2084. Implementation of a Diagnostic Stewardship Algorithm by Bedside Nurses to Reduce Unnecessary Urinary Cultures in Hospitalized Adult Patients

BACKGROUND: Urine cultures (UCx) are often ordered in patients without symptoms of urinary tract infection. A pilot study was conducted to assess the impact of a nurse-driven UCx diagnostic stewardship intervention for adult inpatients. METHODS: We interviewed eight nurses to determine the feasibility of a nurse-driven UCx stewardship intervention. Based on their feedback, an algorithm with appropriate indications for UCx was developed (Figure 1) and approved by physicians and nurses for piloting on a 24-bed medicine unit at The Johns Hopkins Hospital. UCx orders/100 patient-days (PD) were trended with statistical process charts in the intervention and a control unit. Nurses used the algorithm to guide discussions with ordering providers and to suggest instances where UCx may be unnecessary (“intervention”). Nurses were educated on an antibiotic (abx) use safety and appropriate testing during live sessions prior to algorithm implementation. Two study team members reviewed all UCx ordered in the intervention unit 12 months before and 6 months after the intervention for appropriateness based on algorithm criteria. Feedback on UCx order appropriateness and case-based discussion were provided to nurses via in-person meetings post intervention. Data were compared using the χ (2) or the Mann–Whitney test as appropriate. The rate of UCx orders before and after the intervention were compared using a standard incident ratio (IRR). RESULTS: With algorithm implementation, the mean rate of UCx orders/100 PD decreased from 2.7 to 1.8 (39% decrease) in the intervention unit (IRR 0.61, 95% confidence intervals (CI) 0.45–0.82, P = .16). Mean UCx order rates in the control unit were 2.49 and 2.99, respectively (Figure 2). Characteristics of patients reviewed for appropriateness were similar between the two study periods: median age 63 (IQR 39, 74) vs. 56 (IQR 45, 76), female sex 65% vs. 61%, on hemodialysis 7% vs. 11%, urinary catheter present 20% vs. 29%. The proportion of inappropriate UCx decreased from 59% (98/165) to 50% (32/64) (P = 0.16). There were 8 and 1 cases of asymptomatic bacteriuria inappropriately treated in the pre- and post-intervention periods, respectively (42 and 7 abx days). CONCLUSION: With the appropriate training and tools, nurses can steward UCx and reduce unnecessary testing and abx use [Image: see text] [Image: see text] DISCLOSURES: Sara E. Cosgrove, MD, MS, Basilea: Consultant; Theravance: Consultant.