Application of computer assisted anaemia management algorithms in haemodialysis patients produces predictable haemoglobin outcomes regardless of the erythropoietic agent or frequency of administration: results of a randomised study

Anaemia management in renal disease is directed by guidelines and standards. UK Renal Association Standards suggest 85% haemodialysis (HD) patients should have a Haemoglobin (Hb) ≥10 g/dl within 3 months of commencing HD. European Best Practice Guidelines advocate a higher value of Hb≥11 g/dl. We have demonstrated the application of a computerised decision support system (with predetermined dose intervention Hb lower threshold and ceiling of 11 and 12 g/dl respectively), to achieve UK Renal Association Standards in a highly consistent fashion in the context of thrice-weekly subcutaneous (sc) Epoetin (EPO). Based within PROTON software, it uses current Hb and trends, serum ferritin, percentage of hypochromic red cells, recent transfusion burden and current dose of EPO, to advise the following month's prescription of EPO and intravenous iron. We have recently conducted a single-centre randomised study of weekly sc EPO-β (NeoRecormon) and Darbepoetin-α (Aranesp/ DA) in an iron replete population, adapting the algorithms from a thrice-weekly sc EPO-β regime. Of the 217 HD patients randomised, 81 remained in each arm at 9 months. We analysed Hb values per protocol of the 162 patients at randomisation, and of the 81 patients in each treatment arm in the first and final month. Distribution of data was analysed using Mann-Whitney analysis and paired t-testing. Frequency distribution curves were also created to display data distributions graphically. The similarity of the five curves is striking. This is borne out by scrutiny of the table, demonstrating the mean, standard deviation, median and 15th/85th centiles of each set of data. No statistical differences were demonstrated in Hb distribution between the x3/week EPO population and the DA or EPO arms at 1 or 9 months (Mann-Whitney), nor between DA and EPO arms at months 1 and 9 (Mann-Whitney), nor within DA and EPO arms when 1 and 9 month data were compared (paired t-test). Despite the distribution of Hb in each arm diverging during a period of dose stabilisation (from months 3 to 7, where significant differences in Hb values were demonstrable), systematic dose alterations locked the subsequent values: DA doses fell in response to a rising Hb, with the opposite occurring with the use of weekly EPO. This compensation of doses, regardless of the agent or frequency of administration, confirmed the ability of the computerised decision support system to produce consistent, predictable Hb outcomes in a variety of circumstances.