Development and Validation of Clinical Prediction Rules for Bleeding for Patients on Anticoagulant Therapy for Venous Thromboembolism.

Oral anticoagulants (for example warfarin) are commonly used for the long-term
treatment of venous thrombosis. Oral anticoagulants work well to prevent new
thrombotic events. However, physicians and patients need to weigh the benefits
of oral anticoagulant therapy (preventing new blood clots) against its
potential harm (side effects such as bleeding).

Physicians are not able to predict which patients are likely to have bleeding
outcomes. Much effort has gone into developing ways to predict which patient
are at risk of clotting but almost no work has gone into ways of predicting
which patients would be at high risk of bleeding. Investigators have previously
published rules to predict major bleeding or have indicated clinical variables
that suggest an increased risk of major bleeding but no rule to predict major
bleeding on anticoagulants exists that can be recommended for clinical
practice.
This information is required to balance off the risk-benefits and to enable
physicians and patients to understand the risks and benifits of taking these
medications. This study will develop a tool that can be used to predict
bleeding risk in patients taking oral anticoagulant therapy. The ability to
accurately predict hemorrhagic risk will enable intervention studies and have
direct clinical applicability with regard to individualized decision making
regarding the benefit/risk ratio of long term OAT and will result in a better
control of these drugs.

General objective:
To develop or validate a clinical prediction rule for major bleeding in
patients on oral anticoagulant therapy (OAT) who have been safely
anticoagulated without bleeding or venous thromboembolism (VTE) recurrence for
6 months since diagnosis and are being considered for long-term OAT.

Specific Objectives:
To determine in patients with unprovoked VTE on OAT beyond 6 months:
1) a new bleeding risk prediction rule
2) if previous rules (Beyth, van der Meer) are valid for the prediction of
major bleeding
3) if variables used in previously published prediction rules are reproducible
between observers
4) the factors that influence OAT control
5) the degree to which being over the therapeutic INR range influences major
bleeding, independent of OAT control.

Observational, prospective multicenter cohort study.

Enrollment: A visit is planned for blood sampling and the patient will be asked
about things that might affect the dose of oral anticoagulant or risk of
bleeding (e.g. age, current or past health problems, medications).

After the enrollment, a half yearly follow-up is planned, this will be done by
an interview by telephone. Patients will be asked questions about bleeding and
the signs and symptoms of a new thrombotic event. There will also be questions
about changes in health situation and medications.

This is an observational study so there are no risks involved. Participating
in the study will not change the risk of a new thrombotic event or bleeding.