Development of Predictive Models to Inform a Novel Risk Categorization Framework for Antibiotic Resistance in E. coli-Causing Uncomplicated Urinary Tract Infection

Clinical Infectious Diseases, 2024

An Analysis Group team, in collaboration with researchers from the University of Pittsburgh, the University of Maryland School of Medicine, and GSK, has coauthored a study detailing a novel framework that clinicians can use to inform antibiotic prescribing for patients with E. coli-caused uncomplicated urinary tract infections (uUTIs), the most common outpatient infection in the US. The study is the first to use predictive modeling to categorize patient risk profiles for antimicrobial resistance (AMR) as low, moderate, and high risk across four commonly prescribed antibiotic classes for uUTI.

Despite guideline recommendations from the Infectious Diseases Society of America (IDSA) on appropriate first-line antibiotic agents for the treatment of uUTI, 86% of patients are prescribed antibiotic agents that are not considered first-line therapies. These inappropriate prescribing practices contribute to the growth of AMR, treatment failure, persistent uUTI symptoms, adverse patient health outcomes, and increased health care costs. This study was conducted to help provide clinicians with more reliable and efficient approaches for supporting patients with uUTI.

A research team including Managing Principals Lisa Pinheiro and Jimmy Royer and Managers Kalé Kponee-Shovein and Chi Gao drew on electronic health data from 87,487 patients with confirmed E. coli-related uUTI in the US. The investigators then developed and validated predictive models to estimate probabilities of E. coli AMR to four commonly prescribed classes of antibiotics for uUTIs (nitrofurantoin, trimethoprim-sulfamethoxazole, β-lactams, and fluoroquinolones). The predictive models were found to outperform those in existing literature, and the risk categorization framework provided strong separation of risk between patients with truly resistant and truly not resistant E-coli isolates.

The study appeared in Clinical Infectious Diseases in April 2024. Funding for this study was provided by GSK.

Read the study

Authors

Shields RK, Cheng WY, Kponee-Shovein K, Indacochea D, Gao C, Kuwer F, Joshi AV, Mitrani-Gold FS, Schwab P, Ferrinho D, Mahendran M, Pinheiro L, Royer J, Preib MT, Han J, Colgan R