Abstract Background Bloodstream infections (BSIs) represent a poor prognosis in living-donor liver transplant recipients (LDLT-Rs). Some patients develop recurrent BSIs. We evaluated the impacts of ER-BSIs on outcomes in LDLT-Rs. Methods All LDLT-Rs with follow-up data between January 2008 and December 2016 were included. Early BSIs (E-BSIs) defined as BSI events within 2 months after LDLT. ER-BSI was defined as new-onset BSI within 2 months due to another pathogen ≥48-hour interval, or relapse of BSI by the same pathogen ≥1-week interval with negative cultures in between. Logistic regression model was used to analyze risk factors for 1-year mortality. An associated factor of E-BSI and ER-BSI were also evaluated. Results Among 727 LDLT-Rs, 108 patients had 149 events with 170 isolated pathogens of E-BSI. Twenty-eight patients (25.9%, 28/108) experienced ER-BSI. Enterococcus (37.6%) was the most common pathogen. Intra-abdominal infection was the most common focus in the first episode of E-BSI and even significantly more common in ≥ second (59.3% vs. 82.9%, P = 0.007). Intra-abdominal and/or biliary complications were risk factors for both E-BSI and ER-BSI. Whereas high MELD score, longer cold ischemic time and longer recipient operative time were associated with E-BSI, longer post-transplant intensive care unit stay and longer donor operative time was associated with ER-BSI. 1-year survival rates of patients with or without single event of E-BSI were 81.3% and 92.4%, respectively. Patients having ER-BSI showed significantly low 1-year survival rates of 28.6% (Figure 1). ER-BSI was the most relevant risk factor for 1-year mortality (adjusted OR = 8.26; 95% CI = 4.30–15.88). Conclusion LDLT-Rs with ER-BSI showed very low survival rates accompanying with intra-abdominal and/or biliary complications. Clinicians should aware to prevent recurrence of BSI focusing on intra-abdominal complications in order to improve clinical outcomes of LDLT-R. Disclosures All authors: No reported disclosures.

Abstract Background Infections caused by multidrug-resistant (MDR) organisms are associated with poorer clinical outcomes and higher economic burden. However, there has been limited data on the impact of MDR infection on the performance status of patients. Methods Patients with bloodstream infections by S. aureus, E. faecium, E. coli, K. pneumoniae, P. aeruginosa, and A. baumannii have been identified prospectively as a part of a multicenter nationwide surveillance for antimicrobial resistance. Medical records of the patients enrolled from July 2015 through December 2016 were reviewed for demographic, clinical, microbiologic characteristics, and patient outcome. MDR was defined as MRSA, VRE, and nonsusceptibility to one or more agents in three or more different classes of antibiotics for Gram-negative bacteria. Performance status was evaluated by Eastern Cooperative Oncology Group (ECOG) Performance Status before admission and at discharge. Primary outcome was any decline in ECOG at discharge. Multiple logistic regression was used to identify independent risk factors for ECOG decline. Results A total of 19 hospitals participated to the network. The numbers of subjects were 410 for S. aureus, 392 for E. faecium, 708 for E. coli and K. pneumoniae, and 678 for P. aeruginosa and A. baumannii. In univariate analysis, bacteremia by MDR organisms was associated with ECOG decline only in patients with P. aeruginosa (18.4% vs. 10.3%, OR 1.962, 95% CI 1.132–3.399) and A. baumannii (27.6% vs. 11.8%, OR 2.834, 95% CI 1.328–6.045) infections. Patients with MDR K. pneumoniae infection had lower risk of ECOG decline (6.6% vs. 15.8%, OR 0.378, 95% CI 0.183–0.780). Multivariable analysis also showed that infection by MDR organism was independently associated with ECOG decline in patients with P. aeruginosa or A. baumannii infections (OR 2.068, 95% CI 1.478–2.895), but not with other MDR organisms. Comorbidities and initial ECOG showed higher effect size in patients with S. aureus and E. faecium infections. Conclusion In this large multicenter nationwide study, bloodstream infections caused by MDR P. aeruginosa and A. baumannii were associated with higher risk of decline in performance status at discharge. MDR status did not show association in infections by other species. Disclosures All authors: No reported disclosures.

Information on the most current status of antimicrobial resistance (AMR) in local and national levels has critical importance. However, collection and analysis of a large number of antimicrobial susceptibility test (AST) results often results in additional workload in healthcare facilities and latency in final reporting. We sought to develop an automated nationwide surveillance network in Korea. Data collection servers were set up at each participating institutions, which collects AST results of every bacterial isolate from blood, cerebrospinal fluids, urine, and respiratory specimens. Collected results are anonymized and transmitted to central data server every day without human input. End-user can perform various analyses using data warehouse server through web interface. Only first isolates of same species from individual patients were included in analysis. A total of 19 hospitals located in various regions in Korea participated to the network. From January 2015 through December 2017, AST results of 347,356 isolates were collected. The proportion of MRSA among S. aureus (n = 17,761) was 65.3%, which declined gradually from 71.5 to 62.3% during study period (P < 0.001). The proportion of VRE increased from 29.3 to 36.3% (P = 0.001). Resistance rates of E. coli (n = 63,628) to third and fourth generation cephalosporins, fluoroquinolone, and piperacillin–tazobactam were 31.6, 23.0, 44.0, and 4.2%, respectively. Resistance rates of K. pneumoniae (n = 16,875) to same classes were 32.2, 28.1, 31.0 and 19.1%, respectively. Among E. coli and K. pneumoniae, 0.4 and 4.3% were resistant to carbapenem. Resistance rates of P. aeruginosa (n = 12,895) to carbapenem was 30.5%. However, 72.7% of A. baumannii isolates (n = 9,885) were resistant to carbapenem. Colistin resistance rate was still low at 0.5%. We have established a fully automated nationwide surveillance network for AMR in Korea. Our system provided data on the most current status of AMR, which revealed increase in resistance rates among major Gram-negative pathogens compared with previous studies. Schematic diagram of the Korean Antimicrobial Resistance Surveillance Network (KARSNet). Temporal trends of the resistance rates of S. aureus and E. faecium. Temporal trends of the resistance rates of major Gram-negative species. All authors: No reported disclosures.

