Citrate anticoagulation for continuous renal replacement therapy in small children

A retrospective review was conducted in 30 children (mean body weight 8.69 ± 5.63 kg) who had received at least 24 h of continuous arterio–venous or venous–venous hemodiafiltration in the University Hospital, Poznan, Poland, between January and December 2012. HA was used as the standard anticoagulation method (n = 14) at the University Hospital until the middle of 2012 when it was replaced by RCA which was used thereafter in the treatment of new patients (n = 16). No special selection or randomization was used. Disease categories in these children included nephrotic syndrome (2 patients), multiorgan dysfunction due to cardiac surgery (18 children), sepsis (3 children), leukemia (3 children) and metabolic diseases (2 children). CRRT was started due to clinical features and the presence of life-threatening edemas despite enhanced diuresis. Therefore, in some cases, CRRF was started before the failure stage according to pediatric RIFLE (Risk, Injury, Failure, Loss, End-stage kidney disease) criteria [8]. Children on RCA-CRRT did not differ from those on HA-CRRT in terms of age, body weight and creatinine level. Initial clinical and laboratory data are given in Table 1.

Ethics Committee approval was not required for this observational study because all data reported, as well as anticoagulation method assignment, are part of our routine medical procedures and guidelines.

CRRT was performed using the Prismaflex system (Gambro, Lundia AB, Lund, Sweden) including the Prismaflex HF-20 or —60 hemofiltration system (Gambro). Vascular access varied based on the size of the child. In neonates weighing <5 kg, two central catheters were used with minimum size of 20G for arterial access and 18G for venous access. In older children double-lumen accesses were used.

RCA-CRRT was performed using a commercially available pre-dilution citrate solution (Prismocitrate 18/0; Gambro) and a commercially available dialysate normocarbonate solution (Prism0cal B22; Gambro), according to the Intensive Care Units Therapeutic Recommendation for CRRT by Gambro, which is based on the recommendations of Tolwani et al. [9]. Prismocitrate 18/0 consists of 0.5 % trisodium citrate solution, which contains 18 mmol/l citrate without citric acid and 140 mmol/l sodium. In relation to blood flow rate, the citrate solution rate was set to achieve a circuit citrate concentration of 3 mmol/l blood flow. The dosage of citrate in millimoles per liter blood flow was defined as the pre-dilution citrate load. The pre-dilution citrate load was modified, if needed, to obtain circuit ionized calcium in the range 0.25–0.50 mmol/l (post-filter sample). To assess the dosage of citrate in millimoles per hour delivered to the patient, we calculated the estimated citrate load. The post-dilution bicarbonate solution rate was adjusted to obtain a total dialysis dose of >35 ml/kg/h. Calcium glubionate and calcium gluconate (10 and 5 % solution, respectively, in children weighing <10 kg) were infused in the separate central catheter if a double-lumen catheter was used or in a venous single catheter. Calcium compensation was defined as a percentage of supplemented calcium solution when compared to calcium loss with dialysate, and it was set to the machine according to systemic calcium. Calcium solution flow was calculated automatically based on pre-dilution citrate load, dialysate flow and type of calcium solution used. Systemic ionized calcium was maintained in the normal range from 0.9 to 1.2 mmol/l.

In patients with HA-CRRT the conventional heparin protocol was applied as follows: mean initial bolus of unfractionated heparin was in the range 20–30 IU/kg and subsequent continuous infusion of heparin ranged from 10 to 20 IU/kg/h, adjusted to achieve the target activated partial thromboplastin time (APTT) ratio of approximately 1.5. PrismaSol 2 (Gambro) was used as a dialysate and replacement fluid. Blood flow rate (BFR) was determined as 2–3.6 ml/kg/min and ranged from 16 to 24 ml/min. The recommended minimum BFR for the smallest HF20 filter is 20 ml/min. In the children included in our study, it was very difficult due to small-sized catheters to obtain a higher BFR. Hypotension, which is one of the complications that can arise at the time the patient is being connected to the hemodiafiltration system, especially cardiac patients, was defined as a fall in the mean blood pressure of >20 mmHg relative to baseline during the first 60 min after connection to the CRRT system. To avoid hypotension, the circuits were primed with 5 % albumin solution or with blood. Multiorgan dysfunction syndrome (MODS) was defined as the presence of at least three failed organs. Organ system failure was defined using the international pediatric sepsis consensus conference definitions [10]. The presence of acute liver failure was defined as the abrupt loss of liver function without pre-existing liver disease and by determining laboratory liver function parameters [elevated aspartate aminotransferase and coagulopathy with a prothrombin time of ≥15 s or an international normalized ratio of ≥1.5]. Due to general anesthesia, the encephalopathy criterion was not considered [11]. Electrolyte disturbance during CRRT was as categorized as follows: hypocalcemia (ionized systemic calcium <0.9 mmol/l), hypercalcemia (ionized systemic calcium >1.2 mmol/l), hyponatremia (sodium <129 mmol/l), hypernatremia (sodium >143 mmol/l), metabolic acidosis (pH <7.35, HCO₃ <22 mmol/l) and metabolic alkalosis (pH >7.45 and HCO₃ >26 mmol/l). The risk of bleeding in patients who had undergone cardiac surgery was assessed based on blood transfusion rate (ml/kg/day).