(117) Susceptibility Reporting and Antibiotic Prescribing for UTIs in The Inpatient Setting: A Nudge Toward Improved Stewardship

Publication Number: 1162

Background: Urinary tract infections (UTIs) are common in the inpatient setting. Although many UTI-causing pathogens are susceptible to oral beta-lactam antimicrobials, it is not feasible to directly test each potential antimicrobial on each urine sample; instead, pathogen susceptibility is extrapolated from other agents. To improve antibiotic prescribing for UTI, our tertiary care hospital added cephalexin to the susceptibility report that accompanies positive urine culture results in the electronic health record (EHR).

Objectives: To evaluate prescribing trends of cephalexin, other oral beta-lactams, fluoroquinolones, and other antibiotics for UTIs in the inpatient setting, before and after the susceptibility reporting change.

Methods: An interrupted time-series analysis was conducted. Among 1491 patients with positive urine cultures who received at least one antibiotic with a listed UTI indication during their inpatient stay, we measured the weekly prevalence (%) of patients who received each antibiotic group (cephalexin, other oral beta-lactams [amoxicillin-clavulanate, cefdinir, cefuroxime], fluoroquinolones [levofloxacin, ciprofloxacin], and ceftriaxone). The pre-intervention period was 9/2018-3/2019 and the post-intervention period was 9/2019-3/2020. We plotted prevalence of each antibiotic group over time and used segmented linear regression to estimate the intervention’s impact on each group’s time trend.

Results: At study baseline in 9/2018, the weekly prevalence of antibiotic use for UTI was 11% for cephalexin, 26% for other oral beta-lactams, 51% for ceftriaxone, and 29% for fluoroquinolones. Fluoroquinolone use decreased by 11% during the pre-intervention period (95%CI: -17%, -5%) and by 8% (95%CI: -13%, -3%) after the intervention (P for trend deflection 0.7). In contrast, during the pre-intervention period, trends were flat for cephalexin, ceftriaxone, and other oral beta-lactams (all P for pre-intervention slopes were >0.4). During the post-intervention period, use increased for ceftriaxone (6% [95%CI: 3%, 9%]); post-intervention use also increased for cephalexin (5% [95%CI: -3%, 12%]) and other oral beta-lactams (4% [95%CI -8%, 15%]), but these trends were imprecise and not statistically significant at alpha=0.05.

Conclusions: For inpatients with UTIs, the change in urine culture susceptibility reporting was associated with small increases in cephalexin and ceftriaxone use, coincident with ongoing downward trends for fluoroquinolones. Low-resource EHR-based interventions may confer considerable benefit for antimicrobial stewardship efforts in this clinical setting, and larger real-world studies are needed to replicate and contextualize these findings.