Clinical and experimental studies of intraoperative autotransfusion using a new filtration device

Abstract

Abstract The Haemocell S-350 device has recently been introduced for intraoperative autotransfusion. The system uses a novel membrane filter to process shed blood. In the first part of this study a 0.2-μm pore size filter was used in a randomized trial comparing the use of autotransfusion (n = 8) with bank blood controls (n = 9) during aortic reconstruction. This part of the trial was abandoned because of unexpected non-surgical bleeding. Bank blood requirements fell from a median of 3.0 (range 0.0-9.0) units to 1.5 (range 0.0–7.0) units when autotransfusion was used, but these patients had a greater perioperative blood loss (1791 (range 932–3104) versus 1140 (range 440–3840) ml). There was evidence of postoperative heparin excess with an activated partial thromboplastin time ratio of 1.3 (range 0.9–3.0) versus 1.0 (range 1.0–1.2) in controls and an activated clotting time of 206 (range 143–280) versus 137 (range 107–142) s. This was confirmed by raised plasma heparin levels and a prolonged thrombin time normalized by protamine. To improve performance a 0.6-μm pore size filter was studied in ten patients. Filtration efficiency doubled from 19 to 38 per cent. Electron micrographs demonstrated better filter clearance, but 44 per cent of the original concentration of heparin remained in the reinfusate. The S-350 device may be an attractive alternative to centrifugation for intraoperative autotransfusion but, until efficiency is improved, it should only be used for cardiovascular surgery when excess heparin can be reversed with protamine.