Factors Associated with Acute Community-Acquired Pyelonephritis Caused by Extended-Spectrum β-Lactamase-Producing Escherichia coli.
Mónica Romero NietoSara Maestre VerdúVicente GilCarlos Pérez BarbaJose Antonio Quesada RicoMaría de Los Reyes Pascual-PérezPublished in: Journal of clinical medicine (2021)
This study aimed to identify the factors associated with the presence of extended-spectrum ß-lactamase-(ESBL) in patients with acute community-acquired pyelonephritis (APN) caused by Escherechia coli (E. coli), with a view of optimising empirical antibiotic therapy in this context. We performed a retrospective analysis of patients with community-acquired APN and confirmed E. coli infection, collecting data related to demographic characteristics, comorbidities, and treatment. The associations of these factors with the presence of ESBL were quantified by fitting multivariate logistic models. Goodness-of-fit and predictive performance were measured using the ROC curve. We included 367 patients of which 51 presented with ESBL, of whom 90.1% had uncomplicated APN, 56.1% were women aged ≤55 years, 33.5% had at least one mild comorbidity, and 12% had recently taken antibiotics. The prevalence of ESBL-producing E. coli was 13%. In the multivariate analysis, the factors independently associated with ESBL were male sex (OR 2.296; 95% CI 1.043-5.055), smoking (OR 4.846, 95% CI 2.376-9.882), hypertension (OR 3.342, 95% CI 1.423-7.852), urinary incontinence (OR 2.291, 95% CI 0.689-7.618) and recurrent urinary tract infections (OR 4.673, 95% CI 2.271-9.614). The area under the ROC curve was 0.802 (IC 95% 0.7307-0.8736), meaning our model can correctly classify an individual with ESBL-producing E. coli infection in 80.2% of cases.
Keyphrases
- escherichia coli
- urinary tract infection
- klebsiella pneumoniae
- mental health
- healthcare
- urinary incontinence
- biofilm formation
- end stage renal disease
- blood pressure
- ejection fraction
- chronic kidney disease
- liver failure
- prognostic factors
- machine learning
- metabolic syndrome
- stem cells
- intensive care unit
- risk factors
- pregnant women
- big data
- electronic health record
- cystic fibrosis
- bone marrow
- drug induced
- hepatitis b virus
- deep learning
- extracorporeal membrane oxygenation