Objective: Continuous renal replacement therapy is commonly used in the treatment of acute kidney injury. Although the optimal anticoagulation system is not well defined, citrate has emerged as the most promising method. We evaluated the data of 143 patients with acute kidney injury subjected to citrate-based continuous venovenous hemodiafiltration.
Design: Retrospective cohort study.
Setting: Intensive care unit of tertiary care private hospital.
Patients: Patients with acute kidney injury treated from February 2004 to July 2006.
Interventions: None.
Measurements and main results: The main cause of acute kidney injury was sepsis (58%). The mean dialysis dose was 36.6 mL/kg/hr allowing for excellent metabolic control (last tests: creatinine, 1.1 mg/dL; urea, 46 mg/dL). No significant bleeding, severe electrolyte, or calcium disorders were observed. Of the 418 filters used, almost 28,000 hrs of treatment, hemofilter patency was 68% at 72 hrs. Hospital mortality was 59%, and 22% of survivors were dialysis-dependent at the time of discharge. Within our sample, we identified 21 patients with liver failure (mean prothrombin time index, 21% vs. 67%, p < 0.001). This group presented with a lesser median systemic ionized calcium level (1.06 vs. 1.12 mmol/L, p < 0.001) and similar mean total calcium level (8.5 vs. 8.6 mg/dL, not significant), compared with patients without liver failure. These subjects also showed acidemia (median pH, 7.31 vs. 7.40, p < 0.001); however, they exhibited higher levels of lactate (median 29 vs. 16 mg/dL, p < 0.001), chloride (mean 109 vs. 107 mEq/L, p = 0.045) and had a trend to higher mortality rate (76% vs. 56%).
Conclusions: Besides a trend toward higher mortality rate observed in the group with liver failure, we found that citrate-based continuous venovenous hemodiafiltration allowed an effective dialysis dose and reasonable filter patency.