P159: Identifying the cause for inappropriate urine cultures in a Canadian urban academic emergency department

Abstract

Introduction: Inspired by the Choosing Wisely® campaign, St. Michaels Hospital (SMH) launched an initiative to reduce unnecessary tests, treatments and procedures that may cause patient harm. Stakeholder engagement identified inappropriate ordering of urine culture & sensitivities (C&S) in the emergency department (ED) as a focus area. Inappropriate urine C&S increase workload, healthcare costs and detection of asymptomatic bacteriuria which can lead to unnecessary antibiotics. The project’s purposes were to describe the scope of inappropriately ordered urine C&S in the SMH ED and to conduct a root-cause analysis to inform future quality improvement interventions. Methods: Criteria for determining appropriateness was developed a priori using evidence-based guidelines from the University Health Network together with additional literature review. A retrospective chart review was performed on all urine C&S ordered in the ED from Jun 1 Aug 30, 2016. Each chart was reviewed for order appropriateness, demographic information and ordering provider. All inappropriate urine C&S were reviewed to identify root causes which were then grouped into common themes. A pareto chart was constructed to analyze the frequency of causes. Results: Of 425 urine C&S ordered, 75 (17.7%) were inappropriate. The top 3 reasons were: inappropriate urosepsis work-ups (53%), order processing errors (17%) and inappropriate work-ups for weakness (16%). Inappropriate urosepsis work-ups were defined as urine C&S that were ordered empirically despite there being a clear focus for infection elsewhere (i.e. cough, cellulitis) and in the absence of urinary symptoms. Order processing errors were defined as urine C&S which were sent despite there being no documented order. Inappropriate testing was more likely to occur overnight, in females and when a urine routine and microscopy was not ordered prior to C&S. 29% of patients with inappropriate C&S received antibiotics. Conclusion: 17.7% of urine C&S ordered in the SMH ED during the 3-month study period were inappropriate. The top cause was septic patients who were empirically tested despite having another source for infection identified from the outset. A possible reason for this is the recent ED emphasis on early recognition of sepsis which may encourage early use of antibiotics and empiric urine C&S. One question to resolve is whether a 17.7% overutilization rate is sufficient to make it a target for change. Interventions designed to reduce inappropriate urine C&S may inadvertently increase the number of missed cultures in patients admitted with sepsis not yet diagnosed. Next steps involve discussions between the ED, Internal Medicine, Infectious Disease and Microbiology, and patient partners to identify patient-centered change ideas and sustainable strategies. This may involve establishing guidelines for ordering urine C&S and incorporating lab services to provide oversight into urine C&S processing.