Summary. As part of a study of the thrombogenic properties of dialysis membranes a test‐cell has been devised in which the retention of platelets in the presence of different membranes can be measured. This provides a test for platelet adhesiveness which in certain aspects is more sensitive and reproducible than a glass bead column technique. The normal range for platelet adhesiveness was found to be 31 ± 9% using whole blood and 19 ± 3% using platelet‐rich plasma. The platelet adhesiveness value was largely independent of the packed red cell volume but was, in part, dependent upon the presence of adenosine diphosphate. The membrane test‐cell system was sensitive enough to demonstrate a constant and significant fall in platelet adhesiveness following salicylate ingestion. It was convenient in use as blood anticoagulated with 3.8% sodium citrate allowed delays of up to 1 hr without influencing the result. Heparin was found to have an inhibitory effect on platelet adhesiveness. The use of twin test‐cells allowed platelet adhesion to two different dialysis membranes to be studied simultaneously. It was found that fewer platelets adhered to Cuprophan PT 150 membranes than to PT 300 membranes suggesting that PT 150 membranes may be less thrombogenic in vivo.

The importance of reducing blood loss from any source in patients on regular dialysis treatment is generally recognised. This minimises blood transfusion requirements, and the consequent risks of serum hepatitis. Blood loss occurring from Scribner shunts and arteriovenous fistulae during routine dialysis has been measured. The blood loss from this source has been estimated as 1 to 2 litres per annum, and this may exceed the loss due to residual blood trapped in coil dialysers. The risks to the patients and staff due to this source of blood loss are emphasised.

In a study of potassium metabolism in patients with advanced chronic renal failure, total body potassium was measured in 36 non-dialysed patients using the Merlin mobile whole-body monitor. These values were compared with the expected ‘normal’ values as estimated from known regression relationships of total body potassium on height and age and on height, weight and age. Five of the 36 patients had measured total body potassiums significantly greater than expected. A clinical condition in which excessive total body potassium exists does not appear to have been previously reported.