537. Impact of Active Surveillance Testing (AST) on Rates of Hospital-acquired Carbapenem-Resistant Enterobacteriaceae (CRE)

Background Active surveillance testing (AST) for Carbapenem-resistant Enterobacteriaceae (CRE) to identify and isolate asymptomatic carriers has been recommended to help prevent patient to patient transmission. Optimal screening population, frequency, and testing method remain a subject of debate. Methods Beginning in 2012, all clinical cultures yielding a CRE isolate in an 898-bed teaching hospital were reviewed to determine whether the isolate was hospital-acquired (HA). HA CRE rates per 10,000 patient-days were calculated. From 1/2013 to 6/2015, in-house, culture-based point prevalence surveys were performed on rectal swabs from rotating units using the CDC recommended method. 7/2015 through 8/2016, culture-based AST was outsourced to a reference laboratory and AST was expanded to include high-risk patients on admission with weekly sweeps on high-risk units. Of note, revised CLSI breakpoints were implemented by our laboratory in 7/2016, which resulted in an increase in CRE detections. Surveillance was suspended from September 2016 to January 2018 when we resumed AST utilizing in-house PCR for KPC, NDM, OXA48, IMP and VIM mechanisms. Rates of HA CRE were compared between surveillance periods. Cohorting of patients in select units, focus on hand hygiene and isolation, antibiotic stewardship, and CHG bathing were ongoing throughout all time periods. Results 510 rectal swabs in 424 patients were positive for CRE. Additional clinical cultures yielding CRE were absent in 83% of those patients, so would otherwise have gone undetected. Of those patients with both positive AST and clinical culture, 70% had a positive AST result prior to their clinical culture (range 0–997 days, average 94 days, median 14.5 days prior to clinical culture). Compared with preceding periods with no surveillance, on admission and weekly CRE AST, whether utilizing culture based or PCR based screening, was associated with significantly lower rates of HA CRE. (See Table 1). Rates of HA CRE during the initial point prevalence AST period were unchanged compared with periods with no surveillance. Community-onset CRE did not significantly change in any of the time periods monitored (Figure 2). Conclusion On admission and weekly AST was associated with a significant decrease in HA CRE in a large teaching hospital. Disclosures All authors: No reported disclosures.