[Effects of plasma substitutes on hemostasis].

OBJECTIVE: Data synthesis on haemostasis effects of cristalloids and colloids and clinical implications for their use for plasma volume replacement. DATA SOURCES: Data were searched in the Medline database from 1954 to 2000 using the following key-words: cristalloids, colloids, albumin, gelatin, dextran, hydroxyethyl starch, haemostasis, von Willebrand disease, haemodilution. DATA EXTRACTION: Publications from 1954 to 1990 were selected depending on the quality of their methodology. Most of articles published after 1990 and all types including case report were accepted. DATA SYNTHESIS: Cristalloids induces a moderate hypercoagulable state with 10 to 30% haemodilution. Hypocoagulation is observed above 50% haemodilution. Albumin does not impair hemostasis except with a 50% or more haemodilution where hypocoagulation is observed. Dextran dramatically impairs haemostasis and fibrinolysis. With increasing dose, a progressive decrease of all von Willebrand multimers, mostly the largest, is observed. Till 50% haemodilution, gelatin has a moderate impact on hemostasis, but platelet aggregation is moderately modified. However this moderate impairment of haemostasis may potentiate the haemostatic effect of other colloids when used in association with gelatin. More than 30% haemodilution with hydroxyethyl starch (HES) has a serious effect in vitro on platelet function and fibrinoformation. In most studies in human, less than 20 ml.kg-1 plasma volume replacement has no clinical impact, but in some evaluations postoperative bleeding is more important with HES, particularly HES 450, in comparison to other colloids. With HES 450 and HES 200 highly substituted (0.6 of degree of substitution) intravascular cumulation of large molecules leads to type I von Willebrand syndrome when doses overtake 80 ml.kg-1. Dextran and HES are prohibited in patients with impaired haemostasis due to congenital disease (haemophilia and von Willebrand disease) or acquired defect (thrombocytopenia). Caution is required in patients with renal failure or receiving antithrombotic or non-steroidal anti-inflammatory agents. Patients without a haemorrhagic diathesis must not received more than 1.5 g.kg-1.j-1 of dextran and restrictive conditions of use must be respected with HES.
CONCLUSION: Except isotonic cristalloids, all colloids induce haemostastic changes particularly for haemodilution over 30%. Effects are more pronounced with HES and dextran.