Interventions and Randomization
The patients receiving different dosing schedules of rhEPO were randomly divided into three groups. The patients in group A received 10,000 IU of subcutaneous rhEPO (1 ml) daily from preoperative day 5 to postoperative day 3 (9 doses); the patients in group B received 1 ml of subcutaneous normal saline daily from preoperative day 5 to day 3 and then 10,000 IU of subcutaneous rhEPO daily until postoperative day 3 (6 doses); and the patients in group C received 1 ml of subcutaneous normal saline daily from preoperative day 5 to the day before surgery and then 10,000 IU of subcutaneous rhEPO daily from the surgery day to postoperative day 3 (4 doses). If the Hb level exceeded 150 g/L, rhEPO was stopped. All patients received a daily dose of 100 mg intravenous ferric carboxymaltose when given subcutaneous rhEPO, which has been widely accepted as an effective partner with rhEPO for autologous blood donation in patients undergoing arthroplasty (9).
A random allocation sequence was computer-generated and concealed in consecutively numbered, opaque, sealed envelopes by a research statistician who was not involved in the data analysis. One experienced surgeon enrolled the patients, and another recorded the basic details. The envelope was opened after the patients were enrolled, and the study medication and placebo were prepared by a dedicated nurse who was not involved in daily patient care or outcome measurement. Patients, surgeons, anaesthesiologists, care providers, and data collectors were blinded to the allocation sequence. The placebo (normal saline) has the same appearance (colourless and clear liquid) as rhEPO.