2340. Diagnostic Stewardship: Survey of Urine Culturing and C. difficile Testing Practices Amongst Oregon Microbiology Labs

Abstract Background Testing for urinary tract infection (UTI) and Clostridiodes difficile infection (CDI) poses diagnostic and antimicrobial stewardship challenges. Both diagnoses hinge on local microbiology laboratory algorithms. For UTI testing, the definition of “abnormal” urinalysis and the use of reflex urine cultures, both of which alter the frequency of bacteriuria detection, likely differs between laboratories. For CDI, pretest probability, choice and sequence of diagnostic tests are likely variable and impact the chances of accurate diagnosis. Methods To understand laboratory practices and determine variations in local testing algorithms, we deployed a self-administered survey to microbiology laboratories serving Oregon healthcare facilities via SurveyMonkey in September 2018. Responses were collected through April 2019. We analyzed a subset of questions focused on UTI and CDI diagnosis. Results Of 51 surveyed laboratories, response rate was 86% (n = 44). 91% of respondents (n = 40) process bacterial cultures. 47.5% (n = 19) primarily perform urine culture when ordered, whereas the remainder primarily perform cultures in a reflex algorithm when ordered (n = 12; 30%) or a reflex algorithm automatically (n = 9; 22.5%) (Figure 1). The definition of an abnormal urinalysis varied widely (Figure 2). 15% (n = 6) of laboratories reported considering changes to their workflow; two cited a goal of reducing unnecessary testing. Of the 32 laboratories that perform in-house C. difficile testing, the assays and sequence in which they were implemented in testing algorithms varied substantially (Figure 3) and most commonly included NAAT testing. Seven (21.8%) laboratories reported recently changed practices; these changes did not favor any particular algorithm. 84.2% (n = 32) reported stool rejection criteria to limit unnecessary testing, but these criteria varied (Figure 4). Conclusion Wide variation exists in laboratory workflows for UTI and CDI diagnoses in Oregon, suggesting lack of consensus on optimal practices. Encouragingly, multiple labs described recently implemented or planned interventions to reduce unnecessary testing for both infections. This snapshot will inform statewide education and interventions to optimize testing and help prevent patient and population harm. Disclosures All authors: No reported disclosures.