncology wards presented the highest rates of non-albicans Candida species. Candidemia because of C. albicans was more frequent than non-albicans Candida in Northern versus Southern Europe (odds ratio, 2.3; 95% confidence interval, 1.8-2.9; P < 0.001). The all-cause mortality at 30 days was 14.4%. All-cause mortality was higher among patients admitted to the neonatal or pediatric intensive care units than other wards. Over time, no significant changes in species distribution were observed. Conclusions: This first multicenter European study shows unique characteristics of the epidemiology of pediatric candidemia. The insights obtained from this study will be useful to guide clinical management and antifungal stewardship.

Logistic regression (LR) is the most common prediction model in medicine. In recent years, supervised machine learning (ML) methods have gained popularity. However, there are many concerns about ML utility for small sample sizes. In this study, we aim to compare the performance of 7 algorithms in the prediction of 1-year mortality and clinical progression to AIDS in a small cohort of infants living with HIV from South Africa and Mozambique. The data set (n = 100) was randomly split into 70% training and 30% validation set. Seven algorithms (LR, Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Naïve Bayes (NB), Artificial Neural Network (ANN), and Elastic Net) were compared. The variables included as predictors were the same across the models including sociodemographic, virologic, immunologic, and maternal status features. For each of the models, a parameter tuning was performed to select the best-performing hyperparameters using 5 times repeated 10-fold cross-validation. A confusion-matrix was built to assess their accuracy, sensitivity, and specificity. RF ranked as the best algorithm in terms of accuracy (82,8%), sensitivity (78%), and AUC (0,73). Regarding specificity and sensitivity, RF showed better performance than the other algorithms in the external validation and the highest AUC. LR showed lower performance compared with RF, SVM, or KNN. The outcome of children living with perinatally acquired HIV can be predicted with considerable accuracy using ML algorithms. Better models would benefit less specialized staff in limited resources countries to improve prompt referral in case of high-risk clinical progression.

Introduction and Objective: Community-acquired pneumonia (CAP) is one of the most common reasons of prescribing antibiotics for children, often with overuse of broad-spectrum antibiotics. The aim of this study is to describe the antibiotic prescriptions for Italian children with CAP, at the primary care level. Study Design: Retrospective cohort study conducted among children 3 months–14 years of age with CAP, enrolled in Pedianet (http://www.pedianet.it) from January 1, 2009 to December 31, 2018. Antibiotic treatment was defined as narrow-spectrum (NS-ABT) if amoxicillin and broad-spectrum (BS-ABT) if amoxicillin/clavulanic acid, cephalosporins or any combination. Crude and adjusted logistic regressions for the odds of receiving NS-ABT were conducted (all episodes of CAP and per patient). A P value <0.05 was considered statistically significant. Results: Among 9691 CAP, 7260 episodes from 6409 children followed by 147 pediatricians were analyzed. The 16.7% of CAP [1216/7260, 95% confidence interval (CI): 15.9%–17.6%] received an NS-ABT while 53.3% (3863/7260, 95% CI: 52%–54.4%) received BS-ABTs and 30% (2181/7260, 95% CI: 28.9%–31.1%) macrolides. Within 10 years, a slight but increasing trend of NS-ABT prescription was observed (P < 0.001). Factors independently associated with reduced odds of receiving an NS-ABT compared with BS-ABT including macrolides were being older than 5 years [odds ratio (OR) 0.45, 95% CI: 0.39–0.52], living in Central/Southern Italy (OR 0.13, 95% CI: 0.10–0.16) and being exposed to ABT 3 months before (OR 0.61, 95% CI: 0.53–0.70). These findings were confirmed comparing NS-ABT versus BS-ABT excluding macrolides (n = 5079) and when the analysis was limited to index CAP. Conclusion: Our findings report a very limited prescription of narrow-spectrum antibiotics for Italian children with CAP.

