Antimicrobial resistance in patients with decompensated liver cirrhosis and bacterial infections in a tertiary center in Northern Germany.

Abstract

Background and aims: Bacterial infections are common in patients with decompensated liver cirrhosis and a leading cause of death. Reliable data on antibiotic resistance are required to initiate effective empiric therapy. We here aim to assess the antimicrobial resistance profile of bacteria among patients with liver cirrhosis and infection.

Methods: Overall, 666 cirrhotic patients admitted to Hannover Medical School between January 2012 and April 2018 with ascites were assessed for bacterial infection. In case of infection, bacteria cultured from microbiological specimens of ascites, blood or urine were identified and analyzed for resistances against common antibiotic agents. Furthermore, analyses compared two periods of time and community-acquired vs. nosocomial infections.

Results: In 281 patients with infection, microbiological sampling was performed and culture-positive results were obtained in 56.9%. Multidrug-resistant (MDR)-bacteria were found in 54 patients (19.2%). Gram-positive organisms were more common (n = 141/261, 54.0%) and detected in 116/192 culture-positive infections (60.4%). Comparing infections before and after 2015, a numerical decline for MDR-bacteria (23.8% vs. 15.6%, p = 0.08) was observed with a significant decline in meropenem resistance (34.9% vs. 19.5%, p = 0.03). MDR-bacteria were more frequent in the case of nosocomial infections. Of note, in ascites the majority of the tested bacteria were resistant against ceftriaxone (73.8%) whereas significantly less were resistant against meropenem (27.0%) and vancomycin (25.9%).

Conclusions: In our tertiary center, distinct ratios of gram-positive infection with overall low ratios of MDR-bacteria were found. Adequate gram-positive coverage in the empiric therapy should be considered. Carbapenem treatment may be omitted even in nosocomial infection. In contrast, 3rd generation cephalosporins cannot be recommended even in community-acquired infection in our cirrhotic population.