Controlling factors in hemofiltration

Abstract

Ultrafiltration of bovine plasma and blood (hemofiltration) was performed in a dual-parallelmembrane test device over a wide range of hematocrits (0–85%), plasma protein concentrations (3–20 gm%), channel heights (0.022–0.073 cm) and flow rates. The plasma ultrafiltration data was used to determine film theory equation constants by regression techniques similar to that which had been done by others in the past. The ratios of hemofiltration to plasma ultrafiltration rates were then correlated by an independent analytical function of hematocrit, total protein concentration, and the superficial wall shear rate. The filtration flux ratio has a Fricke equation limit at low shear rates which would be below unity while a maximum filtration flux ratio greater than one was observed (and correlated) at high shear rates with hematocrits below 42%. The form of the hemofiltration correlation was also applied to previously published augmentation data in a pure shear field without plasma proteins. The design ramifications of the correlation include the determination of when addition augmentation methods are needed. The combination of both plasma ultrafiltration and hemofiltration correlations show how each variable affects and controls hemofiltration. The correlations are for small percentages of the incoming flow being filtered. For large flow fractions being filtered the correlations would have to be integrated along the axial length combined with mass balances. The hemofiltration equations can yield slightly lower fluxes at the lowest channel heights and slightly higher fluxes for the larger channel heights, presumably due to the Fahraeus—Lindqvist effect. The channel heights and shear rates utilized included conditions where this effect should be considerable as well as conditions where it should not be present. The regression techniques were applied to all the data.