Patient level risk factors for Carbapenem-Resistant Enterobacteriaceae at a low prevalence tertiary care center in the United States

This is the authors' abstract. We don't add key points for this paper.

Purpose: To characterize the clinical epidemiology of Carbapenem-Resistant Enterobacteriaceae (CRE) at Virginia Commonwealth University Health System (VCUHS). Methods & Materials: A manual chart review was performed for all CRE-positive cultures identified at VCUHS between January 2014 and May 2016. The chart review assessed patient risk factors for CRE infection such as individual carbapenem utilization, recent hospital admission, and residence in a long term care facility. Subsequently, a point prevalence survey was performed on 2 high risk units in June 2016. As part of an ongoing quality improvement initiative, patients voluntarily agreed to a peri-rectal swab that was inoculated onto HardyChromTM CRE agar for incubation in ambient air, 35 °C, for 24 hrs. Results: From January 2014 through May 2016, 39 out of 12,947 Enterobacteriaceae clinical cultures were positive with CRE (0.3%). Of these, 25 of 33, had a positive Modified Hodge Test (76%). Most specimens were from urine (N=23, 59%), followed by blood (N=4, 10%). Klebsiella spp. comprised 49% of organisms isolated; E. coli comprised 23%, Enterobacter spp. 24%, Citrobacter freundii complex and Serratia marcescens were both isolated from single specimens (3%). Patient level risk factors included recent admission(s) in 15/39 (38%) and skilled nursing or long term care residency in 13/39 (33%). Most frequent associated medical conditions were neurologic (26%), surgical complications (10%), transplantation (13%) and malignancy (15%). Of the inpatients (32/39), between 1 and 5 specimens were isolated from each of 18 different units without a pattern suggesting clustering. Eighteen of 33 patients with available pharmacy data were exposed to a carbapenem in the 30 days preceding their positive culture (55%) with an average duration of 8 days. The point prevalence survey tested 19 of 23 surgical ICU patients and 14 of 22 medical ICU patients. All of the surveillance swabs were negative for CRE. Conclusion: These data suggest that the CRE at VCUHS are imported and/or individually developed at the patient level rather than transmitted within the hospital.