Estimating plasma volume from red cell volume

Following the withdrawal of the only licensed supply of 125I-HSA in 1997, most UK centres now simply estimate plasma volume from a knowledge of the red cell volume and venous haematocrit. We compared measured and estimated plasma volume in 107 consecutive patients who had had red cell and plasma volume measured independently in the conventional way. In only 40% of cases was the estimated value accurate to within +/- 5%, and discrepancies of more than +/- 10% were seen in 35%. As there appears to be no simple way of predicting those patients for whom an estimation is inappropriate, it is recommended that plasma volume is measured independently in all patients, using a reputable source of unlicensed 125I-HSA.