Acquired bleeding disorders are common. They complicate well defined clinical disorders which can be detected by history and examination. Inherited bleeding disorders are uncommon, but can be detected by careful clinical assessment, including family history. Clinical assessment has high sensitivity, although low specificity, for the presence of a bleeding disorder. In contrast, both the sensitivity and specificity of routine laboratory screening are low. Both false negative and false positive results are common with the basic laboratory 'screening tests'. In a patient without a suggestive history, these tests are inappropriate. In a patient with a suggestive history, they may well be inadequate.
Normal haemostasis involves two processes:
- platelet adhesion to areas of vascular injury with the subsequent formation of a platelet thrombus
- surface or tissue contact mediated activation of coagulation with sequential enzymic reactions culminating in the formation of a fibrin thrombus
These two processes are interactive, although their dominance varies in different sites of the vascular system. Activation of haemostasis is accompanied by activation of the fibrinolytic system which should eventually achieve partial or complete removal of the thrombus.
Clinically important abnormalities of these mechanisms are common and a logical and cost effective approach to the detection of patients at risk of bleeding is required.
Acquired bleeding disorders
These disorders occur in well defined clinical settings and the history and physical examination are a sensitive screen for these problems. The commonest cause of an acquired
bleeding disorder is drug therapy:
- aspirin or other nonsteroidal anti inflammatory drug (NSAID) causing platelet dysfunction. The occurrence and severity of bleeding is variable.
- cytotoxic drugs causing thrombocytopenia
- oral anticoagulant therapy inhibiting the synthesis of the vitamin K dependent coagulation factors
Various disorders may also be associated with a bleeding tendency.
- Chronic liver disease can result in a variety of haemostatic abnormalities ranging from deficiency of vitamin K dependent coagulation factors, through more global coagulation factor deficiencies to low grade disseminated intravascular coagulation (DIC).
- Patients with renal failure may have significant platelet dysfunction.
- Those with myeloproliferative disorders may also have significant abnormalities of platelet function.
- Patients with lymphoproliferative or autoimmune disease may develop auto antibodies to coagulation factors or platelets.
- Paraproteins in patients with plasma cell dyscrasias may interfere with platelet function or fibrin formation.
Inherited bleeding disorders
These disorders are less common and the vast majority of patients will have a personal and/or family history of excessive bleeding. The history is a sensitive, though not specific, screen for these patients. It is important to enquire into bleeding during and after surgical or dental procedures; a return to the operating theatre, blood transfusion, or a need for packing or suture of dental sockets would increase the likelihood of there being an underlying problem.
Menorrhagia, recurrent epistaxis and easy bruising are sensitive indicators, although associated with a high rate of false positives. The perception of menstrual blood loss is variable and often influenced by family norms; the frequency of pad/tampon change may assist in determining its significance. Menorrhagia commencing after pregnancies and childbirth and uninfluenced by hormone supplements is unlikely to be due to an inherited haemostatic disorder. Peripartum bleeding is rare in patients with von Willebrand's disease (arguably the most common inherited haemostatic disorder) and the absence of such a history should not influence the decision as to whether investigation is required.
'Spontaneous' haemarthroses and major muscle bleeds are characteristic of haemophilia A (factor VIII deficiency) or B (factor IX deficiency).
Gastrointestinal bleeding and/or epistaxis as isolated problems are unlikely to be due to an inherited bleeding disorder. However, mucosal surfaces should always be inspected for the characteristic lesions of hereditary haemorrhagic telangiectasia. This condition is under diagnosed.
If a positive family history is obtained, the pattern of affected individuals may suggest autosomal (e.g. von Willebrand's disease) or Xlinked (e.g. haemophilia) inheritance.
Bleeding time (BT)
The skin bleeding time reflects 'primary haemostasis', the interaction of platelets with arterioles and capillaries to form a platelet plug (Fig. 1). The test is rarely useful and should not be performed if the platelet count is <100 x 109/L and if the patient has taken aspirin in the preceding 7-10 days, or a NSAID in the past 1-4 days (depending on the half life of the specific drug). Although the effect of these drugs on the bleeding time is variable, their use renders a prolonged bleeding time uninterpretable. If the result will be uninterpretable, the test should not be done. As patients may not be aware of the aspirin content of prescribed or over the counter medications, both the requesting doctor and the laboratory should check the drug history before testing.
