Risha L Moskalewicz1, Leidy L Isenalumhe2, Cindy Luu3, Choo Phei Wee4, Alan L Nager5. 1. Division of Emergency Medicine, Department of Pediatrics, University of Minnesota Children's Hospital, Minneapolis, MN. Electronic address: rishKP@gmail.com. 2. Division of Hematology and Oncology, Department of Pediatrics, Children's Hospital Los Angeles, Los Angeles, CA. Electronic address: Isenalumhe@moffitt.org. 3. Division of Emergency Medicine, Department of Pediatrics, University of Minnesota Children's Hospital, Minneapolis, MN. Electronic address: ciluu@chla.usc.edu. 4. Biostatistician II, Biostatistics Core, Clinical Research Support Office, Children's Hospital Los Angeles, Los Angeles, CA. Electronic address: cwee@chla.usc.edu. 5. Division of Emergency and Transport Medicine, Department of Pediatrics, Children's Hospital Los Angeles, University of Southern California and Keck School of Medicine, Los Angeles, CA. Electronic address: nager@chla.usc.edu.
Abstract
OBJECTIVE: To examine clinical characteristics associated with bacteremia in febrile nonneutropenic pediatric oncology patients with central venous catheters (CVCs) in the emergency department (ED). BACKGROUND: Fever is the primary reason pediatric oncology patients present to the ED. The literature states that 0.9% to 39% of febrile nonneutropenic oncology patients are bacteremic, yet few studies have investigated infectious risk factors in this population. METHODS: This was a retrospective cohort study in a pediatric ED, reviewing medical records from 2002 to 2014. Inclusion criteria were patients with cancer, temperature at least 38°C, presence of a CVC, absolute neutrophil count greater than 500 cells/μL, and age less than 22 years. Exclusion criteria were repeat ED visits within 72 hours, bloodwork results not reported by the laboratory, and patients without oncologic history documented at the study hospital. The primary outcome measure is a positive blood culture (+BC). Other variables include age, sex, CVC type, cancer diagnosis, absolute neutrophil count, vital signs, upper respiratory infection (URI) symptoms, and amount of intravenous (IV) normal saline (NS) administered in the ED. Data were analyzed using descriptive statistics and a multiple logistic regression model. RESULTS: A total of 1322 ED visits were sampled, with 534 enrolled, and 39 visits had +BC (7.3%). Variables associated with an increased risk of +BC included the following: absence of URI symptoms (odds ratio [OR], 2.30; 95% CI, 1.13-4.69), neuroblastoma (OR, 3.65; 95% CI, 1.47-9.09), "other" cancer diagnosis (OR, 4.56; 95% CI, 1.93-10.76), tunneled externalized CVC (OR, 5.04; 95% CI, 2.25-11.28), and receiving at least 20 mL/kg IV NS (OR, 2.34; 95% CI, 1.2-4.55). The results of a multiple logistic regression model also showed these variables to be associated with +BC. CONCLUSION: The absence of URI symptoms, presence of an externalized CVC, neuroblastoma or other cancer diagnosis, and receiving at least 20 mL/kg IV NS in the ED are associated with increased risk of bacteremia in nonneutropenic pediatric oncology patients with a CVC.
OBJECTIVE: To examine clinical characteristics associated with bacteremia in febrile nonneutropenic pediatric oncologypatients with central venous catheters (CVCs) in the emergency department (ED). BACKGROUND: Fever is the primary reason pediatric oncology patients present to the ED. The literature states that 0.9% to 39% of febrile nonneutropenic oncologypatients are bacteremic, yet few studies have investigated infectious risk factors in this population. METHODS: This was a retrospective cohort study in a pediatric ED, reviewing medical records from 2002 to 2014. Inclusion criteria were patients with cancer, temperature at least 38°C, presence of a CVC, absolute neutrophil count greater than 500 cells/μL, and age less than 22 years. Exclusion criteria were repeat ED visits within 72 hours, bloodwork results not reported by the laboratory, and patients without oncologic history documented at the study hospital. The primary outcome measure is a positive blood culture (+BC). Other variables include age, sex, CVC type, cancer diagnosis, absolute neutrophil count, vital signs, upper respiratory infection (URI) symptoms, and amount of intravenous (IV) normal saline (NS) administered in the ED. Data were analyzed using descriptive statistics and a multiple logistic regression model. RESULTS: A total of 1322 ED visits were sampled, with 534 enrolled, and 39 visits had +BC (7.3%). Variables associated with an increased risk of +BC included the following: absence of URI symptoms (odds ratio [OR], 2.30; 95% CI, 1.13-4.69), neuroblastoma (OR, 3.65; 95% CI, 1.47-9.09), "other" cancer diagnosis (OR, 4.56; 95% CI, 1.93-10.76), tunneled externalized CVC (OR, 5.04; 95% CI, 2.25-11.28), and receiving at least 20 mL/kg IV NS (OR, 2.34; 95% CI, 1.2-4.55). The results of a multiple logistic regression model also showed these variables to be associated with +BC. CONCLUSION: The absence of URI symptoms, presence of an externalized CVC, neuroblastoma or other cancer diagnosis, and receiving at least 20 mL/kg IV NS in the ED are associated with increased risk of bacteremia in nonneutropenic pediatric oncologypatients with a CVC.