Development of a Pharmacist-Led Emergency Department Antimicrobial Surveillance Program
Methods
This project took place at the RRVAMC, a 229-bed tertiary academic medical center that serves > 60,000 patients annually. The RRVAMC ED has 20 beds and received about 29,000 visits in 2014. Patients were eligible for initial evaluation if they had a urine culture collected in the ED within the 91-day period from September 1, 2015 to November 30, 2015. Patients were included for data analysis if it was documented that they were treated for actual or clinically suspected, based on signs and symptoms, uncomplicated UTI, complicated UTI, or UTI with pyelonephritis. Patients did not need to have a positive urine culture for inclusion, as infections could still be present despite negative culture results.6 Patients with positive cultures who were not clinically symptomatic of a UTI and were not treated as such by the ED provider (ie, asymptomatic bacteriuria) were excluded from the study.
Data collection took place via daily chart review of patient records in both the Computerized Patient Record System and Decentralized Hospital Computer Program medical applications as urine cultures were performed. Data were gathered and assessed by a postgraduate year-2 internal medicine pharmacy resident on rotation in the ED who reviewed cultures daily and made interventions based on the results as needed. The pharmacy resident was physically present within the ED during the first 30 days of the project. The pharmacy resident was not within the direct practice area during the final 61 days of the project but was in a different area of the hospital and available for consultation.
Primary data collected included urine culture results and susceptibilities, empiric antimicrobial choices, and admission status. Other data collected included duration of treatment and secondary antibiotics chosen, each of which specifically evaluated those patients who were not admitted to the hospital and were thus treated as outpatients. Additional data generated from this study were used to identify empiric antibiotics utilized for the treatment of UTIs and assess for appropriate selection and duration of therapy within this institution.
Results
During the study period, 722 urine cultures were collected in the ED and were included for initial evaluation. Of these, 127 were treated by the ED provider pursuant to one of the indications specified and were included in the data analysis. Treatment with an antimicrobial agent provided adequate coverage for the identified pathogen in 112 patients, yielding a match rate of 88%. As all included cultures were collected in suspicion of an infection, those cultures yielding no growth were considered to have been adequately covered. 
Nearly half (45%) of treatment plans included a fluoroquinolone. Of those treated on an outpatient basis, fluoroquinolones were even more frequently used, comprising 50 of 82 (61%) courses. Ciprofloxacin was the most frequently used treatment, used in 39 of the 82 outpatient regimens (48%). Cephalexin was the second most common and was used in 14 outpatient regimens (17%), followed by levofloxacin (15%) (Figure 1). 
Mismatched cultures, or those where the prescribed antibiotic did not provide adequate coverage of the identified pathogens based on susceptibilities, occurred at a rate of 12%. Follow-up on these cultures was determined largely by the patient’s admission status. The majority of mismatched cultures were addressed by the inpatient team (10/15) upon admission. 