The bleeding time may be abnormal in acquired and inherited disorders of platelet function, von Willebrand's disease, the rare afibrinogenaemias and inherited disorders of collagen. It may also be prolonged in patients with 'senile purpura', but the test is neither appropriate nor useful in this condition.
The value of the bleeding time as a 'screening test' of haemostasis is severely limited by its lack of specificity and sensitivity and its use cannot be recommended. The lack of specificity is the result of its susceptibility to physiological and technical variables. This so called 'standardised' technique varies between laboratories and between operators within a given laboratory. It is also an insensitive test and may be normal in patients with mild to moderate von Willebrand's disease and in those with disorders of platelet function. Although the BT can be prolonged by aspirin and other NSAIDs, is often abnormal in uraemia and may be abnormal in patients with myeloproliferative disorders, the presence and degree of an abnormality do not correlate with the risk of bleeding. The test cannot be recommended as a predictor of surgical bleeding.1
Activated partial thromboplastin time (APTT)
The APTT reflects the activity of coagulation factors in the intrinsic system and the final common pathway of coagulation (Fig. 1). It is performed by recalcifying citrated plasma in the presence of a 'surface' activator and a 'partial thromboplastin', simulating platelet membrane phospholipid.
A deficiency of a specific coagulation factor (e.g. factor VIII in haemophilia A), a coagulation factor inhibitor, a lupus inhibitor, or inhibition of coagulation by heparin may prolong the APTT. Although it can be prolonged in patients receiving warfarin or those with vitamin K deficiency or severe liver disease, it is less sensitive to these defects than the prothrombin time.
Incorrect specimen collection or handling may result in either a false positive (prolongation) or a false negative (shortening) APTT result.
Causes of an incorrect APTT
Basic tests of haemostatsis.
In practice, the main uses of the APTT are:
- monitoring full dose, continuous infusion heparin therapy. For the laboratory's therapeutic interval to be meaningful, the heparin sensitivity of the APTT reagent used must have been established by the laboratory. The APTT has no place in the management of patients receiving low molecular weight heparin or heparinoids.
- the detection of significant coagulation factor deficiencies in patients with a history suggestive of an inherited bleeding disorder.
- In the absence of such a clinical history, the yield from the APTT is very low.2,3 A normal APTT does not exclude a mild, but clinically significant, coagulation factor deficiency, as most APTT reagents only detect single coagulation factor deficiencies when the level is <35% of normal. Thus the APTT has little or no value when used as a 'routine' preoperative screening test.
- the detection of a lupus inhibitor in patients with a history of recurrent fetal loss or recurrent venous and/or arterial thrombosis. Although there are more sensitive screening tests (e.g. the kaolin clotting time, KCT) for the lupus inhibitor, in practice, many patients are detected because of a prolongation of the APTT. In spite of sometimes being called the lupus anticoagulant, this antiphospholipid antibody is associated with a tendency to thrombosis, not with a bleeding tendency.
- the detection of coagulation factor inhibitors (antibodies). Although antibodies to factor VIII or, rarely, factor IX, may develop in haemophilia, they may also occur in patients with autoimmune or lymphoproliferative disorders, in the postpartum period and in previously normal elderly patients. In the nonhaemophilia groups, unexplained bleeding of recent onset with an isolated prolongation of the APTT should arouse suspicion that an inhibitor, usually against factor VIII, may be present.
Prothrombin time (PT and INR)
The PT reflects the activity of the 'extrinsic system' and 'final common pathway' of coagulation (Fig. 1). It is measured by recalcifying citrated plasma after the addition of a 'complete' thromboplastin (e.g. a suspension of human or animal brain) which simulates tissue factor. Increasingly, thromboplastins produced by recombinant DNA technology are being used for this test. Compared with the APTT, the PT is more sensitive to the coagulation defect induced by oral anticoagulant therapy and less sensitive to the effect of heparin.
