The body has its own natural anticoagulation mechanism. A person should consider using artificial anticoagulation when he or she has enough of an increased risk of forming a thrombus to justify the risks of artificial anticoagulation. Some of the circumstances that make long-term artificial anticoagulation attractive are: having certain types of prosthetic heart valves, having atrial fibrillation, and recurrent stroke, pulmonary embolism or deep vein thrombosis in the legs. Reasons for short-term (3 to 6 months) anticoagulation include first episodes pulmonary embolism or deep vein thrombosis, documented left ventricular thrombus following acute myocardial infarction.

Several new medicines for oral anticoagulation have recently become available. The new medicines which are FDA approved include dabigatran (Pradaxa), rivaroxaban (Xarelto), and apixaban (Eliquis). These are promising because the dose for each patient usually does not have to be adjusted, there are no dietary restrictions, they take effect more quickly than the gold standard medicine, warfarin, and in some cases the studies suggest that they are more effective than warfarin.
It is important to know that there is no antidote for any of these new medicines. Unlike with warfarin, which can be neutralized if the patient begins to bleed, the only options with these agents is to give blood and wait until the medicine wears off. This takes at least 12 hours and, depending on the medicine, 24 or more hours, and can be a real problem if the bleeding is in a critical area such as inside the head.
In addition, because they are new, many physicians are being cautious is using them until their side effects are better understood. You may wish to ask your doctor whether one of these medicines would be better for you than warfarin, but do not be surprised by either, "I'd rather wait and see" or by "Yep, this would be a good choice for you to consider."

Oral anticoagulation using warfarin is the oldest method of long-term anticoagulation so physicians have much experience with its benefits and its risks. The brand name form of warfarin is Coumadin, which many clinicians prefer because of its reliability. Experience has proven that when cost-cutting efforts have forced people to change the form of warfarin they take every day, they have an increased rate of complications. (Note: neither we nor our consultants have any interest in Coumadin or its manufacturer other than as health care providers who are satisfied with the performance of the product.)

Warfarin is taken once a day. Its dose must be adjusted daily, based on the result of a blood test called the "Prothrombin time", until the blood test result is satisfactory day after day. Then, it is tested weekly, and finally monthly (as a practical matter). The reason for all this testing is that the dose required for safe and effective anticoagulation varies a lot from person to person. Moreover, it varies in the same person from week to week depending on the diet. Because warfarin acts by opposing the body's ability to use Vitamin K to the proteins it needs for normal coagulation, the dose of warfarin changes if a person eats more or less Vitamin K than he or she usually does. Everyone who takes warfarin should get a booklet from his or her doctor listing the amount of Vitamin K in a variety of foods. Foods that contain a particularly large amount of Vitamin K include beef liver, pork liver, green tea, broccoli, chick peas, kale, turnip greens, and soybean products including soybean oil, soybeans themselves, soy milk, and tofu.

More information is available here. Medicines can also change the effect of warfarin on the body. Everyone who takes warfarin should make sure her or his doctor considers the effect on the warfarin whenever a medication change is made. More information on Coumadin is available here.

The best dose of warfarin (or of any medicine, for that matter) is the lowest dose that accomplishes the desired result. Considerable research is still going on to determine this dose for warfarin. The problem is difficult because the problems that come with too little warfarin are so bad that it is not ethical to give a dose that has much chance of being too low. On the other hand, higher doses of warfarin are clearly associated with higher risks of major bleeding that can be life-threatening. Current medical thinking relies on the International Normalized Ratio, or INR, of the blood test called the prothrombin time.

The INR is calculated mathematically in a way that corrects for differences in the method used for measuring the prothrombin time. It makes the results more comparable from place to place and at different times in the same place. The INR is a ratio of a clotting time compared to "normal". A normal person would have an INR of about 1.0. People with atrial fibrillation are often told to maintain an INR of 2.0 to 3.0. People with artificial valves and other reasons to have a strong reason to develop dangerous blood thrombi (like clots, but inside the body) are often advised to maintain an INR of 2.5 to 3.5. Newer research suggests that sometimes the INR can be maintained at a lower level (that is, closer to 1.5), but each person should consult his or her own physician about the best INR for him or her.