House A A, Pham B, Pagé D E
Ottawa General Hospital, Ontario, Canada.
Nephrol Dial Transplant. 1998 Jul;13(7):1763-9. doi: 10.1093/ndt/13.7.1763.
Before the routine use of recombinant human erythropoietin (rHuEpo), patients dialysed by peritoneal dialysis (PD) received fewer blood transfusions than patients on haemodialysis (HD). We compared transfusion practices in these groups now that the use of rHuEpo has become standard, while controlling for variables known to influence anaemia of end-stage renal disease (ESRD). Maintenance rHuEpo doses were also compared.
Data were examined for 157 HD and 126 PD patients during a 2-year period. Potential confounders included age, gender, albumin, iron deficiency, parathyroid hormone (PTH), underlying renal disease, comorbid illness, renal transplant, dialysis adequacy and duration. An intent-to-treat analysis was used, with sensitivity analyses to account for change in treatment and transplant.
Mean haemoglobin (Hb) was not different (10.47 g/dl for HD, 10.71 g/dl for PD; P = 0.45). Mean monthly transfusion rate was higher for HD (0.47 units per month vs 0.19; P < 0.01). More HD patients received at least one transfusion (52.9 vs 40.9%; P < 0.01). The maintenance rHuEpo dose was higher for HD (7370 U/week vs 5790 U/week; P = 0.01). The only factors associated with risk of being transfused were dialysis duration and mode of dialysis (less risk for PD, odds-ratio 0.57; 95% confidence interval 0.35-0.92).
Despite the routine use of rHuEpo, HD patients received more blood and rHuEpo than PD patients to achieve the same Hb. No patient factors were identified to account for this difference. The use of fewer transfusions and less rHuEpo in PD represents an advantage over HD in terms of both cost and safety.