In practice, the main uses of the PT are:
- monitoring oral anticoagulant therapy. For this purpose, the PT result is expressed as an international normalised ratio or INR, which provides a result standardised for local reagents and methodology. The therapeutic interval varies for specific indications and there is currently no consensus in the literature. The North American literature4 recommends an INR of 2.0-3.5, and specifically an INR of 2.5-3.5 for patients with mechanical prosthetic valves. In Europe5, an overall therapeutic interval of 2.5-4.8 is recommended; 3.6-4.8 for patients with mechanical valves. One concern is that some of the differences relate to reagent sensitivity. It has been suggested6 that an INR of 2.0-3.0 is adequate for the treatment and secondary prevention of venous thromboembolism, and an INR of 2.5-3.5 is appropriate for those with mechanical prosthetic valves, perhaps partly because of the reduced thrombogenicity of modern valves.
- assessing patients with hepatocellular disease. The PT is considered to be a sensitive test of liver function; a prolonged PT, particularly in alcoholic liver disease, is often partly due to concomitant dietary vitamin K deficiency.
- detection of vitamin K deficiency; particularly in the alcoholic, in small bowel malabsorption and after prolonged fasting or vomiting, especially if associated with broad spectrum antibiotic therapy.
A prolonged PT may be seen in patients with a lupus inhibitor, although the APTT is generally more sensitive. The PT has only a limited role in the assessment of patients with a history suggestive of an inherited bleeding disorder, as factor VII deficiency is rare. As with the APTT, use of the PT as a 'routine' preoperative screening test has little or no value.2,3
Thrombin time (TT)
The thrombin time assesses the conversion of fibrinogen to fibrin and is measured by adding thrombin to citrated plasma. The TT only detects abnormalities of fibrinogen and of fibrin formation (Fig. 1).
In practice, the thrombin time is used mainly by the laboratory, rather than as a requested test, to:
- detect heparin in a specimen with an unexplained prolongation of the APTT. A prolonged thrombin time which corrects with protamine sulphate confirms the presence of heparin in the sample.
- assist in the diagnosis of disseminated intravascular coagulation (DIC) in which the thrombin time is prolonged due to the presence of hypofibrinogenaemia and fibrin degradation products (FDP) which interfere with fibrin polymerisation. In DIC, the TT only partially corrects with protamine sulphate.
- detect the rare inherited disorders, hypofibrinogenaemia and a fibrinogenaemia.
- detect dysfibrinogenaemia (an abnormal fibrinogen molecule with abnormal function). Inherited dysfibrinogenaemia is rare, but acquired dysfibrinogenaemia may be seen in patients with hepatocellular carcinoma. A similar functional abnormality may be seen in patients with myeloma, due to the paraprotein interfering with fibrin polymerisation.
Investigations should address a diagnostic question, rather than being applied as a 'routine'. Provision of an accurate history on the request form and consultation with the pathologist as to appropriate testing should assist the laboratory to answer that question.
(See also Dental implications)
The following statements are either true or false.
1. The risk of postoperative bleeding can be calculated from the bleeding time and the activated partial thromboplastin time.
2. Most acquired bleeding disorders are drug induced.
Answers to self-test questions
- Lind SE. The bleeding time does not predict surgical bleeding [see comments]. Blood 1991;77:2547-52. Comment in: Blood 1992;79:2495-7.
- Eisenberg JM, Clarke JR, Sussman SA. Prothrombin and partial thromboplastin times as preoperative screening tests. Arch Surg 1982;117:48-51.
- Suchman AL, Griner PF. Diagnostic uses of the activated partial thromboplastin time and prothrombin time. Ann Intern Med 1986;104:810-6.
- Hirsh J, Dalen JE, Deykin D, Poller L. Oral anticoagulants. Mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest 1992;102(4 Suppl):312S-326S.
- Loeliger EA. Therapeutic target values in oral anticoagulation justification of Dutch policy and a warning against the so called moderate intensity regimens [published erratum appears in Ann Hematol 1992;64:253]. Ann Hematol 1992;64:60-5.
- Saour J, Gallus A. Warfarin: is it time to reduce target ranges again? Aust NZ J Med 1993;23:692-6.