Application of the Weighted-Incidence Syndromic Combination Antibiogram (WISCA) to guide the empiric antibiotic treatment of febrile neutropenia in oncological paediatric patients: experience from two paediatric hospitals in Northern Italy.

BACKGROUND: Guidelines about febrile neutropenia in paediatric patients are not homogeneous; the best empiric treatment of this condition should be driven by local epidemiology. The Weighted-Incidence Syndromic Combination Antibiogram (WISCA) addresses the need for disease-specific local susceptibility evidence that could guide empiric antibiotic prescriptions based on outcome estimates of treatment regimens obtained as a weighted average of pathogen susceptibilities. This study developed a WISCA model to inform empirical antibiotic regimen selection for febrile neutropenia (FN) episodes in onco-haematological paediatric patients treated at two Italian paediatric tertiary centres. METHODS: We included blood cultures from patients with a bloodstream infection and neutropenia admitted to the Paediatric Haematology-Oncology wards in Padua and Genoa Hospitals from 2016 to 2021. WISCAs were developed by estimating the coverage of 20 antibiotics as monotherapy and of 21 combined regimens with a Bayesian probability distribution. RESULTS: We collected 350 blood cultures, including 196 g-negative and 154 g-positive bacteria. Considering the most used antibiotic combinations, such as piperacillin-tazobactam plus amikacin, the median coverage for the pool of bacteria collected in the study was 78%. When adding a glycopeptide, the median coverage increased to 89%, while the replacement of piperacillin-tazobactam with meropenem did not provide benefits. The developed WISCAs showed that no monotherapy offered an adequate coverage rate for the identified pathogens. CONCLUSIONS: The application of WISCA offers the possibility of maximizing the clinical utility of microbiological surveillance data derived from large hospitals to inform the choice of the best empiric treatment while contributing to spare broad-spectrum antibiotics.

Validation of Administrative Data and Timing of Point Prevalence Surveys for Antibiotic Monitoring.

Importance: Point prevalence surveys (PPSs) are used globally to collect data on antibiotic prescriptions. However, the optimal frequency for data collection to ensure comprehensive understanding of antibiotic use and to target and monitor stewardship interventions remains unknown. Objective: To identify the optimal frequency for collecting data on antibiotic use among the pediatric population through PPSs leveraging administrative data. Design, Setting, and Participants: This prognostic study used a cross-sectional validation approach and was conducted in pediatric outpatient and inpatient settings in the Veneto region of Italy. Antibiotics were classified according to the World Health Organization Access, Watch and Reserve criteria. Prescribing rates of access antibiotics were analyzed for pediatric inpatients with records dated between October 1, 2014, and December 31, 2022, and outpatients with records dated between January 1, 2010, and December 31, 2022. The study included children younger than 15 years with an antibiotic prescription who were admitted to the pediatric acute care unit or evaluated by a primary care pediatrician. Data analysis was performed from October 2023 to January 2024. Main Outcomes and Measures: An algorithm was developed to identify optimal time frames for conducting PPSs. This approach sought to minimize the discrepancy between quarterly and yearly PPS results, aiming to accurately estimate annual antibiotic prescribing rates in both inpatient and outpatient settings (primary outcome). External validity of the optimal PPS time frames derived from outpatient data when applied to the inpatient setting was also investigated. Validation involved assessing the effectiveness of administrative data in identifying strategic PPS periods for capturing inpatient antibiotic use patterns (secondary outcome). Results: This analysis included 106 309 children: 3124 were inpatients (1773 males [56.8%]) and 103 185 were outpatients (53 651 males [52.0%]). A total of 5099 and 474 867 antibiotic prescriptions from inpatients and outpatients were analyzed, respectively. Outpatients tended to be older than inpatients, with a median age of 3.2 (IQR, 1.3-6.3) years vs 2.6 (IQR, 0.6-6.6) years, respectively, and with a lower burden of clinical comorbidities (≥1 comorbidity: 6618 [6.4%] vs 1141 [36.5%], respectively). The algorithm successfully identified distinct time frames within the calendar year from inpatient and outpatient records optimized for PPS data collection. Rates obtained from the quarterly PPS during these identified periods exhibited greater agreement with annual antibiotic prescribing rates (inpatient: r = 0.17, P < .001; and outpatient: r = 0.42, P < .001) than those derived from the yearly PPS (inpatient: r = 0.04, P = .58; and outpatient: r = 0.05, P = .34), with a Δ reduction of up to 89.8% (where Δ represents the percentage point change in antibiotic prescribing rates). Furthermore, the optimal PPS time frames gleaned from the outpatient data demonstrated robust applicability to the inpatient setting, yielding comparable results in both scenarios. Conclusions and Relevance: This study evaluated the potential of administrative data in determining the optimal timing of PPS implementation. The quarterly PPS balanced precision and sustainability, especially when implemented during strategically selected periods across different seasons. Further studies are needed to validate the algorithm used in this study, especially in post-COVID-19 pandemic years and different settings.