Background: In 2023, UVA Health’s Catheter-Associated Urinary Tract Infection (CAUTI) Coalition identified diagnostic stewardship as a contributing factor in 57% of all CAUTI events. An independent review of all CAUTIs involving hospitalist physicians from 2022-2023 found that diagnostic stewardship was a key driver. Inappropriate urine testing can lead to treating asymptomatic bacteriuria with antibiotics, which can contribute to antibiotic resistance, longer hospital stays, and risk for Clostridium difficile infections. Despite the publication of institutional guidelines for UTI diagnosis and treatment in October 2022, adherence has been inconsistent.
Purpose: This project aims to reduce the inappropriate treatment of asymptomatic bacteriuria by reducing unnecessary urinalysis (UA) orders. A 12-month retrospective chart review evaluated UA and antibiotic ordering practices across three acute care medicine units. Findings showed that only 47% of reviewed UAs had appropriate indications documented. Of those without, 25% received antibiotics, resulting in inappropriate antibiotic use for asymptomatic bacteriuria about 13% of the time. A Pareto chart identified the top three reasons for inappropriate UA orders: isolated fever without urinary symptoms, altered mental status with another likely cause, and diagnostic evaluation for non-infectious conditions like acute kidney injury.
Description: Embedded clinical decision support (CDS) for urine testing orders was redesigned to address common causes of inappropriate testing. The new CDS prompts clinicians to select appropriate UA indications and discourages pan-culturing for fever and encephalopathy. This change was implemented in Epic August 2024, alongside frontline clinician education about the updated orders. Early data shows a steep decline in UA with reflex culture orders compared to prior baseline (see Figure 1) and a corresponding increase in UA with reflex microscopy orders. No increase in SIRS alerts, chosen as a balancing measure, has been observed. Infectious Disease faculty and clinical pharmacists have identified some cases where clinicians did not appropriately order UA with reflex culture for suspected UTI. Preliminary review point to these orders originating outside the urine testing panel, and alternative urine testing orders are being removed from Epic’s preferences list in order to drive consistent use of the urine testing panel. Chart reviews will be performed post-implementation to assess the appropriateness of UA testing.
Conclusions: Inappropriate urine testing, despite institutional guidelines, is common and contributes to unnecessary antibiotic use and related harms. Inappropriate orders were often due to clinician behaviors like pan-culturing, reflexively ordering UA with culture for acute encephalopathy, or ordering UA with culture instead of UA with reflex microscopy for non-infectious kidney conditions. The enhanced CDS has significantly reduced UA with reflex culture orders. Further evaluations will determine whether the new urine testing panel has reduced low-value testing and inform additional improvements.
