The purpose of this single center, randomized, control, open-labeled study is to evaluate the effect and safety of RCA versus no anticoagulation for CRRT in patients with liver failure and high risk bleeding.

Liver failure (LF) is a common co-morbidity in critical care patients who need continuous renal replacement therapy (CRRT). Patients with LF are usually associated with impaired coagulation, impaired metabolic ability of anticoagulants, and increased bleeding risk. KDIGO guideline recommended no-anticoagulation for CRRT in patients with liver failure and increased bleeding risk. However, the averaged CRRT circuit lifespan under no-coagulation was reported to be 7-8 hours in patients with liver failure. Commonly, a CRRT regimen needs more than 24 hours treatment, which means 3-4 filters replacement for one regimen in liver failure patients underwent no-anticoagulation CRRT. Several observational studies suggested that regional citrate anticoagulation (RCA) during CRRT was effective and safe in patients with liver failure. Therefore, the current opinions on the anticoagulation strategy for CRRT in patients with liver failure and high bleeding risk are controversial. Therefore, the purpose of this single center, randomized, control, open-labeled study is to evaluate the effect and safety of RCA versus no anticoagulation for CRRT in patients with liver failure and high risk bleeding.

Inclusion Criteria:

• Liver failure (acute liver failure and chronic liver failure)
• High bleeding risk
• Scheduled CRRT treatment
• Informed consent

Exclusion Criteria:

• Use of other anticoagulants
• Uncorrectable hypoxemia (PaO2 < 60 mmHg) or systemic hypoperfusion shock
• Pregnancy or lactation
• Fistula, CRRT treatment time < 12 hours