2147. Multi-modal Quality Improvement Study to Reduce C. difficile Detection and Infections in Hospitalized Children

Abstract

Abstract Background Our institution identified rising rates of Clostridiodes difficile (C. difficile) results and infections, which coincided in part with the introduction of gastrointestinal panels (GIP) testing. We aimed to reduce the rate of positive C. difficile results and hospital-acquired infections (HAI) by 20%. Methods We engaged key stakeholders to identify key drivers, develop interventions, and evaluate outcomes through several plan-do-study-act cycles. We conducted a multi-modal intervention that involved 1) provider education, 2) C. difficile clinical pathway development, 3) improved preventative cleaning measures, and 4) electronic medical record (EMR)-based GIP changes including a) added option for GIP with or without C. difficile (run on same testing platform, but C. difficile results were suppressed if not ordered), and b) built in testing restrictions with optional approval process (Table 1). Our population included all hospitalized children (< 18 years old) who had a stool test performed at a large quaternary-care children’s hospital over 30 months (2018-2020). Our primary outcomes were 1) rate of C. difficile positive results and 2) C. difficile HAI rates (defined as a positive test and symptom onset on or after day three of admission; adjudicated in real-time as part of routine surveillance) per 10,000 patient-days (PDs). Our process measure was the rate of testing capable of detecting C. difficile per 10,000 PDs. As a balancing measure, we evaluated the proportion of suppressed results that were ultimately released based on judgement of a infectious disease specialist who monitored these results in real-time for 7 months. Results We identified 2,001 stool tests performed for 1,982 encounters (Table 2). After education and EMR-based changes, we found special cause variation with a 44% decrease in the rate of positive C. difficile results and 55% decrease in HAI rates (Figures 1,2). There was a 44% decrease in testing for C. difficile, and only 2% (2/89) of suppressed results were ultimately released based on expert monitoring. The majority of patients (76%) with suppressed C. difficile results had a coinfection or were < 1 old. Conclusion Education coupled with EMR-based testing changes resulted in an impactful and sustained decrease in C. difficile detection and HAI rates. Disclosures Samuel R. Dominguez, MD PhD, Biofire DIagnostics: Advisor/Consultant|Biofire DIagnostics: Grant/Research Support|DiaSorin Molecular: Advisor/Consultant|Karius: Advisor/Consultant|Pfizer: Grant/Research Support.