Urinary tract infections and recurrent urinary tract infections pose significant burdens on patients and healthcare systems. Testing and treatment strategies are increasingly important in the age of antibiotic resistance and stewardship.
To evaluate the cost effectiveness of urinary tract infection (UTI) testing and treatment strategies with a focus on antibiotic resistance.
We designed a decision tree to model four strategies (Figure 1) for managing UTIs: (1) empirical antibiotics first, followed by culture-directed antibiotics if symptoms persist; (2) urine culture first, followed by culture-directed antibiotics; (3) urine culture at the same time as empirical antibiotics, followed by culture-directed antibiotics if symptoms persist; (4) symptomatic treatment first, followed by culture-directed antibiotics if symptoms persist. To model both patient and society-level concerns, we built three versions of this model with different outcome measures: quality-adjusted life-years (QALY), symptom-free days, and antibiotic courses given. Societal cost of antibiotic resistance was modeled for each course of antibiotics given. The probability of UTI and the level of antibiotic resistance were modeled from 0-100%. We also extended the model to account for patients requiring catheterization for urine specimen collection.
In our model, the antibiotic resistance rate was based either on the local antibiotic resistance patterns for patients presenting with sporadic UTIs, or on rate of resistance from prior urine cultures for patients with recurrent UTIs. With the base case assumption of 20% antibiotic resistance, urine culture at the same time as empirical antibiotics was the most cost-effective strategy and maximized symptom-free days. However, empirical antibiotics was the most cost-effective strategy when antibiotic resistance is below 6%, while symptomatic treatment was the most cost-effective strategy when antibiotic resistance was above 80%. To minimize antibiotic use, symptomatic treatment first was always the best strategy followed by urine culture first. Sensitivity analyses with other input parameters did not affect the cost effectiveness results. When we extended the model to include an office visit for catheterized urine specimen, empirical antibiotics became the most cost effective option.
We developed models for UTI management strategies that can be interpreted for patients initially presenting with UTIs or those with recurrent UTIs. Our results suggest that in most cases, urine culture at the same time as empirical antibiotics is the most cost-effective strategy and maximizes symptom-free days. Empirical antibiotics first should only be considered if the expected antibiotic resistance is very low. If antibiotic resistance is expected to be very high, symptomatic treatment is the best strategy and minimizes antibiotic use.

Copyright © 2021 Elsevier Inc. All rights reserved.

Author