A qualitative analysis of diagnostic testing, antibiotic selection, and quality improvement interventions for uncomplicated urinary tract infections.

• Main Authors: Mark Pinkerton, Jahnavi Bongu, Aimee James, Jerry Lowder, Michael Durkin
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2020-01-01
• Series: PLoS ONE
• Online Access: https://doi.org/10.1371/journal.pone.0238453

<h4>Background</h4>Uncomplicated urinary tract infections (UTIs) can often be diagnosed based solely on symptoms and should be treated with a short course of narrow spectrum antibiotics. However, clinicians often order urine analyses and prescribe long courses of broad spectrum antibiotics.<h4>Objective</h4>The objectives of our study are: 1) Understand how primary care providers and residents clinically approach UTIs and 2) to understand specific opportunities, based on provider type, to target future antibiotic stewardship interventions.<h4>Design and participants</h4>We conducted semi-structured qualitative interviews of community primary care providers (n = 15) and internal medicine residents (n = 15) in St. Louis, Missouri from 2018-2019. A 5-point Likert scale was used to evaluate participant preferences for possible interventions. Interviews were transcribed, de-identified, and coded by two independent researchers using a combination inductive and deductive approach.<h4>Key results</h4>Several common themes emerged. Both providers and residents ordered urine tests to "confirm" presence of urinary tract infections. Antibiotic prescription decisions were often based on historical practice and anecdotal experience rather than local susceptibility data or clinical practice guidelines. Community providers were more comfortable treating patients over the phone than residents and tended to prescribe longer courses of antibiotics. Both community providers and residents voiced frustrations with guidelines being difficult to easily incorporate due to length and extraneous information. Preferences for receiving and incorporating guidelines into practice varied. Both groups felt benchmarking would improve prescribing practices but had reservations about implementation. Community providers preferred pragmatic clinical decision support systems and nurse triage algorithms. Residents preferred order sets.<h4>Conclusions</h4>Significant opportunities exist to optimize urinary tract infection management among residents and community providers. Multifaceted interventions that include provider education, synthesis of guidelines, and pragmatic clinical decision support systems are needed to improve antibiotic prescribing and diagnostic testing; optimal interventions to improve UTI management may vary based on provider training level.