Continuous veno-venous hemodiafiltration or hemofiltration: Impact on calcium, phosphate and magnesium concentrations

Background and Objectives: Different techniques of continuous renal replacement therapy (CRRT) might have different effects on calcium, phosphate and magnesium concentrations. Accordingly, we tested whether continuous veno-venous hemodiafiltration (CVVHDF) or continuous venovenous hemofiltration (CVVH) would achieve better control of these electrolytes. Design: Retrospective controlled study Setting: Two tertiary Intensive Care Units Patients: Critically ill patients with acute renal failure (ARF) treated with CVVHDF (n=49) or CVVH (n=50) Interventions: Retrieval of daily morning ionized calcium, phosphate and magnesium before and after the initiation of CRRT for up to 2 weeks of treatment. Measurements and Results: Before treatment, both groups had a high incidence of abnormal ionized calcium concentrations (57.2% for CVVHDF vs 46.0% for CVVH; NS). After treatment, both groups showed a significant increase in serum calcium concentration over the first 48 h (p=0.041 vs p=0.0048) but hypercalcemia was more common during CVVHDF (15.3% vs 0.4%; p<0.0001). However, in both groups, hypocalcemia remained common (30.9% vs 36.7%; NS). Before treatment, abnormal serum phosphate concentrations were also common (65.1% for CVVHDF vs 78.1% for CVVH; NS). After treatment, both groups achieved a significant reduction of serum phosphate within 48 hours (p<0.0001 in both groups). There was no difference in the prevalence of abnormal phosphate levels during treatment (45.5% vs 42.4%; NS). Before treatment, both groups had a high incidence of abnormal magnesium concentrations (50.0% for CVVHDF vs 51.2% for CVVH; NS). During treatment, there was no significant change in serum magnesium concentrations during the first 48 hours or in the prevalence of abnormal magnesium concentrations (56.3% vs 63.4%; p=0.13). However CVVHDF was associated with a higher prevalence of hypomagnesemia (8.1% vs 0.4%; p&LT;0.0001) and a lower incidence of hypermagnesemia (48.2% vs. 63.0%; p=0.0014). Conclusions: In critically ill patients with ARF calcium, phosphate and magnesium were commonly abnormal and they were only partly corrected by CRRT CVVH and CVVHDF had a different effect on serum magnesium concentrations.