How to prescribe optimal haemodialysis

The characteristics of patients starting renal replacement therapy (RRT) have changed significantly over the last few years. The last EDTA-ERA Registry Report [1] showed that 37% of patients starting RRT in 1992 were over 65 years old. The median age of patients starting RRT was 62 years. On the other hand, primary renal disease has also changed considerably over the past years. In Europe, 17% of the patients starting RRT in 1992 had diabetes mellitus. It is important to note that in 1977 only 3.7% of new patients were diabetic. Although this 17% incidence of diabetics is only a little more than half of that in the United States, where the last USRDS report [2] showed that 33% of new patients were diabetic, this still indicates an important change in the characteristics of new patients. At the beginning of 1995 in Spain, there were nearly 10 000 patients living with a functioning transplant and 15000 patients on dialysis [3]. Of these patients on dialysis, only 30% are on a waiting list for transplant. This figure is similar to the rest of Europe and is a consequence of the increase in age of the patients and of the existence of co-morbidity. The fundamental conclusion to be drawn then, is that the only possible RRT for the majority of patients is dialysis, and therefore it is of vital importance to define optimal dialysis. This concept of optimal dialysis seems especially important taking into account that for the majority of patients with end-stage renal disease (ESRD) the only treatment possible is dialysis for the rest of their lives. Besides this, when renal transplant results for the best age group are analysed, the 50% graft survival is approximately 11 years [4]. A possible definition of optimal dialysis is that which offers low mortality and morbidity rates, and therefore a high long-term survival with a good quality of life. Various factors are related to good long-term survival; in different studies, the most important of these is nutrition [5,6]. Nevertheless, the efficiency of dialysis is a factor which influences mortality, possibly through its effect on the nutritional state [6-8]. We consider optimal dialysis not that which presents adequate urea kinetic modelling, but rather we coincide with various authors who affirm that survival is the best index of adequate dialysis [9]. Urea kinetic modelling is probably the most objective way of measuring the efficiency of dialysis, and therefore of prescribing optimal dialysis, but there are a series of alterations in ESRD which cannot be measured by urea kinetic modelling. In Figure 1 we can observe the mean Kt/V in patients undergoing different treatment techniques, ranging from conventional haemodialysis with a cuprophan membrane and acetate dialysis fluid to dialysis with the same membrane and bicarbonate, including high flux haemodialysis with 1.9 m polysulfone membrane and two haemodiafiltration techniques (PFD and AFB). All of these patients have very similar Kt/V, but show a series of differences which we will analyse and which are now shown by urea kinetic modelling. The alterations we will review and which should be taken into account when prescribing optimal dialysis are the following: acidosis, hyperphosphataemia, dry weight, /?2-microglobulin, and biocompatibility.