Postoperative Rather Than Preoperative Neutropenia Is Associated With Early Catheter-related Bloodstream Infections in Newly Diagnosed Pediatric Cancer Patients

• Published in: The Pediatric infectious disease journal Vol. 41; no. 2; p. 133
• Main Authors: Cher, Wen Qi, Lee, Vanessa, Wang, Rachel, Cheah, Sue-Mei, Lee, York Tien, Saffari, Seyed Ehsan, Tan, Chin-Bee, Chong, Chia-Yin, Lam, Joyce C M, Loh, Amos H P
• Format: Journal Article
• Language: English
• Published: United States 01.02.2022
• Subjects: Adolescent; Antibiotic Prophylaxis - statistics & numerical data; Bacteremia - complications; Bacteremia - epidemiology; Catheter-Related Infections - complications; Catheter-Related Infections - epidemiology; Central Venous Catheters - adverse effects; Child; Child, Preschool; Female; Humans; Infant; Leukemia - complications; Leukemia - epidemiology; Leukemia - surgery; Male; Neutropenia - complications; Neutropenia - epidemiology; Perioperative Period - statistics & numerical data; Retrospective Studies
• ISSN: 1532-0987
• DOI: 10.1097/INF.0000000000003315

The relationship of early catheter-related bloodstream infections (CRBSIs) with perioperative neutropenia and antibiotic prophylaxis is not well established. We sought to evaluate perioperative factors associated with early CRBSIs in newly diagnosed pediatric cancer patients, particularly hematologic indices and antibiotic use. We retrospectively reviewed national registry records of newly diagnosed pediatric cancer patients with port-a-caths inserted using standardized perioperative protocols where only antibiotic use was not regulated. Thirty-day postoperative CRBSI incidence was correlated with preoperative factors using logistic regression and with postoperative blood counts using linear trend analysis. Among 243 patients, 17 CRBSIs (7.0%) occurred at median 14 (range, 8-28) postoperative days. Early CRBSIs were significantly associated with cancer type [acute myeloid leukemia and other leukemias (AML/OLs) vs. solid tumors and lymphomas (STLs): odds ratio (OR), 5.09; P = 0.0036; acute lymphoblastic leukemia vs. STL: OR 0.83; P = 0.0446] but not preoperative antibiotics, absolute neutrophil counts and white blood cell counts. Thirty-day postoperative absolute neutrophil counts and white blood cell trends differed significantly between patients with acute lymphoblastic leukemia and STLs (OR 0.83, P < 0.05) and between AML/OLs and STLs (OR 5.09, P < 0.005), with AML/OL patients having the most protracted neutropenia during this period. Contrary to common belief, low preoperative absolute neutrophil counts and lack of preoperative antibiotics were not associated with higher early CRBSI rates. Instead, AML/OL patients, particularly those with prolonged neutropenia during the first 30 postoperative days, were at increased risk. Our findings do not support the use of empirical preoperative antibiotics and instead identify prolonged postoperative neutropenia as a major contributing factor for early CRBSI.