Continuous venovenous hemodiafiltration with a low citrate dose regional anticoagulation protocol and a phosphate-containing solution: effects on acid–base status and phosphate supplementation needs

Abstract

Recent guidelines suggest the adoption of regional citrate anticoagulation (RCA) as first choice CRRT anticoagulation modality in patients without contraindications for citrate. Regardless of the anticoagulation protocol, hypophosphatemia represents a potential drawback of CRRT which could be prevented by the adoption of phosphate-containing CRRT solutions. The aim was to evaluate the effects on acid–base status and phosphate supplementation needs of a new RCA protocol for Continuous Venovenous Hemodiafiltration (CVVHDF) combining the use of citrate with a phosphate-containing CRRT solution. To refine our routine RCA-CVVH protocol (12 mmol/l citrate, HCO3- 32 mmol/l replacement fluid) (protocol A) and to prevent CRRT-related hypophosphatemia, we introduced a new RCA-CVVHDF protocol (protocol B) combining an 18 mmol/l citrate solution with a phosphate-containing dialysate/replacement fluid (HCO3- 30 mmol/l, Phosphate 1.2). A low citrate dose (2.5–3 mmol/l) and a higher than usual target circuit-Ca2+ (≤0.5 mmol/l) have been adopted. Two historical groups of heart surgery patients (n = 40) underwent RCA-CRRT with protocol A (n = 20, 102 circuits, total running time 5283 hours) or protocol B (n = 20, 138 circuits, total running time 7308 hours). Despite higher circuit-Ca2+ in protocol B (0.37 vs 0.42 mmol/l, p < 0.001), circuit life was comparable (51.8 ± 36.5 vs 53 ± 32.6 hours). Protocol A required additional bicarbonate supplementation (6 ± 6.4 mmol/h) in 90% of patients while protocol B ensured appropriate acid–base balance without additional interventions: pH 7.43 (7.40–7.46), Bicarbonate 25.3 (23.8–26.6) mmol/l, BE 0.9 (-0.8 to +2.4); median (IQR). No episodes of clinically relevant metabolic alkalosis, requiring modifications of RCA-CRRT settings, were observed. Phosphate supplementation was needed in all group A patients (3.4 ± 2.4 g/day) and in only 30% of group B patients (0.5 ± 1.5 g/day). Hypophosphatemia developed in 75% and 30% of group A and group B patients, respectively. Serum phosphate was significantly higher in protocol B patients (P < 0.001) and, differently to protocol A, appeared to be steadily maintained in near normal range (0.97–1.45 mmol/l, IQR). The proposed RCA-CVVHDF protocol ensured appropriate acid–base balance without additional interventions, providing prolonged filter life despite adoption of a higher target circuit-Ca2+. The introduction of a phosphate-containing solution, in the setting of RCA, significantly reduced CRRT-related phosphate depletion.