Frequency of Bacteremia and Urinary Tract Infection in Pediatric Renal Transplant Recipients

Background: Our primary goal was to determine the frequency of bacteremia and urinary tract infections (UTI) in pediatric renal transplant recipients presenting with suspected infection within 2 years of transplant and to identify clinical and laboratory factors associated with bacteremia. Methods: We conducted a retrospective cross-sectional study for all pediatric (<18 years old) renal transplant recipients seen at 3 large children's hospitals from 2011 to 2018 for suspected infection within 2 years of transplant date, defined as pyrexia (>38 degrees C) or a blood culture being ordered. Patients with primary immunodeficiencies, nontransplant immunosuppression, intestinal failure, and patients who had moved out of the local area were excluded. The primary outcome was bacteremia or UTI; secondary outcomes included pneumonia, bacterial or fungal meningitis, respiratory viral infections, and antibiotic resistance. The unit of analysis was the visit. Results: One hundred fifteen children had 267 visits for infection evaluation within 2 years of transplant. Bacteremia (with or without UTI) was diagnosed in 9/213 (4.2%) and UTIs in 63/189 (33.3%). Tachycardia and hypotension were present in 66.7% and 0% of visits with documented bacteremia, respectively. White blood cell (12,700 cells/mm(3) vs. 10,900 cells/mm(3); P = 0.43) and absolute neutrophil count (10,700 vs. 8200 cells/mm(3); P = 0.24) were no different in bacteremic and nonbacteremic patients. The absolute band count was higher in children with bacteremia (1900 vs. 600 cells/mm(3); P = 0.02). Among Gram-negative pathogens, antibiotic resistance was seen to 3rd (14.5%) and 4th (3.6%) generation cephalosporins, 12.7% to semisynthetic penicillins, and 3.6% to carbapenems. Conclusions: Bacteremia or UTIs were diagnosed in one-quarter of all pediatric renal transplant recipients presenting with suspected infection within 2 years of transplant. Evaluations were highly variable, with one-third of visits not having urine cultures obtained. No single demographic, clinical or laboratory variable accurately identified patients with bacteremia, although combinations of findings may identify a high-risk population.