A Large Retrospective Analysis of Urinary Tract Infection Encounters

Objectives: To retrospectively evaluate urinary tract infection (UTI) encounters at a tertiary care academic medical center for the risks of antibiotic administration or not, including antibiotics-related complications and progression to sepsis within seven days.

Methods: The Discovery Data Mart administered by the UCLA Office of Health Informatics and Analytics was queried to identify direct patient encounters for acute cystitis based on ICD codes from January 1, 2013 through July 1, 2021. Patient demographics, visit type, and clinical data were obtained. Only women with uncomplicated UTI were eligible for inclusion. Women with complicated UTI or septicemia at presentation were excluded.

Results: 168,358 encounters met inclusion criteria. Outpatient visits represented 87.7% of encounters; emergency department and inpatient encounters represented 9.9% and 2.4%, respectively. Just over half (54.9%) of these resulted in an antibiotic prescription. (Table 1) Urine culture positivity was exceptionally low overall, but significantly lower in women who were prescribed antibiotics (3.2% versus 4.8%, p< 0.001). Rates of progression to any form of sepsis within seven days were very low but significantly higher in women who did not receive antibiotics (0.09% versus 0.14%, p=0.025). Rates of urosepsis (defined as positive blood culture and urine culture) and pyelonephritis, however, did not differ with antibiotic administration (p=1.00 and 0.912, respectively). Women who received antibiotics had significantly higher rates of future visits for dermatologic (0.5% versus 0.4%, p< 0.001), hypersensitivity (0.1% versus 0.05%, p< 0.001), and genitourinary (1.4% versus 1.2%, p< 0.001) adverse events than those who did not. In patients prescribed accepted first-line antibiotics for UTI (e.g., nitrofurantoin, trimethoprim/sulfamethoxazole), we observed a lower risk than in those prescribed non-first-line agents (e.g., fluoroquinolones) of requiring future visits to manage gastrointestinal (7.4% vs. 12.5%, p< 0.001), renal (0.3% vs. 1.5%, p< 0.001), cardiovascular (0.1% vs. 0.4%, p=0.003) and pulmonary (0.7% vs. 2.9%, p< 0.001) complications. Of the 0.14% of outpatients who progressed to sepsis within 7 days, only 17% of those in whom urine cultures were obtained (4% of all patients with sepsis) had positive urine cultures consistent with urinary infection; rates of urine culture positivity did not differ when stratified by whether antibiotics were prescribed at initial presentation (p=0.122).

Conclusions: Half of encounters resulted in antibiotics prescriptions, however there were no differences in rates of progression to sepsis, urosepsis, or pyelonephritis within 7 days among outpatient women despite significantly more antibiotics-associated adverse events. Providers should wait to confirm urine culture positivity prior to prescribing antibiotics given the very low rates of urine culture positivity and progression to sepsis. While this study is limited by the accuracy of ICD coding, our results suggest that most uncomplicated UTIs are diagnosed inaccurately and that higher rates of sepsis when antibiotics are not given may reflect poor recognition of other infectious conditions.