55 - Role of urine gram stain in the management of young febrile infants with a suspected urinary tract infection
Friday, April 28, 2023
5:15 PM – 7:15 PM ET
Poster Number: 55 Publication Number: 55.112
Borja Gomez, Cruces University Hospital, Barakaldo, Pais Vasco, Spain; Ana Mier, Hospital de Cruces, Bilbao, Pais Vasco, Spain; Alberto Ugedo, Cruces University Hospital, Gamiz-Fika, Pais Vasco, Spain; Javier Benito, Cruces University Hospital, Bilbao, Pais Vasco, Spain; Santiago Mintegi, Hospital Universitario Cruces, Bilbao, Pais Vasco, Spain
Pediatric Emergency Physician Cruces University Hospital Barakaldo, Pais Vasco, Spain
Background: The American Academy of Pediatrics recommends defining a urinary tract infection (UTI) as a positive urine culture (UC) with associated pyuria/leukocyturia. Urine gram stain is considered a predictor of positive UC, but it is not usually included in the initial decision making.
Objective: To analyze the performance of the urine gram stain to predict a positive UC in young febrile infants with leukocyturia.
Design/Methods: Secondary analysis of a prospective unicenter registry, including all the infants ≤90 days old attended in our pediatric emergency department with fever without source. For this subanalysis, we included those infants attended between 2010 and 2021 with leukocyturia and in whom a urine gram stain was requested (urethral catheterization). We excluded patients with a gram stain showing mixed flora. We analyzed the sensitivity, specificity and positive and negative predictive values (PPV and NPV) of the gram stain (defined as positive if any organisms were seen) for predicting a confirmed UTI (positive UC: >10,000 UFC/mL of a unique bacterium).
Results: Among the 2,537 infants attended, 547 (21.5%) had leukocyturia. All of them had a UC obtained and 371 (67.8%) had a urine gram stain. We excluded 3 infants with a gram stain showing mixed flora.
Among the 368 included infants, 279 (75.8%) infants had a positive gram stain and 304 (82.6%) had a positive UC (274; 90.1% by E. coli).
Rate of positive UC was higher among infants with a positive gram stain (97.1% vs 37.1% of those with a negative gram stain; p< 0.01), showing a sensitivity of 89.1% (95%CI 85.1-92.2%), a specificity of 87.5% (95%CI 77.2-93.5%) a PPV of 97.1% (95%CI 94.4-98.5%) and a NPV of 62.9% (95%CI 52.5-72.2%). Sensitivity of the gram stain was lower in patients with a confirmed UTI by bacteria others than E. coli (70.0% vs 91.2% for confirmed UTIs by E. coli; p< 0.01).
Three of the 89 febrile infants with leukocyturia and a negative Gram stain (3.3%) were diagnosed with a bacterial infection other than UTi: two bacteremias by S. pneumoniae and S. aureus, respectively, and one bacterial gastroenteritis by Salmonella typhimurium.
Conclusion(s): Urine gram stain may help in the clinical decision making in young febrile infants with leukocyturia. Only one third of the patients with leukocyturia and no microorganisms in the urine gram stain will be finally diagnosed with a confirmed UTI. These patients have a higher rate of UTIs by bacteria others than E. coli. Bacterial infections other than UTI should be considered in febrile infants with leukocyturia and negative gram stain.
