While urinary tract infections (UTIs) are frequently encountered in clinical practice, the clinical decision-making involved in the diagnosis of hospital-acquired UTIs is not well described in the literature. The purpose of our study was to identify the clinical data most commonly used to diagnose hospitalized patients with UTI and to investigate the potential impact that a reflex urine culture model (whereby urine cultures are only performed if a specific urinalysis [UA] or microscopy threshold is reached) could have on diagnosis.
We performed a retrospective chart review of adult patients admitted to an urban university hospital with positive urine cultures >48 hours after admission between January 1, 2015 and February 28, 2015. Patient demographics, clinical symptoms, urine studies, and adverse effects were collected. We then applied a reflex urine culture model to our data using two thresholds: 1) any abnormality in leukocyte esterase, nitrite, blood, or protein on UA, and 2) urine microscopy with >10 white blood cells per high power field (WBC/hpf).
In total, 258 patients with positive urine culture were reviewed. Eighty patients were treated for UTI. The strongest predictors of treatment were the presence of >100k Colony-Forming Units (CFUs) in culture (OR 7.55, p = .0001) and an abnormal UA (OR 5.40, p = .002). Seven treated patients (9%) experienced an adverse reaction. Applying the reflex culture model requiring abnormal UA, 9% of treated patients would not have been cultured. Moving the threshold to >10 WBC/hpf, the number rose to 41%.
At our institution, clinicians relied on high colony counts and abnormal urinalyses to guide UTI treatment in hospitalized patients. Though pyuria alone is not diagnostic of a UTI, it can be supportive, and a large proportion of treated patients did not have significant pyuria. These findings highlight areas for clinician education. Additionally, our study suggests reflex urine cultures in the hospital setting may reduce the number of urine cultures performed on samples with lower likelihood of true infection. This in turn could improve diagnostic accuracy and decrease laboratory costs, antibiotic usage, and adverse effects, making it a potentially useful antimicrobial stewardship tool.
M. Brister, None
P. Axelrod, None
T. Fekete, None
D. Mueller, None