Variability of relative blood volume during haemodialysis

Background. A decrease in blood volume is thought to play a role in dialysi s-related hypotension. Changes in relative blood volume (RBV) can be assess ed by means of continuous haematocrit measurement. We studied the variabili ty of RBV changes, and the relation between RBV and ultrafiltration volume (UV), blood pressure, heart rate, and inferior caval vein (ICV) diameter. Methods. In 10 patients on chronic haemodialysis, RBV measurement was perfo rmed during a total of one hundred 4-h haemodialysis sessions. Blood pressu re and heart rate were measured at 5-min intervals. ICV diameter was assess ed at the start and at the end of dialysis using ultrasonography. Results. The changes in RBV showed considerable inter-individual variabilit y. The average change in RBV ranged from -0.5 to -8.2% at 60 min and from - 3.7 to -14.5% at 240 min (coefficient of variation (CV) 0.66 and 0.35 respe ctively). Intra-individual variability was also high (CV at 60 min 0.93; CV at 240 min 0.33). Inter-individual as well as intra-individual variability showed only minor improvement when RBV was corrected for UV. We found a si gnificant correlation between RBV and UV at 60 (r = -0.69; P < 0.001) and a t 240 min (r = -0.63; P < 0.001). There was a significant correlation betwe en RBV and heart rate (r = -0.39; P < 0.001), but not between RBV or UV and blood pressure. The level of RBV reduction at which hypotension occurred w as also highly variable. ICV diameter decreased from 10.3 +/- 1.7 mm/m(2) t o 7.3 +/- 1.5 mm/m(2). There was only a slight, although significant, corre lation between ICV diameter and RBV (r = -0.23; P < 0.05). The change in IC V-diameter showed a wide variation. Conclusions. RBV changes during haemodialysis showed a considerable intra- and inter-individual variability that could not be explained by differences in UV. No correlation was observed between UV or changes in RBV and either blood pressure or the incidence of hypotension. Heart rate, however, was s ignificantly correlated with RBV. Moreover, IVC diameter was only poorly co rrelated with RBV, suggesting a redistribution of blood towards the central venous compartment. These data indicate that RBV monitoring is of limited use in the prevention of dialysis-related hypotension, and that the critica l level of reduction in RBV at which hypotension occurs depends on cardiova scular defence mechanisms such as sympathetic drive.