Objective: To identify clinical and laboratory risk factors for invasive infection in febrile oncology patients seen in the Pediatric Emergency Department, depending on the type of cancer. Methods: We conducted a prospective observational study of febrile oncology patients seen between 2016 and 2023 at the Pediatric Emergency Department of a tertiary teaching hospital. Invasive infection was defined microbiologically (isolation of a bacterial pathogen in sterile fluid or a commensal bacterial species in 2 different blood cultures) or clinically (death, meeting Phoenix criteria for sepsis, receiving inotropic support, or developing acute complications or sequelae). We performed separate multivariate analyses for hematologic cancers and solid tumors. Hematologic cancers were further classified considering the myelotoxicity of the chemotherapy received. We classified the following as high-risk hematologic cancers: acute lymphoblastic leukemia and non-Hodgkin lymphoma in the induction, reinduction, or consolidation phase or relapse, and acute myeloid leukemia in any phase Results: We included 471 episodes: 306 hematologic cancers and 165 solid tumors. The median age was 4 years (interquartile range: 2 to 10) and the median duration of fever was 2 hours (interquartile range: 1 to 3). Invasive infections were diagnosed in 69 cases (14.6%). The rate was higher among patients with high-risk hematologic cancers (n=45, 23.2%), than among those with solid tumors (n=18, 10.9%; P =0.002) or with low-risk hematologic cancers (n=6, 5.4%; P <0.001). Among patients with hematological cancers, presenting a high-risk cancer (OR: 6.006; 95% CI: 2.459-18.200) and elevated procalcitonin levels (OR: 1.668; 95% CI: 1.205-2.571) were predictors of invasive infection AUC for the model: 0.718 (95% CI: 0.640-0.795). Only age (OR: 1.145; 95% CI: 1.036-1.267) was found to be an independent risk factor in patients with solid tumors. Conclusions: In patients with hematologic cancers and very recent onset fever, the type of cancer and procalcitonin level are useful for predicting the risk of invasive infection.
Gómez et al. (2026) studied this question.