Atrial Fibrillation is a heart condition in which people are treated with blood thinners such as warfarin to decrease the risk of stroke. Large studies have shown that when patients adjust their own dose of warfarin, similar to insulin, results are better.

The purpose of this study is to evaluate whether implementing this method of warfarin management is beneficial in a Canadian primary care clinic. Patients will be educated on how to adjust their own warfarin doses when necessary using simple charts. The success of patient self management will be compared against management by a physician.

BACKGROUND Oral anticoagulants have seen large popularity due to their ease of administration and proven benefit in a variety of conditions. In an examination of patients in Manitoba, 7.2% of the elderly population was being treated with warfarin therapy, a proportion which is likely similar across Canada. For atrial fibrillation, the dose of warfarin is titrated using INR as a guide to a target of 2.0-3.0, a therapeutic range with significant evidence of decreased thromboembolic events while minimizing the risk of major hemorrhage2. Other indications for anticoagulation therapy include presence of a mechanical heart valves, a history of deep vein thromboses or pulmonary embolism(s), or other pro-thrombotic conditions.

Ever since the introduction of warfarin into routine clinical practice, maintenance of a therapeutic INR in patients has proven to be a difficult task, with patients only spending an average of 57-66% of the time within therapeutic range. As the majority of these patients in Canada are treated by primary care providers, busy family physicians must keep a watchful eye on INR values, which are constantly changing due to diets with varying amounts of Vitamin K, not to mention the effects of interacting drugs and other confounders.

In an attempt to find alternative strategies for anticoagulation treatment several randomized control trials have been undertaken in the last two decades examining the idea of "Patient Self-Management" (PSM) of anticoagulation. In a scheme similar to diabetics monitoring their blood glucose values and altering insulin dosages, patients in these trials have been taught to self-test their INR values using home electronic devices and then adjust their warfarin doses accordingly. Data has been published showing improved control in therapeutic range and decreased complication rates of thromboembolic events and major hemorrhages. There was no significant decrease in mortality in the above studies, although most were not powered to detect this as sample sizes were small (49-737 participants). In 2006 a systematic review and meta-analysis of 14 randomized control trials (RCT's) was published involving 3049 participants comparing "Patient Self-Monitoring" (ie. no adjustment of dose) and/or PSM to standard management Data from seven PSM studies was included which demonstrated a significant decrease in thromboembolic events, a non-significant decrease in major hemorrhagic events, and a significant decrease in mortality. Furthermore, it was shown that patients were more satisfied with PSM strategies. These data are reflected in the American College of Chest Physicians guideline entitled Pharmacology and Management of Vitamin K Antagonists, "In patients who are suitably selected and trained, PST [patient self-testing] or PSM is an effective alternative treatment model. We suggest that such therapeutic management be implemented where suitable".

In an attempt to compare the success of anticoagulation control in all arenas, Walraven and associates examined 67 studies involving 50 208 patients with 57 154 patient-years from anticoagulation clinics, clinical trials, and community practises. Study investigators examined many variables which may have influenced differences in therapeutic control and concluded that study setting was the greatest predictor. They quoted the following rates of anticoagulation success: PSM 72%, RCT's 66%, anticoagulation clinics 66%, and community primary care clinics 57%. It was also shown that PSM was associated with a significant improvement in anticoagulation control.

In considering the practise of family medicine in Canada, two facts regarding anticoagulation control are apparent: (a) family physicians shoulder the majority of the treatment responsibility; and (b) control is likely not optimal. PSM in RCT's have shown to improve therapeutic control in patients when compared to "usual care"5,7 and also to care from anticoagulation clinics in tertiary care hospitals. This improved control has shown to be associated with improved clinically relevant endpoints. Thus, it would be logical to suggest that implementing PSM strategies in Canadian primary care settings would have the greatest return. There are several problems with this logic though. Firstly, RCT's all involved extensive training sessions with nurse clinicians or specialized physicians at tertiary care centres to educate participants in the fundamentals of warfarin therapy and the adjustment regimes. These resources are simply not available for family physicians. Secondly, "point of care" electronic INR testing devices were provided to all patients, which would cost an average patient approximately $1000 with the added cost of testing strips.

OBJECTIVE It is the purpose of this small study to determine the feasibility of implementing a PSM strategy in an average Canadian primary care practise.

STUDY DESIGN The study will involve patients from a single family private practise in Chilliwack, BC. It will be carried out in a randomized, unblinded cross-over trial for a duration of eight months (four months per arm).

All patients from the practise of interest will be considered for the study and will be evaluated for inclusion and exclusion criteria as well as desire to participate. Criteria for participation is based on the design of above mentioned studies which showed benefits of PSM. 20 patients will then be randomized to two groups using a computer randomizer. Group A will be assigned to PSM for a duration of four months and then switched to physician-management for the remaining four months of the study. Group B will follow the opposing schedule. Patients in the PSM group will obtain their INR values from a community laboratory and then adjust their Warfarin dose. Dosage adjustments will be guided by dose adjustment nomograms (provided to patients) which are based on the recommendations of the BC Guidelines and Protocols Advisory Committee. Data from PSM will be compared to results from physician-management and also to accepted values for community anticoagulation control in the literature.

INTERVENTION All patients considered for the study will be introduced to the idea of PSM and theory behind anticoagulation in a normal office visit with a physician of approximately half an hour. The consent form will also be explained and given to the patients. They will then return for a follow-up in approximately one week after considering the undertaking. If participation is desired and consent is obtained the patient will be randomized into either group A or group B. Patients of group A will have the task of managing their own warfarin therapy using the provided nomograms. Patients of group B will continue to be managed by their physician. In the second office visit patients will be told to which group they were randomized and if they are in group A they will then be instructed on how to use the warfarin adjustment nomograms provided. After four months the groups will switch to the alternate management strategy. This time those in group B will be instructed on how to use the nomograms during an office visit. Patients in both groups will use warfarin tablets of two dosage strengths, 1 mg and 5 mg. All INR testing will be done at the community laboratory and results will be made available to the patient for dosage adjustment, as well as the physician. The physician's office will be available for appointments and phone calls from study subjects to discuss and support the process at any time.

RESULTS All INR results will be forwarded to the community physician's office and will thereby be recorded as data to determine the proportion of values in therapeutic range. Statistical calculations will be done according to accepted practices for analysis of cross-over studies by a researcher with experience in statistical analysis. Patients will be given a structured treatment-related quality of life survey after the before and after the PSM phases and will explore topics of general treatment satisfaction, self-efficacy, strained social network, daily hassles, and distress. The survey consists of 32 questions and has been previously described and validated.4 Other outcomes including office visits utilized for issues of PSM, and complications of sub- or supra-therapeutic coagulation will be recorded. Secondary outcomes are not expected to be significant due to the limited power of the study.

Inclusion Criteria:

• Over 18 years of age
• Anticoagulation to a target of 2.0-3.0
• Warfarin treatment for > 3 months
• previous medication treatment adherence
• competence judged by demonstrable ability to utilize drug adjustment nomograms
• understand the basic theory of anticoagulation therapy

Exclusion Criteria:

• Coagulopathic disease
• significant psychiatric illness
• significant language barrier
• poor visual acuity