EFFICACY AND SUSCEPTIBILITY OF EMPIRICAL ANTIBIOTIC THERAPY IN HOSPITALIZED CIRRHOTIC PATIENTS WITH MULTIDRUG-RESISTANT BACTERIAL INFECTION

Background and aims
Multidrug-resistant bacterial infections (MDR-BI) are emerging problems and associated with poor outcomes in cirrhotic patients. This study aimed to identify risk factors and assess efficacy of empirical antibiotics (ATB) in cirrhotic patients with MDR-BI.
Methods
We consecutively enrolled hospitalized cirrhotic patients with bacterial infection (BI) at Chulalongkorn University Hospital, Thailand, between 2018 and 2023. Microbiology and clinical information were collected at baseline and during hospitalization. MDR-BI was defined as acquired resistance to at least one agent in three or more antimicrobial categories. ATB adherence was categorized according to the recommendation (J Hepatol. 2014 Jun;60(6):1310-24). Clinical responsiveness was defined as vital signs stability, symptoms and/or laboratory improvement.
Results
Overall, 359 cirrhotic patients were enrolled. Mean age was 66±16 years, mean Child-Pugh, and MELD score were 9±2 and 19±8, respectively. There were 218 patients with positive bacterial cultures (blood and body fluids), and 85 (38.5%) patients had MDR-BI. Most common MDR bacteria were Escherichia coli (22.9%), Acinetobacter baumannii (5%), and Klebsiella pneumoniae (5%). Multivariate analysis showed healthcare-associated and hospital-acquired infection, invasive procedure within three months prior to admission, urinary tract infection, and ventilator use were independent predictors of MDR-BI. ATB susceptibility was significantly lower in patients with MDR-BI compared to those with non-MDR-BI (47% vs. 78.2%, p<0.001). Patients with MDR-BI had decreased clinical response to empirical ATB (50% vs. 64.2%, p=0.04) and higher incidence of septic shock (20.2% vs. 9.7%, p=0.03) than those with non-MDR-BI. In MDR-BI, carbapenems and piperacillin-tazobactam showed similar antimicrobial susceptibility, while they exhibited higher antimicrobial susceptibility than third-generation cephalosporins. In addition, carbapenem had a similar clinical response rate to piperacillin-tazobactam, but greater rate than third-generation cephalosporins (Figure 1). Susceptibility to empirical ATB resulted in a lower incidence of septic shock (10% vs. 29.5%, p=0.03) and ventilator use (5% vs. 22.7%, p=0.02) compared to antimicrobial non-susceptibility. Empirical ATB was adherent to the recommendation in 61.6% of patients. Rate of ATB adherence was comparable between MDR and non-MDR BI (p=0.22). Patients with BI who received adherent ATB exhibited more significant clinical response (69.1% vs. 34.9%, p<0.001) and resolution of infection (89% vs. 77.8%, p=0.04) than those with non-adherent ATB treatment.
Conclusions
Empirical broad-spectrum ATB enhanced ATB susceptibility and clinical response in hospitalized cirrhotic patients at risk for MDR-BI. Adherence to empirical ATB recommendations improved the outcomes in cirrhotic patients with BI.