Urine testing is associated with inappropriate antibiotic use and increased length of stay in emergency department patients

Richard Childers; Ben Liotta; Jesse Brennan; Phoebe Wang; Jacob Kattoula; Thien Tran; Henry Montilla-Guedez; Edward M Castillo; Gary Vilke

doi:10.1016/j.heliyon.2022.e11049

Urine testing is associated with inappropriate antibiotic use and increased length of stay in emergency department patients

Heliyon. 2022 Oct 12;8(10):e11049. doi: 10.1016/j.heliyon.2022.e11049. eCollection 2022 Oct.

Authors

Richard Childers¹, Ben Liotta¹, Jesse Brennan¹, Phoebe Wang¹, Jacob Kattoula¹, Thien Tran¹, Henry Montilla-Guedez¹, Edward M Castillo¹, Gary Vilke¹

Affiliation

¹ Department of Emergency Medicine, University of California, San Diego, San Diego, CA, United States.

Abstract

Background: Exposing patients with a low probability of disease to diagnostic testing with poor test characteristics leads to false positive results. Providers often act on these false results, which can cause unnecessary evaluation and treatment. The treatment of asymptomatic bacteriuria is discouraged, but it still frequently occurs in the inpatient setting; it is less studied in the Emergency Department (ED). In this study, we examine associations between urine testing, inappropriate antibiotic use, and length of stay in discharged ED patients at risk of urinary tract infection (UTI) misdiagnosis.

Methods: A cohort of discharged ED patients at risk of UTI misdiagnosis was created by pulling visit information for patients presenting with abdominal pain, chest pain, headache, vaginal bleeding in pregnancy, and elderly females with weakness or confusion. Predictors of urine testing, and urinary tract infection treatment were determined with logistic regression analysis. A chart review of a representative sample of this cohort was then completed screening for the presence of urinary tract symptoms and urine culture results. Linear regression analysis was then used to generate an adjusted mean difference in length of stay between patients who had urine testing compared to those who did not.

Results: About a quarter of chest pain and headache patients had urine testing, while approximately 75% of abdominal pain patients, vaginal bleeding in pregnancy, and elderly females with weakness or confusion did. Except for chest pain patients, the UTI treatment rate was more than double the positive culture rate, indicating overtreatment. A diagnosis of UTI is based on a combination of UTI symptoms and positive urine cultures, yet only about 15% of patients treated for UTI met these criteria. Lastly, in all chief complaint groups, the length of stay was significantly longer-30 min or more-for those who had urine testing compared to matched controls.

Conclusions: In this observational study of patients at risk of UTI misdiagnosis, urine testing was associated with inappropriate antibiotic use and delayed discharge. There is pressure on providers to perform diagnostic testing, but in patients without specific UTI symptoms, urine testing might cause more harm than benefit.

Keywords: Antibiotic stewardship; Asymptomatic bacteriuria; Length of stay; Urinary tract infections.