Risk factors for carbapenem-resistant Enterobacteriaceae (CRE) carriage at the time of hospital admission Siddharth Mookerjee,1 Eleonora Dyakova,1 Frances Davies,1 Alison Holmes,1,2 Jonathan Otter1,2 1. Imperial College Healthcare NHS Trust; 2. Imperial College London.  j.otter@imperial.ac.uk @jonotter 1. Introduction Carbapenem-resistant Enterobacteriaceae (CRE) are an important emerging threat.1 Our hospitals in London screen all admissions to ICU, renal, vascular, and haematology, and overnight hospitalisation in the past 12 months or overseas residence are used as screening triggers for all other admissions.2 Here, we report risk factors for CRE carriage identified at the time of hospital admission. 3. Results 1050 (1.4%) of 74319 screened admissions carried CRE; less than half of these were carbapenemase producers. Risk factors for CRE carriage in multivariable analysis were previous hospitalisation at our hospitals in the past 12 months (78% of positive patients had previous hospitalisation; OR 3.3, 95% CI 2.8-3.8) and admitting speciality (Table). Compared with vascular patients, patients admitted to renal (OR 2.0, CI 1.4-2.9), haematology (OR 2.1, CI 1.5-3.2), and paediatrics (OR 1.9, CI 1.2-2.8) were more likely to carry CRE, and patients admitted to cardiology (OR 0.5, CI 0.3-0.8), the emergency department (OR 0.6, CI 0.4-0.8), oncology (OR 0.6, CI 0.4-0.9), and surgery (OR 0.4, CI 0.3-0.7) were less likely to carry CRE. Patient age was not significantly associated with CRE carriage. The number of screens and the rate of CRE cases in admitted patients increased (from 2 per 1000 patient days in December 2016 to 5 in October 2017) (Figure). However, the proportion of positive screens remained at around 1.5%, suggesting this was increased ascertainment of an existing pool of carriers, rather than an expanding pool of carriers. Please export the PowerPoint document as a PDF) and upload the PDF into the system. Do not use a font size smaller than 16. 2. Methods All patients screened for antibiotic-resistant Gram-negative bacteria in the first 24 hours or 30 days prior to a hospitalisation (to capture pre-admission screening) between December 2015 and November 2017 were included. A rectal swab was plated onto chromogenic media (ColorexTM mSuperCARBA™, E&O Laboratories, UK) and antibiotic susceptibility determined (using EUCAST antimicrobial disc susceptibility analysis).3 CRE was defined as Enterobacteriaceae resistant to either ertapenem or meropenem. Risk factors for CRE were identified using univariable and multivariable logistic regression. Figure: Trends in CRE rate and number of screens Variable CRE negative (n=73269) CRE positive (n=1050) % positive p OR (95% CI) Median age (range) 61 (0-117) 60 (0-107) NS Speciality Vascular 2049 33 1.6 Ref Haematology 3976 182 4.4 .000 2.2 (1.5-3.2) Renal 4658 167 3.5 2.0 (1.4-2.9) Paediatrics 1449 64 4.2 .003 1.9 (1.2-2.8) Emergency department 18168 152 0.8 .004 0.6 (0.4-0.8) Oncology 6224 62 1.0 .008 0.6 (0.4-0.9) Cardiology 3377 20 0.6 .005 0.5 (0.3-0.8) Surgery 5396 36 0.7 .001 0.4 (0.3-0.7) Other 7346 46 Previous hospitalisation 33899 823 2.4 3.3 (2.8-3.8) 4. Conclusions Renal, vascular, and haematology patients are higher risk specialities for CRE, and paediatrics could also be considered a high risk speciality. Previous hospitalisation is also an important risk factor for CRE identified at the time of hospital admission.4,5 We lacked information on exposure to antibiotics, which will be included in future analyses. Increased screening detected more cases, suggesting that CRE was going undetected previously. Prospective surveillance studies are required to understand the epidemiology of CRE. References: 1. Savard & Perl. Clin Microbiol Infect 2014;20:854-61. 2. PHE. Acute trust toolkit for the early detection, management and control of carbapenemase-producing Enterobacteriaceae. 3. www.eucast.org/clinical_breakpoints/. 4. Otter et al. J Antimicrob Chemother 2016;71:3556-61. 5. Torres-Gonzalez et al. PLoS ONE 2015;10:e0139883. P values and odds ratios refer to multivariable analysis. Only specialities significantly associated with a CRE positive screen are included. Disclosures: JO is a consultant to Gama Healthcare and Pfizer. Table: Variables associated with a CRE positive screen Copyright © 2018 The Authors References National Audit Office. The prevention, management and control of Healthcare Associated Infections (HCAI) in hospitals. 2009.