Background: Data for hospital antibiograms are typically compiled from all patients, regardless of disposition, demographics and other comorbidities.
Objective: We hypothesized that the sensitivity patterns for urinary pathogens would differ significantly from the hospital antibiogram in patients that were discharged from the emergency department (ED).
Methods: We evaluated a retrospective cohort of all adult patients with positive urine cultures treated in the 2016 calendar year at an inner-city academic ED. Positive urine cultures defined by our institution's microbiology department. Investigators conducted a structured review of an electronic medical record (EMR) to collect demographic, historical and microbiology records. We utilized a one-sample test of proportion to compare the sensitivity of each organism for discharged patients to the hospital published antibiogram. Alpha set at 0.05.
Results: During the study period, 414 patients were discharged from the ED and found to have positive urine cultures; 20% age > 60 years old, 85% female, 79% Hispanic, 33% diabetic. The most common organisms was E. coli (78%). E. coli was sensitive to Trimethoprim-Sulfamethoxazole for 59% vs. 58% in our antibiogram (p = 0.77), Ciprofloxacin 81% vs. 69% (p < 0. 001), Nitrofurantoin 96% vs 95%; (p = 0.25). K. pneumoniae was sensitive to Trimethoprim-Sulfamethoxazole 87% vs. 80% in our antibiogram (p = 0.26), Ciprofloxacin 100% vs. 92% (p = 0.077), Nitrofurantoin 86% vs 41% (p < 0.001).
Conclusions: For our predominantly Hispanic study group with a high prevalence of diabetes, we found that our hospital antibiogram had relatively good value in guiding antibiotic therapy though for some organism/antibiotic combinations sensitivities were higher than expected.
Keywords: Antibiogram; Antibiotic resistance; Antibiotic sensitivity; Diabetic; UTI; Urinary tract infection.