Urinary tract infection (UTI) is the second most common bacterial infection leading to hospitalization, accounting for 40% of nosocomial infections. Despite the high prevalence of UTI, variability in diagnostic testing and treatment of UTI among hospitalized patients is common and can lead to inappropriate antibiotic use and subsequent antibiotic resistance. We therefore sought to determine the appropriateness of urine culture for UTI and its impact on antimicrobial prescribing.


We randomly selected patients admitted to a large academic center between February 2008 and February 2009 who had urine cultures obtained during the hospital stay. Patients were excluded if they were admitted to intensive care, had a major urinary procedure (e.g., renal transplant, diversion, or stents), were actively being treated for a UTI at the time of admission, were on empiric treatment > 48 hours prior to urine collection while hospitalized, or were obstetrics patients. Retrospective chart review was independently performed by 2 physicians to determine the presence of signs or symptoms of a UTI, presence of a urinary catheter, antibiotic administration, and urine culture results. Determination of the appropriateness for sending urine cultures was compiled from national and professional society guidelines (i.e., from the Centers of Disease Control and Prevention and the Infectious Disease Society of America).


Of the 210 patients included, 97 patients (46%) had an appropriate reason documented to obtain a urine culture. The majority of these (72%) had fever. Urinary symptoms were infrequent, occurring in 22%. In 113 patients (53.8%), no guideline‐accepted criterion for obtaining a urine culture was found. Of these 113 patients, 46 (41%) had no documented indication for culture, whereas 67 patients (59%) had documented reasons that were not consistent with guidelines, including orthopedic procedures (21%) and altered mental status without a urinary catheter (14%). Of all 210 patients, 84% had negative urine cultures. Culture negativity was similar regardless of the presence or absence of indications supported by the guidelines (84% vs. 84%). Of the culture‐negative patients, 45% were on antibiotics within 48 hours prior to urine collection. More than 10% of culture‐negative patients were started on antibiotics for UTI within 72 hours after culture. The kappa statistic on indications for culture was 0.89, indicating excellent interrater agreement.


In more than half of hospitalized patients, urine cultures are obtained outside of accepted criteria, often being sent for reasons other than urinary symptoms. In these scenarios, complicating factors included insufficient supporting data (orthopedic procedures) or nonspecific symptoms (altered mental status), which might include UTI in the differential diagnosis. Urine cultures infrequently generated new antibiotic use, perhaps because of high rates of preexisting antimicrobial use. Guidelines relevant to the hospitalized patient are urgently needed.


S. Hartley ‐ none; S. Valley ‐ none; L. Kuhn ‐ none; A. Malani ‐ none; L. Washer ‐ none; T. Gandhi ‐ none; C. Chenoweth ‐ none; S. Saint ‐ Institute for Healthcare Improvement, Michigan Health and Hospital Association, numerous hospitals and nonprofit health care organizations and medical societies, honoraria for speaking; S. Flanders ‐ Centers for Disease Control and Prevention, Institute for Healthcare Improvement, research funding, Antimicrobial Stewardship Faculty