Background: Unnecessary testing for Clostridioides difficile can lead facilities to overreport laboratory-identified (LabID) events. Because false-positive LabID tests could dilute infection control resources, we developed best practice alerts (BPAs) in the electronic health record, educational materials as well as a follow-up system to help reduce unnecessary testing and, therefore, reduce false-positive results. Methods: Three BPAs were initiated in late August, 2018. Alerts fired when clinicians tried to order repeat C. difficile testing after a positive result, testing within 24 hours of laxative administration and to order a multiplex PCR panel for GI pathogens >48 hours after admission. The GI multiplex PCR test consists of 21 targets, including C. difficile, but it allows for testing solid stool. All alerts gave suggestions for how to proceed (ie, not test for cure from previous positive, wait until laxatives wear off, or call for approval before GI panel) but could be bypassed by clinicians. Educational emails and signage were distributed to all house staff and clinicians in all clinical areas at the start of the program. For each bypassed BPA, infection control physicians contacted the ordering clinician by email or phone to explain why testing was not advised. Results: Between September 5, 2018, and April 23, 2019, 1,217 BPAs were issued: 634 in first half and 583 in the second half. Of these, 268 (22%) were bypassed by clinicians (Fig. 1). There was no significant decrease in bypassing BPAs. In the first half of the intervention, 22% of BPAs were bypassed (141 of 634). In the second 4 months, 22% of BPAS were still bypassed (127 of 583; P = .85). Of the 40 ordering services, 8 had no bypassed BPAs in the first half and 9 had no bypassed BPAs in the second half. Conclusions: Educating providers and following up after bypassed BPAs did not decrease the number of bypassed BPAs. Although fewer BPAs were issued in the second half of the intervention, more analysis is needed to understand whether this decrease is significant. In this study, 268 unnecessary C. difficile tests were ordered over 8 months.

Funding: None

Disclosures: None