Anticoagulation

From Ganfyd

More complicated than most people think.
Most people think it is complicated

This shows worst case curves at upper limit of 95% CI based on INR at time of presentation with stroke in patients anticoagulated for atrial fibrillation^[1]

A fair bit is known as to the benefits and risks of anticoagulation. This is not a simple area although the benefits in several areas such as preventing ischaemic stroke in non-rheumatic atrial fibrillation in people over 65 years old or preventing pulmonary embolism are very well defined.

The curve for benefit in reducing ischaemic stroke with anticoagulation is J-shaped, with the risk of ischaemic stroke increasing if anticoagulation is controlled outside the range of 2 to 2.5.

Doctors may find the real life clinical observation that patients with an INR >4.5 are at a greater a risk of ischaemic stroke as those on no warfarin in non-rheumatic atrial fibrillation so counterintuitive that they fail to evaluate it for the true message of just how important good anticoagulation control is for outcome, or incorrectly tell the patient only that it is their risk of bleeding that increases with increasing INR.

Another example is that the risk of intracranial haemorrhage is an exponential function of INR while the relative risk of a subdural haematoma remains flat for INRs less than 2.4 and then suddenly quadruples. There is more information on the issue of drugs, not just anticoagulants, increasing bleeding risk, here.

Risk with anticoagulation is commonly understood to increase with age, but actually it doesn't in the age range from 60 to 80 in Western populations then suddenly doubles in the over 85s compared to those less than 80.^[2][3]

Unhappily this complexity along with difficulties in managing the knowledge service causes doctors to simplify anticoagulation decisions, regarding it as an art and often making decisions where the science indicates otherwise.

Choice of anticoagulant

Essentially for years this was between parentally administered heparin and orally administered coumadin analogues such as warfarin. Heparin was used for initiation in lifethreating thromboembolic disease, the pregnant and thromboprohylaxis and coumadin analogues in maintenance. Comparative trials have tended to favour low molecular weight heparins over unfractionated heparin in terms of convenience and safety. There is evidence that heparins can be more clinically effective in some indications than warfarin but no evidence for a mortality benefit ^[4]. Again the limited head to head trials of LMWH products seem to show no definite difference at effective dose but there are certainly slightly different indications and the effective dose is indication dependent. As a rule the new anticoagulants (anti-factor Xa agents and direct thrombin inhibitors) tend to be oral, be at least equivalent to LMWH in efficacy, but as new products their long term safety record is yet to be established.

Length of anticoagulation

Guidelines exist and balance risk against benefit assuming single pathology and the population median change with time^[5]. It is an area where ever more refined information allowing evaluation of risk for individuals will accumulate. For example unprovoked thromboembolism has long been known to be associated with a higher risk of recurrent thromboembolism than in patients where the risk factor is addressed. It has only recently been demonstrated that in unprovoked thromboembolism it is females less than 65, with no lower leg hyperpigmentation, oedema or redness of either leg, low D-dimer while taking warfarin and a body mass index less than 30 kg/m that have the lowest risk for recurrence when warfarin is discontinued. Accordingly an argument can now be made, all other things being equal in unprovoked thromboembolism to continue anticoagulation long term in all males and all elderly females or young females with obesity, continued activation of D-dimer or chronic lower leg pathology^[6].

See Also

• Warfarin
• Heparin
• Peri-operative anticoagulation

References