An approach to reducing hemolysis in an axial-flow blood pump

In an attempt to decrease hemolysis caused by an axial-flow blood pump, we studied whether specific speed (Ns) at a design point (determined by flow in m3/min, pump head in m, and pump speeds in rpm), should be kept within the existing engineering standard range (1000 < Ns < 2500) or whether pump speed should be reduced to a minimum (Ns < 1000). Four pumps (A: 14,000 rpm, B: 18,000 rpm, C: 22,000 rpm, and D: 26,000 rpm), each with an impeller 11.8 mm in diameter, were designed to accommodate a flow rate of 5 L/min and a pressure head of 100 mmHg. At this design point, the Ns of each pump was calculated as A:758, B:974, C:1191, and D:1407. Pump performance was observed, and the total efficiency of each pump was calculated. The hemolysis index (HI) was calculated after simultaneous testing in duplicate of all four pumps using fresh goat blood (anticoagulated with citrate-dextrose solution) in a closed mock-loop circuit. Total efficiency of each pump was calculated as A:49%, B:50%, C:45%, and D:22%. In the first hemolytic test, HIs were measured as A:0.066, B:0.18, and C:0.13; a water seal failed in pump D. In the second test, HIs were B:0.077, C:0.0499, and D:0.12; a bearing failed in pump A. It is concluded that a lower level of hemolysis is associated with a pump speed in the minimum range at the design point, even though Ns is outside the standard range.