1852. Rethinking Empirical Treatment for Urinary Tract Infections in the Outpatient Setting

Background Antibiograms can be useful for guiding empirical treatment. The Tufts Medical Center microbiology laboratory generates an antibiogram for the adult primary care (PC) clinic consisting of urinary isolates of E. coli to guide empirical treatment for UTI. Standard antibiograms arranged by organism are of limited utility for infections like UTI which are caused by a wide array of bacteria. Furthermore, some providers only send urine cultures from patients after clinical failure. This results in overestimation of resistance. Methods For 2 months, PC clinicians were instructed to alter practice by sending a UA with reflex to culture for all patients suspected of having UTI. A retrospective chart review collected antibiotic prescriptions, UA and culture results. We generated a combination UTI-specific antibiogram (CUSA) based on data from all urine cultures sent from this clinic. Using the CUSA we developed an empirical UTI treatment algorithm and evaluated prescribing trends before and after its implementation. Results The CUSA as compared with the E. coli urinary antibiogram for the PC clinic is shown in Table 1. Distribution of organisms is represented by Figure 1. Based on the CUSA, a treatment algorithm was developed that included preference for nitrofurantoin and TMP/SMX for patients with cystitis, and allowed for use of TMP/SMX even in pyelonephritis cases without risk factors for resistance. Cephalosporins were advised over fluoroquinolones. Of 304 patients in whom urine cultures were sent pre-implementation, 178 empirical antibiotic prescriptions were written, while 126 were written for 388 patients after implementation. Nitrofurantoin prescriptions increased (47% to 55%), TMP/SMX (19%) and cephalosporin use (4%) remained the same, and fluoroquinolone prescribing decreased (28% to 20%).Table 1: E. coli Urine Antibiogram vs. CUSA Amox Amox/Clav Cephalexin Cefpodoxime Cipro TMP/SMX Nitrofurantoin PC Urine E. coli Antibiogram (99) 54 62 82 84 82 77 98 CUSA (157) 69 78 86 97 83 86 87Figure 1. Conclusion We used a CUSA to develop a treatment guideline more reflective of our causative pathogens and institutional resistance patterns. We saw a shift in usage from antibiotics with more to antibiotics with less collateral damage. Disclosures All authors: No reported disclosures.