506. The Impact of Bowel Management System (BMS) on the Incidence of Hospital-Onset (HO) Clostridium difficile Infection Laboratory-ID Events Despite Diagnostic Stewardship

BACKGROUND: Clostridium difficile infection (CDI) Laboratory identified events are reportable to CMS through the CDC’s NHSN. Diagnostic stewardship has been shown to decrease the incidence by decreasing false-positive incidence. Bowel management systems (BMS) have been associated with transient loss of tone of anal sphincter muscles that result in diarrhea. These episodes of diarrhea may be misdiagnosed as CDI due to a false-positive test result. The objective of this study was to determine whether the use of BMS has resulted in false-positive CDI Lab ID events. METHODS: We performed a retrospective review of all HO-CDI Lab ID events from October 1, 2016 to December, 31, 2017 in a 1,157-bed tertiary academic medical center. Since 2013, several interventions were implemented to decrease the incidence of CDI Lab ID events. These interventions have included: (i) enhanced environmental cleaning, (ii) CDI testing algorithm, (iii) use of hydrogen peroxide enhanced terminal cleaning of high-risk units, and (iv) computer-assisted decision support diagnostic stewardship. Poisson regression analysis was performed to compare incidence rates. A P-value of ≤0.05 was considered significant. RESULTS: A sustained low and decreasing HO-CDI incidence was observed from 2013 to 2017 (7.9, 6.0, 7.1, 6.5 and 5.2 CDI/10,000 patient days; P = 0.01). An incremental decrease was observed when comparing the annual incidence in 2016 to the YTD incidence in 2017 (6.5 vs. 5.2 CDI/10,000 patient days; P < 0.001). Comparing the five quarters before diagnostic stewardship was implemented to post implementation, the CDI incidence decreased from 6.7 to 5.2 CDI events/10,000 patient days (P = 0.009). Of the 180 HO-CDI Lab ID events that occurred post-implementation of the diagnostic stewardship, 31 (17%) were cases in which the computer-assisted alerts were overridden and may have been false positives. An additional 12 (6.7%) cases occurred in patients who had BMS in place within 48 hours and 22 (12%) had BMS in place within 1 week. CONCLUSION: Diagnostic stewardship through computer-assisted decision support is an effective method of reducing false-positive CDI Lab ID events. We found that an additional 12% of the HO-CDI are potentially false positives as these were obtained from patients who had BMS in place immediately before the positive test results. DISCLOSURES: All authors: No reported disclosures.