No abstract

In 1996, we prospectively audited peri-operative transfusion practice in elective surgical patients over a 3-month period. Two-unit transfusions represented 60% of all transfusions. Haemoglobin was measured infrequently prior to transfusion and the main 'trigger' for transfusion was an estimated blood loss in excess of 500 ml. Transfusion guidelines that required the haemoglobin level to be measured immediately before transfusion were introduced. The audit was repeated in 1998; transfusion 'triggers' and the number of transfusions for the two periods were compared. In the second audit, the total number of transfusions decreased by 43%. The mean estimated blood loss associated with a 2-unit transfusion had increased from 608 (373) ml to 1320 (644) ml (p < 0.01) and the estimated haemoglobin concentration after transfusion had decreased from 12.4 (1.8) g.dl-1 to 9.9 (2.4) g.dl-1 (p < 0.01). These results suggest that transfusion guidelines can have a significant impact on clinical practice.

Many renal transplant centres now try to avoid blood transfusion prior to renal transplantation, to avoid alloimmunization due to antibody production against donor antigens usually present on contaminating white cells. Post- or peri-operative transfusions are usually not considered to present problems, since the patient is heavily immunosuppressed. We present a patient who suffered a rare transfusion reaction, that we believe may have initiated a severe vascular rejection of a kidney transplant, probably mediated by Kidd blood group antigens.

Using anti-A and anti-B blood group monoclonal antibodies and fluorescent activated cell sorting of human bone marrow, A (or B) blood group antigen was shown to be on 5.2 +/- 5.9 (mean +/- SD) % of CFU-GEMM and 12.5 +/- 19.6% of the erythroid burst forming cells (designated BFU-GEMM) as defined by the mixed colony assay, and 49.5 +/- 20% of the BFU-E and 83.5 +/- 9.9% of the CFU-E as defined by the erythroid colony assay. This antigen expression on the BFU-GEMM is consistent with the concept that erythroid bursts stimulated by leucocyte conditioned medium are less mature, and are closer in development to the pluripotent stem cell than the BFU-E. These results help to explain the delayed erythropoiesis, and perhaps impaired engraftment of all cell lineages, that may occur in some recipients of ABO incompatible bone marrow transplants with persistent and high anti-A titres.

A retrospective study was carried out to ascertain the blood bank provision required to support a liver transplant programme and to assess the effect of intraoperative aprotinin on blood product requirements in liver transplant recipients with cirrhosis. Sixty patients with end-stage liver disease underwent 62 consecutive orthotopic liver transplants between October 1988 and January 1991. The total and intraoperative requirements of red cells, platelets and fresh frozen plasma (FFP) were analysed for three groups of liver transplant recipients, those without cirrhosis (n = 15), those with cirrhosis (n = 25) and those with cirrhosis who received intraoperative aprotinin (n = 20). Fifteen without cirrhosis had mean total requirements of 15 units of red cells, 18 units of platelets and 16 units of FFP. Twenty patients with cirrhosis who received intraoperative aprotinin had broadly similar requirements. However, blood product requirements for 25 patients with cirrhosis were significantly greater (46 units of red cells, 41 units of platelets, 43 units of FFP, excluding the seven patients with primary biliary cirrhosis). We conclude that a liver transplant programme can be supported by a teaching hospital blood bank. The use of intraoperative aprotinin significantly reduces blood product requirements.