Abstract
OBJECTIVES: To determine the mortality-related risk factors for carbapenem-resistant Enterobacteriaceae (CRE) infection in hospitalized patients and to compare the clinical efficacy of different antimicrobial regimen. METHODS: Data were retrospectively collected from a 3,500-bed regional medical center between January 2021 and June 2022. Mortality-related risk factors were analyzed by the Cox proportional regression model for multivariate analysis. RESULTS: 120 patients were included and the all-cause mortality was 20.8% (25/120). Multivariate analysis showed that age (HR = 1.035, 95%CI: 1.002-1.070, P = 0.036), SOFA score (HR = 1.169,95%CI: 1.066-1.281, P = 0.001), central venous catheter (HR = 3.858, 95%CI: 1.411-10.547, P = 0.009), the length of hospital stay (HR = 0.868, 95% CI: 0.806-0.936, P = 0.000) and combination therapy (HR = 3.152, 95%CI: 1.205-8.245, P = 0.019) were independent mortality risk factors after CRE infection. All patients received definitive therapy and 65.0% (78/120) received sensitive drug treatment. Among those 65.4% (51/78) received combination therapy and 34.6% (27/78) received monotherapy. Subgroup analysis of the non-sepsis group showed significantly lower mortality in monotherapy than in combination therapy (0% versus 22.2%, P = 0.034). Patients who received carbapenem-containing therapy had significantly higher mortality than those who received carbapenem-sparing therapy (31.3% versus 13.9%, P = 0.022). CAZ-AVI-containing therapy presented a lower mortality (19.0%) and a higher 7-day microbiological clearance (47.6%) compared to other antimicrobial regimens, but there were no statistical significance (P>0.05). CONCLUSIONS: Patients with older age, higher SOFA score, central venous catheter, shorter hospital stay after CRE infection may had poor outcomes. Since patients with non-sepsis have a lower mortality rate from monotherapy, combination antibiotic treatment should not be routinely recommended. Patients who received CAZ-AVI-containing therapy presented a lower mortality compared to other antimicrobial regimens without statistical significance, further larger sample size is needed for verification.