The purpose of this study is to compare citrate regional anticoagulation with systemic heparinization in continuous venovenous hemofiltration. The investigators' hypothesis is, that regional citrate anticoagulation with replacement solution containing trisodium citrate, will be associated with lower mortality and less bleeding complications compared to heparin, with also a better filter survival.

Acute renal failure occurs in about 20% of critically ill patients and is associated with increased morbidity and mortality, in spite of modern renal replacement techniques. The latter include continuous venovenous hemofiltration (CVVH) techniques, necessitating anticoagulation of blood entering the extracorporeal circuit to prevent premature clot formation and hemofilter dysfunction. Heparin is commonly used for that purpose, but carries a serious risk of bleeding complications and heparin induced thrombocytopenia. In a subgroup of critically ill patients systemic anticoagulation is absolutely contraindicated. Citrate-anticoagulant CVVH carries the potential advantage of less bleeding complications and prolonged filter survival, but carries the risk of hypocalcaemia, when citrate is inappropriately or insufficiently counteracted by calcium infusion after passage of blood through the filter. In addition, when too much citrate enters the circulation, a metabolic alkalosis may develop, since citrate is converted to bicarbonate by the liver.

Moreover, continuous filtration techniques may attenuate a potentially harmful exaggerated immune response, particularly when high volume filtration (> 6 L/h) is used. Also, the type of anticoagulation may modulate immune responses, as known from biocompatibility studies in intermittent hemodialysis.

In the first part of the research proposal concerning high bleeding risk patients a comparison will be made in a prospective sequential cohort study between no anticoagulation and citrate regarding filter survival time, bleeding risk, dialyser efficacy, circulating immune mediators (such as neutrophil elastase and myeloperoxidase, interleukins, platelet-activating factors, activated complement products, soluble cytokine receptors and adhesion molecules), metabolic balance, and acute renal failure duration. Also, filter survival time will be assessed. The purpose of the second part of the current research proposal is to evaluate in a randomised controlled clinical trial in 350 critically ill patients (18-80 years) with acute renal failure, (2 arms of 175 patients), without an increased bleeding risk (thrombocytes > 40 x 10^9/L, APTT < 60 sec, PT-INR < 2) whether citrate CVVH is better than bicarbonate-heparin CVVH in terms of the same parameters as in the first part of the study but with the addition of mortality as the primary endpoint.

For this purpose a simple predilution system and citrate adjustment protocol will be used and compared to standard heparin dosing. This replacement solution shall be custom made, containing trisodium citrate, no lactate or bicarbonate, no calcium and a low sodium content.

Main objective: Investigation of the mortality during continuous venovenous hemofiltration with systemic anticoagulation with heparin compared with regional anticoagulation with trisodium citrate and also the investigation of the filter survival. Our hypothesis is, that regional citrate anticoagulation with replacement solution containing trisodium citrate, will be associated with less bleeding complications compared to heparin, with also a better filter survival. Most important we want to evaluate the hypothesis that treatment with citrate will result in a lower mortality compared to treatment with systemic heparinization.

Regional anticoagulation with trisodium citrate may also have some potential effects on the immune response as known from biocompatibility studies in intermittent hemodialysis. Bioincompatibility leads to polymorphonuclear cell degranulation as indicated by the release of intracellular granule products such as myeloperoxidase, lactoferrin, lysozyme and elastase. Citrate anticoagulation may lead to a lower polymorphonuclear cell degranulation, since cations play a pivotal role in the process of cell activation and citrate creates an almost calcium-free environment within the dialyser by its virtue to chelate calcium.

Primary endpoints:

Mortality at day 28 after inclusion will be evaluated. Survival time of the first hemofilter used will be determined, including the cause of filter termination and the number of filters used in the first 72 hours; the average filter patency time will be calculated.

Citrate CVVH is stopped and thus also the study, if the patient fulfils one of the following criteria:

1. Total to ionised calcium ratio of more than 2.5.
2. Persistent metabolic alkalosis with a B.E. of more than 10.
3. Clinical signs of hypocalcaemia: tetanic symptoms or prolonged QT interval
4. Progressive non-lactic acidosis (pH < 7.20) during CVVH combined with an increase in anion gap (> 13) without the presence of endo- or exogenous acids other than citrate suggesting citrate accumulation

Patients on heparin developing a HIT will continue CVVH with danaparoid anticoagulation. Patients on heparin developing a bleeding episode will continue CVVH with regional citrate anticoagulation.

Inclusion Criteria:

• Patients admitted on the Intensive Care Unit (ICU) requiring continuous venovenous hemofiltration.
• No high bleeding risk. A high bleeding risk is defined as a platelet count below 40 x 10^9/L or APTT of more than 60 seconds or a PT-INR of more than 2.0 or a recent major bleeding or significant active bleeding i.e. requirement for more than two units of packed red blood cells as a transfusion within 24 hours of initiation of CVVH.

Exclusion Criteria:

• Less than 18 or over 80 years of age.
• Patients administered heparin or coumarins for other reasons will also be excluded.
• Patients with a HIT in known history will also be excluded.