Validation of a Clinical Decision Rule to Identify Risk Factors Associated With Multidrug-Resistant Urinary Pathogens in the Emergency Department

Abstract

Background: Antimicrobial resistance remains a significant obstacle for clinicians when treating patients presenting to the emergency department (ED) with urinary tract infections. Objective: The goal of the proposed study was to validate a previously developed clinical decision rule identifying risk factors for multidrug-resistant (MDR) urinary pathogens. Methods: We conducted a validation study of a previously published clinical decision rule to identify patients with MDR urinary pathogens using a cohort from an urban academic center ED with annual census over 80 000. Using our previously identified clinical risk factors, we determined the sensitivity, specificity, positive likelihood ratio (+LR), and negative LR (−LR) to estimate measures of precision of our clinical decision rule in the validation cohort. Results: Factors associated with MDR urinary pathogen included sex, recent hospitalization, nursing home residency, and catheter placement. Using our previously defined threshold of greater than 1 risk factor, the adjusted model in the validation cohort identified that only nursing home residency was associated with positive MDR pathogen (adjusted odds ratio = 4.13; 95% CI = 1.95-8.77). The clinical decision rule in the validation cohort yielded a sensitivity of 56.4%, specificity of 66.3%, +LR of 1.7, and −LR of 0.7. Conclusion and Relevance: Our clinical decision rule to identify patients at risk for MDR urinary pathogens was unable to be validated in the setting of different antimicrobial resistance patterns. Future studies should evaluate an improved clinical decision rule identifying risk factors associated with MDR pathogens that performs well in varying patient populations.