During the study period, patients undergoing CRRT were included and grouped as before SCT (group A) and after SCT (group B). Group A included patients treated from March 2003 to July 2008 and group B those from August 2008 to April 2016. Before SCT, pediatric CRRT was run by occasional operators, but after the SCT began, ICU nurses joined and began to work as the member of SCT. Double-lumen catheters ranging between 6.5 and 13.5 F in diameter (Gambro Healthcare, Lakewood, CO, USA) were inserted into the central veins depending on the child’s age and weight. Polyarylethersulfone hollow-fiber hemofilters (PAES; the Prismaflex® HF20, Gambro Lundia AB, Lund, Sweden) and polyacrilonytrile hollow-fiber hemofilters (① AN69® membrane before the year 2010; the Prismaflex® M-10/60/100, Gambro Lundia AB, Lund, Sweden; ② AN69® ST membrane since the year 2010; the Prismaflex® ST-60/100, Gambro Lundia AB, Lund, Sweden) were used in all patients, depending on the patient’s weight. HF-20 or M-10 were used in children weighing less than 10 kg; ST-60 or M-60 were used in patients weighing 10–20 kg, and ST-100 or M-100 were used in children weighing more than 20 kg. Commercially prepared bicarbonate-buffered hemofiltration replacement fluid (Hemosol B0; Gambro Healthcare, Seoul, Korea; potassium free), was used as a dialysate and replacement fluid. Potassium chloride (KCl) was added if the patient has a risk of hypokalemia (20 mEq KCl mix in the 5L Hemozol® when serum potassium level ranged from 3.6 to 4.5 mEg/L and 40 mEq KCl mix in the 5L Hemozol® when serum potassium level lowered than 3.6 mEg/L). The blood flow rate was set as 5 mL/kg/min [18 (link)]. The predilution replacement fluid rate or dialysate rate was set at a rate of 2000 mL/1.73 m2/hour [18 (link)]. The mode of CRRT was selected from one of the following, depending on the patient’s status of solute imbalance: continuous veno-venous hemofiltration (CVVH), continuous veno-venous hemodialysis (CVVHD), and continuous veno-venous hemodiafiltration (CVVHDF). These were determined by the pediatric nephrologist and pediatric intensivist through in-depth discussion.
The time to initiate CRRT was decided by the pediatric intensivist, depending on each patient’s clinical condition, such as anuria, oliguria (<0.5 mL/kg/hour), or positive fluid balance, regardless of administration of high doses of diuretics (furosemide more than 1 mg/kg/hour). Anticoagulation was not administered during CRRT initiation; however, our protocol establishes that if the filter was blocked within 12 hours of CRRT initiation, anticoagulation agents such as continuous heparin or nafamostat mesilate infusion via the pre-blood pump port were used. The percentage of fluid overload at CRRT initiation (%FO) was calculated using the following formula [20 (link)]:
At the initiation of CRRT, the following data were obtained for all patients: sex, age, diagnosis, underlying patient conditions, blood flow rate, use of inotropic agents, anticoagulants, and hours to starting CRRT.