Development and implementation of a nurse-led clinical decision support tool for urinary tract infection

Ordering urine cultures in the absence of urinary symptoms contributes to unnecessary antibiotic prescribing for urinary tract infection (UTI). The aim of this study was to develop and implement an electronic health record (EHR)-integrated, nurse-led clinical decision support (CDS) tool for UTI symptom triage to support outpatient antibiotic stewardship.

Exploratory sequential mixed-methods study.

Urban academic family medicine practice.

Eleven nurses.

In the qualitative phase, workflow mapping and semi-structured interviews informed development of an EHR-integrated CDS consisting of a symptom score calculator and triage algorithm. In the quantitative phase, implementation was evaluated using a pre-post design with interrupted time series analysis over 15 months. Implementation outcomes included adoption, fidelity, and usability. Clinical outcomes included symptom documentation, urine testing practices, and antibiotic prescribing. Safety outcomes included pyelonephritis within 30 days.

Qualitative analysis identified four barriers: incomplete symptom documentation, reflexive urine testing, medicolegal concerns about missing pyelonephritis, and patient pressure for antibiotics. The CDS was adopted by all nurses, with 77.8% fidelity and high usability. Documentation of ≥2 UTI-specific symptoms increased at implementation with sustained improvement (slope + 2.4%/month; 95% CI, 0.1-4.7). Urine cultures without microscopy showed sustained reduction (slope -5.5%/month; 95% CI, -7.9 to -3.2). Antibiotic prescribing for asymptomatic bacteriuria decreased (20% to 3%; P = .003) with no increase in pyelonephritis.

Nurse-led, EHR-integrated clinical decision support for UTI triage was associated with sustained improvements in symptom documentation, reduced unnecessary urine culture ordering, and decreased antibiotic treatment of asymptomatic bacteriuria without compromising safety.