Decreasing Treatment for Postoperative Clostridioides difficile Infection in Pediatric Surgery Patients.

PURPOSE: Our hospital was an outlier for postoperative Clostridioides difficile infection (CDI) on the 01/2022 NSQIP-Pediatric Semiannual Report. We instituted a QI initiative to improve testing methods in our hospital and decrease the postoperative CDI treatment rate. METHODS: Surgical patients eligible for NSQIP-Pediatric sampling from 01/2022-07/2025 were analyzed to identify patients treated for CDI within 30 days of surgery. Prior to project initiation, screening tests were often used without confirmatory assays. In 01/2022, an initiative to increase confirmatory testing for patients screening positive for CDI was begun. The prior confirmatory assay in use was a send-out lab that required a new fecal sample and had long turn-around time (5 days). In-house EIA-based confirmatory testing was developed in 06/2024, and it was automatically deployed using the same stool sample. The turn-around time was 2-4 hours, allowing for delay of therapy until confirmation. RESULTS: We identified 16,926 NSQIP-Pediatric eligible patients, and 55 were treated for CDI. Ten patients (18%) were treated despite negative confirmatory testing. With education, confirmatory testing decreased from 82% to 62%, but CDI treatment decreased from 0.58% to 0.26% over the first 18 months. With the implementation of in-house automatic EIA testing, the CDI treatment rate decreased further to 0.07% over the subsequent 12 months. CONCLUSION: Confirmatory cytotoxin testing in an in-house automatic fashion allows accurate treatment of patients with active CDI instead of colonization. Utilization of a QI tool focused on diagnostic and antibiotic stewardship can successfully decrease system-wide healthcare resource utilization and costs, while improving patient care. Reducing inappropriate treatment of colonized patients lowered the NSQIP-P-defined CDI rate without an increase in missed or delayed diagnoses, fulminant CDI, or CDI requiring surgery.