Abstract Background Sequence type (ST) 17 of vancomycin-resistant Enterococcus faecium (VREF) is known to be associated with nosocomial isolates. However, there is no evidence of the effect of ST17 VREF on the patient`s clinical outcome. We conducted a retrospective cohort study to identify ST17 VREF would contribute to developing subsequent bacteremia among VREF-colonized patients. Methods VREF-colonized patients and its non-repetitive rectal VREF isolates were collected between March 2014 and February 2015. Subsequent bacteremia event within 1 year after colonization was reviewed from electronic medical records. STs were identified by multi-locus sequence typing. Cohort was defined as VREF with ST17 or non-ST17. Multivariable cox regression model was used to adjust effect of ST17 for developing subsequent bacteremia. If available, pulsed field gel electrophoresis (PFGE) was conducted to compare similarity between rectal and blood VREF isolates. Results Fifty-two patients with ST17 and 169 patients with non-ST17 VREF carriage were included in each cohort. There were six cases and 10 cases of subsequent bacteremia in cohorts ST17 and non-ST17, and 1-year VREF bacteremia free rates were 85.9% and 90.2%, respectively. There was no significant difference of subsequent bacteremia (P = 0.257) in log-rank test. However, after adjusted in multivariable models, VREF ST 17 was associated with subsequent bacteremia (adjusted relative risk, 4.02; 95% CI, 1.32–12.29, P = 0.015). Of 16 patients who had developed to subsequent VREF bacteremia, 12 VREF blood isolates could be analyzed. Only six cases (50%) of rectal and blood isolates had identical ST, whereas all available ST17 VREF cases (four cases) had identical ST and PFGE pattern (Figures 1 and 2). Patients who had identical ST isolates had shorter time difference than those who had non-identical ST isolates (P = 0.041). Conclusion In our study, ST17 VREF was risk factors of subsequent bacteremia and the strain that showed strong concordance between rectal and blood isolates. Further study is needed to improve clinical outcome of patients carrying VREF using genotype data of rectal VREF isolates. Figure 1: Figure 2: Disclosures All authors: No reported disclosures.

Abstract Background Extended spectrum β-lactamase-producing Enterobacteriaceae (ESBL-PE) is related to inappropriate empirical therapy for acute pyelonephritis. The aim of this study was to investigate whether the delay in appropriate antimicrobial therapy of APN caused by ESBL-PE was associated with patient`s poor outcome or not. Methods A retrospective cohort study was performed at a tertiary-care hospital from January 2014 through December 2016. Patients who had APN caused by ESBL-PEs and were treated with appropriate definite antibiotics for at least 7 days were enrolled. The delay in appropriates antimicrobial therapy was defined as patients who had received appropriate antibiotics 48 hour or later after diagnosis of APN. Primary endpoint was treatment failure defined as clinical and/or microbiological failure. Secondary endpoint was length of hospital stay and recurrence of febrile urinary tract infection by ESBL-EP within 1-year. The propensity score matching and multivariable Cox proportional hazard modeling were used to adjust heterogeneity of each group. Results A total of 175 eligible cases were collected. Escherichia coli (144/175, 82.3%) was the most common pathogen, followed by Klebsiella pneumonia (29/175, 16.5%). 59 (34%) patients received delayed appropriate antibiotics (delayed group) and 116 (66%) patients received appropriate initial empirical antibiotics (appropriate group). Treatment failure was observed in 5 (8.4%) patients and 9 (7.8%) patients in each group. After matching, the risk of treatment failure was similar between the both groups (adjusted odd ratio [aOR] 1.05; 95% confidential index [CI] 0.26–4.15). Mean days of hospitalization length was similar (10.8 days in delayed group vs. 11.1 days in appropriate group; P = 0.717), and recurrence rates was also similar between the both groups (adjusted HR [aHR] 0.92; 95% CI 0.43–2.00, Figure 1). Conclusion The delay in appropriate antimicrobial therapy did not affect the clinical outcome of patients if they were properly treated thereafter. This suggests that prescription of a broad-spectrum antibiotics was not needed as initial empirical antibiotics for the treatment of APN with a potential risk of EBSL-PE.Figure 1. Time to recurrence within 1 year after initial APN episodes. Disclosures All authors: No reported disclosures.