185 Impact of Eliminating a Urinalysis With Reflex to Urine Culture on Urine Testing Patterns in the Emergency Department

Abstract

As antibiotic resistance has increased as a public health threat, attention has focused on interventions aimed at reducing unnecessary antibiotic prescribing. This includes the overtreatment of asymptomatic bacteriuria (ASB), due to unnecessary urine testing with subsequent diagnosis of urinary tract infection (UTI) despite absence of clinical symptoms. With ASB common in some populations, such as older adults, stewardship programs should seek to reduce unnecessary urine testing on asymptomatic patients. A reflex urinalysis (UA) (defined as WBC> 5/hpf = urine culture) order is commonly used in the emergency department (ED) setting due to perceived operational efficiency benefits; however, this can result in unnecessary diagnosis and treatment of UTI. The objective of the study was to evaluate the impact of removing reflex UA from the computerized physician order entry system (CPOE) on overall urine testing patterns in an academic ED. We conducted a retrospective study of all patients who received any urine test in the ED within the six month pre- and post-intervention periods. A one-month washout period was used surrounding reflex UA removal. The intervention consisted of a notification to providers via e-mail who frequently used reflex UA (academic detailing) and removal of reflex UA from personal and ED preference lists and order sets (systems approach). A positive culture was defined as growth of one or more pathogenic organisms at 104 CFU/ml or greater. Patient demographic and urine testing data were extracted from the electronic health record and analyzed using STATA SE15. Primary outcomes were the rates of total cultures and contaminated cultures per 1000 ED visits. Patient safety was assessed by analysis of any missed urine cultures in pyelonephritis or urosepsis cases. Urine culture cost was estimated from the Medicare Clinical Laboratory Fee Schedule. The study included 17,061 patients with 20,029 urine tests (mean age, 45.7 years; 85.0% white; 60.8% female). The six-month average urine culture order rate fell from 63.2 per 1000 ED visits to 35.3 per 1000 post-intervention, a 41% overall reduction (P<0.001). The rate of contaminated cultures was 16.0 per 1000 ED visits before the intervention compared to only 6.9 after (P<0.001) (see Figure). Of those diagnosed with pyelonephritis or sepsis, 72.8% received a urine culture before intervention in contrast to 56.9% after intervention (P<0.001). Cost savings from reduction in urine cultures was estimated to be $10,830. Removal of reflex UA from the CPOE led to significant reductions in the rates of total cultures performed as well as contaminated cultures. However, there was also a significant decrease in urine cultures obtained for conditions where a culture is typically recommended. Ordering systems greatly impact urine testing practices and could have future implications in reducing overtreatment of ASB. Any intervention or change to eliminate reflex UA orders should be accompanied with robust provider education emphasizing the need to separately order urine cultures when indicated.