ADS 25 - An Infection Preventionist Driven Early Warning System Reduces Hospital Onset Clostridioides difficile Laboratory Identified Events

Background: The Centers for Medicare and Medicaid Services (CMS) mandates reporting of hospital-onset (HO) Clostridioides difficile (C. diff). Misidentification of community onset cases as hospital onset can be a significant source of financial burden on healthcare organizations. Development of a symptom- and risk-factor based early warning system, driven by Infection Preventionists and presented in a critical action-item format to clinical leadership daily may reduce numbers of HO C. diff Laboratory-Identified (LabID) events reported to the National Healthcare Safety Network (NHSN).

The purpose of this project is to assist Infection Preventionists (IP) in acute care settings to develop a report in their Electronic Medical Record (EMR) that will allow them to present actionable items to clinical leadership, without placing undue time burden on IP.

Methods: The report scanned output flowsheets each time it was run looking for four components to flag records: 1) An age greater than 1 year old, 2) Charting of aberrant stool quality (loose, frothy or watery), 3) Lack of a C. diff test (Polymerase Chain Reaction or screening) result, or order and 4) Patient admission to an NHSN qualifying inpatient unit for three calendar days. Records that met criteria were flagged for review by IP to check for additional risk factors such as antibiotic use, elevated white blood cells, tube feeding or proton-pump inhibitor drugs, as well as a previous history of C. diff.

Results: The report and EMR charts were reviewed by an IP and cases meeting criteria for testing were communicated to unit leadership. Testing was recommended to providers by unit leadership if appropriate.

Conclusions: Utilization of our EMR’s reports along with an action-focused reporting system led by Infection Prevention reduced HO C. diff LabID events to NHSN by 43% in 2023 over the same period in 2022 (20 cases versus 35 cases, p=0.03).