Impact of continuous veno-venous hemofiltration on acid-base balance

Continuous veno-venous hemofiltration (CVVH) appears to have a significant and variable impact on acid-base balance. However, the pathogenesis of these acid-base effects remains poorly understood. The aim of this study was to understand the nature of acid-base changes in critically ill patients with acute renal failure during continuous veno-venous hemofiltration by applying quantitative methods of biophysical analysis (Stewart-Figge methodology).

We studied forty patients with ARF receiving CVVH in the intensive care unit. We retrieved the biochemical data from computerized records and conducted quantitative biophysical analysis. We measured serum Na+, K+, Mg2+, Cl-, HCO3-, phosphate, ionized Ca2+, albumin, lactate and arterial blood gases and calculated the following Stewart-Figge variables: Strong Ion Difference apparent (SIDa), Strong Ion Difference Effective (SIDe) and Strong Ion Gap (SIG).

Before treatment, patients had mild acidemia (pH: 7.31) secondary to metabolic acidosis (bicarbonate: 19.8 mmol/L and base excess: -5.9 mEq/L). This acidosis was due to increased unmeasured anions (SIG: 12.3 mEq/L), hyperphosphatemia (1.86 mmol/L) and hyperlactatemia (2.08 mmol/L). It was attenuated by the alkalinizing effect of hypoalbuminemia (22.5 g/L). After commencing CVVH, the acidemia was corrected within 24 hours (pH 7.31 vs 7.41, p<0.0001). This correction was associated with a decreased strong ion gap (SIG) (12.3 vs. 8.8 mEq/L, p<0.0001), phosphate concentration (1.86 vs. 1.49 mmol/L, p<0.0001) and serum chloride concentration (102 vs. 98.5 mmol/L, p<0.0001). After 3 days of CVVH, however, patients developed alkalemia (pH: 7.46) secondary to metabolic alkalosis (bicarbonate: 29.8 mmol/L, base excess: 6.7 mEq/L). This alkalemia appeared secondary to a further decrease in SIG to 6.7 mEq/L (p<0.0001) and a further decrease in serum phosphate to 0.77 mmol/L (p<0.0001) in the setting of persistent hypoalbuminemia (21.0 g/L; p=0.56).

CVVH corrects metabolic acidosis in acute renal failure patients through its effect on unmeasured anions, phosphate and chloride. Such correction coupled with the effect of hypoalbuminemia, results in the development of a metabolic alkalosis after 72 hours of treatment.