This article is more than 5 years old. Information may no longer be current.
UTI risk calculator decreases testing in infants with fever
The use of a calculator that can assess the probability of UTI using clinical variables, and later update the probability once lab results are received, reduces testing in nonverbal febrile infants younger than 2 years of age by 8.1%, according to findings published in JAMA Pediatrics.
“Testing for urinary tract infection in young children, whether by catheterization or via the two-step process, is challenging,” Nader Shaikh, MD, MPH, from the division of general academic pediatrics at the Children’s Hospital of Pittsburgh of UPMC, and colleagues wrote. “Accordingly, clinicians obtain samples only when they judge the probability of UTI to be sufficiently high. Estimating the probability of UTI by using each child’s unique set of presenting signs and symptoms can be challenging given the relatively large number or variables that modify the risk.”
To create and examine the efficacy of UTICalc, the researchers conducted a review of electronic medical records of febrile infants aged between 2 and 23 months. All children presented at the Children’s Hospital of Pittsburgh ED.
Shaikh and colleagues then formed an independent training database that included 1,686 patients seen between Jan. 1, 2007, and April 30, 2013. Additionally, a validation database was created and included 384 patients. Training and testing of the predictor were completed using five multivariable logistic regression models. Only clinical variables were contained in the clinical model, and all other models used lab results.
The researchers were aware only of the temperature of the included children, who all had a recorded fever (temperature at least 38°C; 100.4°F) and were aged between 2 and 23 months. The risk of UTI was predicted using each model through culture-confirmed UTI, and subsequent models were merged into UTICalc and implemented for the analysis of medical records.
Of those included in the training database, 72.1% were female and 69.2% were white. The validation database had similar demographics, with females making up 75.8% of this cohort and 52.1% were white. When UTICalc was used as a clinical model, testing for UTIs decreased by 8.1% when compared with the current AAP algorithm (95% CI, 4.2%-12.0%). Additionally, the number of missed infections was reduced from three to zero.
Upon examination of the UTICalc dipstick model, Shaikh and colleagues observed that the number of treatment delays would decrease by 10.6% when compared with the practice of empirically treating all infants that have a leukocyte esterase test with a result of +1 or higher (95% CI, 0.9%-20.4%).
“We envision that UTICalc, should it prove cost-effective, could eventually be incorporated as a decision-support tool in electronic health records and could, to a large extent, be prepopulated before the clinician assesses the child,” the researchers wrote. “Of the five variables in the clinical model, all except absence of other source of fever can be easily obtained by a triage nurse in person or by phone. Because many children have upper respiratory tract symptoms, this variable will also often be known.” – by Katherine Bortz
Disclosures: The authors report no relevant financial disclosures.
