LAUSR

Febrile neutropenia is the most frequent complication in children treated with chemotherapy. Nevertheless, neutropenic children are a very heterogeneous group and invasive bacterial infections concern a minority of patients. Reducing antibiotics would bring many benefits. Yet, we can only explore this strategy if the safety of children is preserved. The main aims of this review were to study the safety and effectiveness of reducing antibiotic use in children with febrile neutropenia in terms of duration, route of administration (oral versus intravenous) and narrowing of antimicrobial spectrum. Cochrane Library, Pubmed and Embase were searched for relevant articles until February 2020. We have included all articles describing controlled trials written in French or in English. The risk of bias was assessed with ROB-2 (Cochrane Handbook for Systematic Reviews of Interventions Version 6.0. 2019, Chap. 8) or ROBINS-1 (Cochrane Handbook for Systematic Reviews of Interventions Version 6.0. 2019, Chap. 25). On 2351 articles, the systematic research retained 13 studies. Nine were used for a meta-analysis comparing oral versus intravenous treatment. We found no pediatric studies concerning de-escalation of empiric broad-spectrum antibiotics. No publication biases were found and almost all of the selected studies were at low risk or with some concern for bias. In comparing oral versus intravenous treatment and early cessation versus continuing antibiotics when no infection is proven, we found no difference in terms of safety (mortality and admission in intensive care unit) and efficacy (need of readmission/antibiotic modification/recurrence of fever). It seems safe and effective to provide oral treatment in low-risk febrile neutropenia and to stop antibiotics when no bacterial infection is proven. Spectrum reduction remains an important topic in pediatric research.