Assessment of Bacteremic Risk in 1100 Febrile Neutropenic Pediatric Oncology Patients 750

Abstract

Recently, several pediatric oncology centers have begun to manage carefully selected patients with fever and neutropenia as outpatients. The premise for such strategies has been that a relatively lower risk group of children with fever and neutropenia can be identified at the time of presentation based upon clinical features. However, this hypothesis has not been proven in a large population of children with febrile neutropenia. To determine whether a group at lower risk of bacteremia exists, we identified 1157 febrile neutropenic episodes in children aged 1 year or greater from three pediatric oncology centers. A total of 167 episodes of bacteremia were identified in this group for a frequency of 14%. Using univariate logistic regression, we assessed the following covariates with respect to risk for bacteremia: age, absolute monocyte, neutrophil, and phagocyte counts ( 250) and temperature elevation ( 39.5). We found that AMC < 100 conferred no increased risk of bacteremia when compared to AMC of 250 or greater (odds ratio.8; 95% CI.34-2.06). In addition, ANC < 100 conferred no increased risk of bacteremia when compared to ANC of 250 or greater (odds ratio 1.13; 95%CI.57-2.2) and APC of 100 or less similarly did not confer increased risk (odds ratio 1.22; 95%CI.68-2.16). Increased temperature above 39.5 also was not associated with bacteremia. Current strategies to identify children with fever and neutropenia at lower risk for bacteremia may be limited, since none of the previously identified risk factors associated with bacteremia were identified in this larger series of pediatric patients.