Introduction. Haemostasis is a vital, complicated process. Many standard orthopaedic operations strain the limits of this process, leading to blood transfusions. The former view is that haemostasis occurs cascade-wise in discrete steps, primary haemostasis followed by coagulation, followed by fibrinolysis. This view has been modified to the insight that although there is a temporal succession of these steps, there is also multiple ante- and retrograde interactions between the various reactions.

The complexity of the haemostatic system also implies individual variance of the effectiveness of haemostasis. Minor haemostatic defects such as mild cases of von Willebrand disease probably occur in several per cent of the population. Furthermore many orthopaedic patients are on medication with ASA, NSAIDs, clopidogrel, antidepressants, warfarin and LMWH, all common drugs which affect haemostasis.

Methods to reduce blood loss. Basal measures include avoidance of hypothermia, appropriate positioning of the patient, appropriate anesthesiological and surgical techniques and if possible discontinuation of unsuitable drugs.

In patients with known haemostatic disorders, substitution of the deficient coagulation factors may improve haemostasis. The same holds true for patients on warfarin medication where substitution with vitamin K, with factors of the vitamin-K dependant complex or with plasma normalizes haemostasis.

Desmopressin stimulates the release of factor VIII and the von Willebrand factor and thus improves platelet function in some subgroups of von Willebrand disease and in platelet dysfunction due to ASA or dextran use. However, the blood-saving effect in patients without these disorders has not been conclusively shown.

In recent years the role of factor VII as a main initiator of coagulation has been stressed. Case reports of effective haemostasis in severe trauma using recombinant factor VII have been published but the experience of its use in orthopaedic surgery is so far limited and the cost is prohibitive for routine use.

During surgery and trauma, the fibrinolytic system is activated with particularly high levels of fibrinolytic markers in the wound. The effect of tranexamic acid, a synthetic fibrinolysis inhibitor has been studied in 17 randomised control trials in knee and hip arthroplasty. The drug significantly reduced blood loss and/or blood transfusion in the majority of these studies. The same findings were reported in 2 studies in spinal surgery.

To exert full effect, tranexamic should be given prophylactically, before the beginning of surgery. In studies at our department, the use of tranexamic acid was highly costeffective as it is significantly cheaper than blood transfusions.

Aprotinin, a protease inhibitor decreasing fibrinolysis has been extensively used in cardiac surgery. It has also been shown to reduce blood loss and blood transfusion in 4 out of 5 RCT:s in major orthopaedic surgery. Neither aprotinin nor tranexamic acid were reported to increase the frequency of postoperative venous thromboembolism.

Fibrin sealant, sprayed onto the wound has also been reported to reduce bleeding in spinal surgery as well as in arthroplasties.

Conclusion It is important to reduce blood loss and the burden of transfusion in orthopaedic surgery. This can be achieved by some simple basal methods as well as by the aid of various drugs to ameliorate haemostasis. At present, tranexamic acid seems to be the most costefficient drug for routine use.

We investigated the effect of a fibrinolytic inhibitor, tranexamic acid, on blood loss and blood transfusion in knee arthroplasty by a randomised, double-blind study of 86 patients. A dose of 10 mg/kg body-weight of either tranexamic acid or placebo was given intravenously shortly before the release of the tourniquet, and repeated three hours later. The mean total blood loss was 730 ± 280 ml in the tranexamic acid group as against 1410 ± 480 ml in the placebo group (p < 0.001).

Both the number of patients receiving blood transfusion and the number of blood units transfused were reduced to one-third in the treated group, and mean postoperative Hb concentrations were significantly higher after prophylaxis. The number of thromboembolic complications was the same in both groups. Tranexamic acid should be given prophylactically in order to be effective.