loading page

Does Fever Response to Acetaminophen Predict Blood Stream Infections in Febrile Neutropenia?
  • +1
  • Duncan Mackie,
  • Dennis Kuo,
  • Megan Paul,
  • Jennifer Elster
Duncan Mackie
University of California San Diego School of Medicine

Corresponding Author:[email protected]

Author Profile
Dennis Kuo
University of California San Diego School of Medicine
Author Profile
Megan Paul
University of California San Diego School of Medicine
Author Profile
Jennifer Elster
University of California San Diego School of Medicine
Author Profile

Abstract

Background: There is a need to identify clinical parameters for early and effective risk stratification and prediction of bacterial blood stream infections (BSI) in patients with febrile neutropenia (FN). 1,2,3,4 Acetaminophen is used widely to treat fever in FN; however, little research exists on whether fever response to acetaminophen can be used as a predictor of BSIs. Objectives: Investigate the relationship between fever response to acetaminophen and bacteremia in FN. Design/Method: A retrospective review of patients (1-21 years old) presenting with FN and bacteremia at Rady Children’s Hospital (2012-2018) was performed. Demographic information, presenting signs/symptoms, degree of neutropenia (ANC > 500 or < 500 cells/µL), absolute monocyte count (AMC), blood culture results, temperatures 1-, 2-, and 6-hours after acetaminophen, and timing of antibiotic administration were examined. Patients were stratified into three malignancy categories: leukemia/lymphoma, solid tumor, and hematopoietic stem cell transplant (HSCT). Patients were matched with culture negative controls based on sex, age, malignancy category, and degree of neutropenia. Results: Thirty-five cohort-control pairs met inclusion criteria (70 presentations of FN). Mean age of cohort was 10.7 years (± 6.3) vs. 10.0 years (± 5.9) for the controls. Twenty were female (57%). Twenty-three pairs were categorized as leukemia/lymphoma (66%), 8 as solid tumor (23%) and 4 as HSCT (11%). Thirty-four pairs (97%) had a presenting ANC < 500 cells/µL. Higher temperature 1-hour post-acetaminophen was associated with bacteremia (p = 0.04). Logistic regression demonstrated that temperature 1-hour post-acetaminophen had significant predictive value for bacteremia (p = 0.011). Area under the receiver operating characteristic curves (AUC-ROC) for logistic regression and classification and regression tree (CART) analysis were 0.70 and 0.71 respectively. Conclusion: While temperature 1-hour post-acetaminophen was higher among patients with bacteremia and was a significant predictor of bacteremia, fever response in isolation lacks sufficient predictive value to impact clinical decision making. Future studies are needed to assess fever responsiveness as an adjunct to existing modalities of FN risk stratification.