Cargando…

1372. Urinary Tract Infections Caused by Gram-Positive Bacteria in Patients Younger than 19 Years: Prediction Analysis in a 13-year Hospital-Based Cohort

BACKGROUND: Urinary tract infection (UTI) is one of the common pediatric bacterial infections. Gram positive (GP) pathogens, in contrast to gram negative (GN) bacilli such as E. coli, are less accounted for pediatric UTI. The aim of this study was to identify predictors to enable clinicians to detec...

Descripción completa

Detalles Bibliográficos
Autores principales: Hsu, Yu-Lung, Chang, Shih-Ni, Kuo, Chin-Chi, Lin, Che-Chen, Lin, Hsiao-Chuan, Lai, Huan-Cheng, Hwang, Kao-Pin, Chiang, Hsiu-Yin
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2020
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7777486/
http://dx.doi.org/10.1093/ofid/ofaa439.1554
Descripción
Sumario:BACKGROUND: Urinary tract infection (UTI) is one of the common pediatric bacterial infections. Gram positive (GP) pathogens, in contrast to gram negative (GN) bacilli such as E. coli, are less accounted for pediatric UTI. The aim of this study was to identify predictors to enable clinicians to detect GP uropathogens from mostly causative GN bacteria in children with UTI. METHODS: This retrospective cohort study identified 26,066 paired urinalysis and urine culture obtained from the pediatric patients during 2003-2016. Of patients with UTI meeting our criterial, we included children with first-time UTI and classified them into GP-UTI and GN-UTI (Figure 1). Demographic, clinical and laboratory data were collected into analysis. We built a multivariable logistic regression model to predict the GP-UTI. The model performance was examined by using calibration and discrimination plots. We demonstrated a nomogram to predict GP-UTI that could be feasible in the clinical practice. Figure 1. Flowchart of the Selection Process of the Study Population (N = 3,783 patients). [Image: see text] RESULTS: Of 3,783 children with first-time UTIs, 166 (4.4%) were infected by GP and 3,617 (95.6%) by GN bacteria. The top 3 pathogens for GP uropathogens were vancomycin-resistant (VR) E. faecalis, S. saprophyticus, and coagulase-negative Staphylococcus. Significant risk factors associated with GP-UTI in the multivariable analysis were: age >= 24 months [odds ratio (OR) 3.40, 95% confidence interval (CI) 1.40-8.26], serum white blood cell (WBC) (compared to ≥14.4 x 10(3)/μL) [OR 2.18, 95% CI 1.26-3.77], hemoglobin (compared to < 11.3 g/dL) [OR 1.89, 95% CI 1.04-3.45], negative urine leukocyte esterase [OR 3.12, 95% CI 1.83-5.33], negative urine nitrite [OR 4.14, 95% CI 1.88-9.14] and urine WBC (compared to ≥420/μL) [OR 2.16, 95% CI: 1.09, 4.26] (Table 1). This model had good discrimination (C-statistic 0.874; 95% CI 0.839-0.908) and calibration performance (Figure 2). By using our nomogram, physicians can estimate the probability of UTI that is caused by a GP pathogen, with a probability ranges from 0.04% to 55% (Figure 3). Table 1. Multivariable Prediction Model for Pediatric Urinary Tract Infections Caused by Gram-Positive Bacteria. [Image: see text] Figure 2. Discrimination Plot (A) and Calibration Plot (B) of the Prediction Model for Pediatric Urinary Tract Infections Caused by Gram-Positive Bacteria. [Image: see text] Figure 3. Nomogram of the Prediction Model for Pediatric Urinary Tract Infections Caused by Gram-Positive Bacteria. [Image: see text] CONCLUSION: VR E. faecalis is the leading GP uropathogen in the children less than two years of age which need notice of infection control. Our proposed prediction model for GP UTI in children could help clinicians detect potential GP uropathogen and enable them to choose adequate antibiotic regimen early. DISCLOSURES: All Authors: No reported disclosures