Background: Although the number of Zika virus (ZIKV) cases has substantially declined in Latin America and the Caribbean since the 2015-2016 outbreaks, the cohort of children born at that time and affected by congenital zika syndrome (CZS) are now around 4-5 years old and experiencing an ongoing impact on their health and development. Gaps in our understanding remain regarding the outcomes of ZIKV exposure in utero and congenital infection and the consequences of congenital zika syndrome (CZS) for health throughout childhood. Methods: The ZIKAction Paediatric Registry is an international multi-centre registry of infants and children with documented ZIKV exposure in utero (i.e. born to mother with confirmed infection in pregnancy) and/or with confirmed or suspected congenital ZIKV infection. Clinical teams at participating sites in Argentina, Brazil and Jamaica conduct retrospective case note reviews of children eligible for inclusion in the Registry and enter pseudonymised data into a central Registry database, with additional data collected prospectively on routine follow-up at some sites. Data collected will include sociodemographic, maternal and pregnancy information, delivery information and newborn assessment, paediatric clinical assessments (physical, neurological, developmental, ophthalmological, audiological), and laboratory results conducted as part of local standard of care. The ZIKAction Paediatric Registry network will conduct pooled analyses to address questions relating to characteristics, health and neurodevelopmental outcomes of this population. The Registry is embedded within a larger programme of research studies conducted by ZIKAction. Discussion: As the health outcomes of children affected by ZIKV continue to unfold, this paediatric registry will provide comprehensive data on their clinical and neurodevelopmental outcomes, growth and management, as well as on later sequelae. This will inform their support and care and provide potential insights on pathogenesis of the disease, of importance to currently affected families and for the response to possible future outbreaks. It will highlight the service needs of the affected populations in Latin America and the Caribbean and allow the identification of potential participants for future studies.

Background While antifungal consumption in immunocompromised patients appears high, data on children are limited. We analyzed antifungal use in hospitalized immunocompromised children across Europe in order to better organize a European pediatric antifungal stewardship programs (pAFS). Methods A multicenter 12-wk modified point-prevalence study was conducted. All patients hospitalized in hematology-oncology (HO) or bone marrow/solid organ transplant (BMT/SOT) units across Europe and receiving systemic antifungals were included. Data on ward demographics and policies were collected once at the beginning; weekly ward and patient data were prospectively collected during the 12-wk study period and entered in REDCap. Systemic antifungals administered were recorded (doses, duration, reason for administration or discontinuation). Results Twenty-one HO and 10 BMT/SOT units from 18 hospitals in 11 European countries participated in the study and 572 antifungal prescriptions were recorded. The most common underlying conditions were: 69% malignancy (81% hematologic, 19% solid tumors), 20% BMT, 6% hematologic diseases except malignancy and 4% primary immunodeficiency. Indication of antifungals was prophylaxis for 439 (77%) and treatment for 133 (23%) prescriptions (62 empirical, 43 pre-emptive, 28 targeted). Most common reasons for empirical, pre-emptive and targeted treatment were antibiotic-resistant febrile neutropenia (52%), abnormalities on chest-CT with/without positive galactomannan (72%) and candidiasis (82%), respectively. For targeted treatment, the most frequent pathogens were C. albicans (50%), C. parapsilosis (11%) and A. fumigatus (11%). Overall, fluconazole and liposomal amphotericin B were the most frequently prescribed agents both for prophylaxis (31% and 21%) and treatment (32% and 23%). However, in BMT patients the most frequently prescribed antifungal agents were fluconazole or micafungin for prophylaxis and caspofungin and voriconazole for treatment (Table). Antifungal agents used per underlying condition Antifungal agents used per underlying condition Conclusion Most systemic antifungal prescribing across European HO and BMT/SOT units is for prophylaxis, and fluconazole is the main antifungal prescribed. Results from this multicenter study can be a first step to guide a Europe-wide pAFS. Disclosures Emmanuel Roilides, MD, PhD, ECMM (Research Grant or Support, Other Financial or Material Support, ECMM grant for this study)