525. The Impact of Switching to Molecular Testing on Clostridium difficile Infection Rates: Large-Scale Assessment Using an Interrupted Time Series Poisson Regression Approach. (26th November 2018)

• Record Type: Journal Article
• Authors: Jabur, Tiago Barbieri Couto; Ilies, Iulian; Baker, Arthur W; Anderson, Deverick J; Benneyan, James
• Is Part Of: Open forum infectious diseases. Volume 5(2018)Supplement 1
• Page Start: S194
• Page End: S194
• Publication Date: 2018-11-26
• Subjects: Communicable diseases -- Periodicals; Medical microbiology -- Periodicals; Infection -- Periodicals; 616.9
• DOI: 10.1093/ofid/ofy210.534
• Languages: English
• ISSNs: 2328-8957

Abstract

Abstract: Background: Clostridium difficile is the most common cause of hospital-acquired infections in the United States, affecting over 500, 000 patients per year at a cost of nearly $5 billion. The reported incidence of C. difficile infections (CDIs) has increased in recent years, partly due to broad adoption of polymerase chain reaction (PCR) testing replacing enzyme-linked immunosorbent assay (ELISA) methods. Our aim was to assess the contribution of this change on reported CDI incidence using a large-scale empirical data set. Methods: We retrospectively analyzed 8 years of CDI surveillance data (2009–2016) collected from 47 hospitals in the Duke Infection Control Outreach Network. During this period, 24 hospitals switched to PCR testing, 10 used ELISA throughout, and 13 used PCR throughout. We used interrupted time series analysis to quantify the relative change in incidence rate (IRR) of CDIs due to the switch from nonmolecular (ELISA) to molecular (PCR) testing. Data were aligned across hospitals at their interruption point, set at the reported test change date or nearest available measurement. Individual hospital and network-wide estimates of the PCR-over-ELISA IRR were determined through Poisson regression, controlling for total patient days, proportion of intensive care unit patient-days as a proxy for acuity, background trends, and previously detected clusters. Results: Average monthly CDI rates significantly increased after the test change from 11.7 to 26.8 per 10, 000 patient-days in hospitals that switched to PCR testing. A similar difference was observed between ELISA-only and PCR-only hospitals, which averaged 12.7 and 21.0 CDIs per 10, 000 patient-days, respectively. Regression analysis yielded hospital-specific test change IRRs ranging from 0.70 (95% confidence interval [CI]: 0.48–1.02) to 3.64 (CI: 2.77–8.46) (Figure 1) and a network-wide IRR of 1.79 (CI: 1.73–1.90). Results also found an increasing background trend of 0.9 CDIs per 10, 000 patient-days per year (CI: 0.7–1.2) (Figure 2), as well as a significant effect of known clusters (IRR of 1.56, CI: 1.48–1.65). Conclusion: Hospitals that switched to molecular testing experienced an average post-change increase of 80% in reported CDI rates, similar to that observed during known cluster periods. Disclosures: All authors: No reported disclosures.