1605. Differences in Clinical Characteristics of Third Generation Cephalosporin Resistance and Treatment Outcomes in Escherichia coli and Klebsiella pneumoniae Bacteremia in Patients with Liver Cirrhosis

BACKGROUND: This study aimed to identify characteristics of third-generation cephalosporin (3GC) resistance in Escherichia coli bateremia (ECB) and Klebsiella pneumoniae bacteremia (KPB) in patients with liver cirrhosis (LC), and to investigate the effects of appropriateness of empirical antibiotic treatment on outcomes. METHODS: We retrospectively collected demographic, clinical and microbiological information on all ECB and KPB episodes in LC patients ≥ 18 years of age hospitalized to a tertiary-care teaching hospital in South Korea from 2007 to 2018. Clinical characteristics associated with 3GC resistance and treatment failure were analyzed using a multivariate logistic regression model. Treatment failure was defined as persistent bacteremia for ≥ 7 days, or relapsed bacteremia ≤ 30 days, or all-cause mortality ≤ 30 days. RESULTS: 3GC resistance rates of E. coli were 30.3% overall and increased significantly during the study period (P=0.001), while the rates of K. pneumoniae were not changed (24.3% overall) (P=0.994). Of total 356 ECB and KPB episodes, 112 were caused by 3GC resistant strains. The factor associated with 3GC resistance was isolation of 3GC resistant strain ≤ 1 year in both ECB (OR, 7.754; 95% CI, 2.094~28.716) and KPB (OR, 2.774; 1.318~5.838). In ECB, beta-lactam or fluoroquinolone treatment ≤ 30 days was another factor associated with 3GC resistance (OR, 2.774; 95% CI, 1.318~5.838), but not in KPB. The factor associated with treatment failure was high MELD score in both ECB (OR, 1.193 at 1 increase; 95% CI, 1.118~1.272) and KPB (OR, 1.163; 95% CI 1.083~1.250). Additionally, in ECB, non-alcoholic LC (OR 3.262; 95% CI 1.058~10.063), high Charlson Comorbidity Index (OR, 1.285; 95% CI 1.066~1.548), and inappropriate empirical antibiotic treatment (OR, 3.194; 95% CI 1.207~8.447) were associated with treatment failure. CONCLUSION: During the study period, 3GC resistance increased in ECB, but not in KPB. In ECB, the severity of the underlying disease and the appropriateness of empirical antibiotics were associated with treatment failure, but there was no correlation in KPB. In ECB of LC patients, the appropriateness of empirical antibiotics was a factor associated with treatment outcome, and is the only correctable factor in the clinical setting. DISCLOSURES: All Authors: No reported disclosures