Discussion

Appropriate urine diagnostic practices are essential to distinguish true infection from ASB and reduce unnecessary antibiotic use. Optimizing test utilization in the ED supports antimicrobial stewardship by improving diagnostic precision while minimizing inappropriate treatment and associated antibiotic-related harms.

Our multicomponent study found that implementation of a stricter urine diagnostic workflow was associated with a significant reduction in UA and UCX utilization, reflecting improved diagnostic stewardship and more efficient resource utilization by limiting testing in patients with a lower likelihood of UTI. After the updates, UCXs more frequently yielded high bacterial counts (>100,000 CFU/mL), while no-growth cultures declined, suggesting improved targeting of clinically significant bacteriuria rather than incidental findings. The reduction in contaminated UCXs further indicates improved specimen quality and diagnostic yield with stricter reflex criteria, potentially decreasing diagnostic uncertainty, repeat tests, and unnecessary antibiotic exposure. Despite these improvements, increases in UCX-NR and UA-NR orders suggest that some providers bypassed the reflex workflow, likely due to concern for missed infections, habitual ordering, or use of stand-alone testing in populations such as neutropenic patients, where UA-RC may have low sensitivity. These findings suggest that this multicomponent initiative alone may be insufficient and highlight the need for continued provider engagement, education, and feedback to sustain diagnostic stewardship workflows.

No significant differences were observed in antibiotic administration in the ED or prescribing at discharge, which suggests that decreased urine testing did not increase empiric antibiotic use and that treatment decisions were primarily guided by clinical assessment rather than urine test results alone. Additionally, no difference in 30-day UTI-related revisits was observed between the pre and postupdate periods. This provides reassurance that reduced urine testing did not compromise patient outcomes and supports the safety of the updated diagnostic approach in the ED setting.

Frequent use of the “other” indication with nonspecific free-text entries suggests usability gaps in the ordering interface. Many entries reflected broad or non-urinary symptoms, indicating that urine testing was often part of a generalized diagnostic evaluation rather than symptom-driven assessment.

Our findings align with prior studies by Penney et al and Caveness et al, which showed that implementing more restrictive UA reflex criteria led to substantial reductions in culture rates without compromising clinically significant detection, but contrast with their reports of stable stand-alone UCX ordering and decreased antibiotic use. ^{Reference Penney, Rodday, Sebastiani, Snydman and Doron7,Reference Caveness, Orvin, Ingram and Bouchard8} Consistent with Keintz et al, whose study found that most free-text “other” overrides in UA-RC orders were inappropriate and rarely based on symptoms, our results also highlight gaps in indication selection and opportunities to refine options, improve usability, and enhance standardization. ^{Reference Keintz, Marcelin, Rupp and Van Schooneveld9}

This study has several limitations. First, its retrospective design and reliance on automated electronic health record data extraction without manual chart review may affect data accuracy. UTI-related revisits were identified using diagnosis codes alone, which may not reflect true infection or clinical relevance. Limited documentation of symptoms and provider rationale restricted assessment of testing appropriateness and clinical decision-making. Second, antibiotic prescribing after laboratory results were not evaluated. Third, although no major practice changes were noted, the analysis is vulnerable to temporal confounding, including seasonal variation and secular trends unrelated to the UA-RC changes. Future studies incorporating chart review are needed to better evaluate clinical outcomes and appropriateness of urine testing following diagnostic workflow updates.

In conclusion, changes to ordering practices, including removal of UA-RC from select order sets, required indication selection, and stricter reflex criteria, significantly reduced ED UA and UCX utilization without impacting the rate of repeat visit for UTI within 30 days or initial antibiotic prescribing.