Improvement of anticoagulant treatment using a dynamic decision support algorithm: a Danish Cohort study

Area of Science:

• Pharmacology
• Medical Informatics
• Biostatistics

Background:

• Warfarin is a widely prescribed vitamin K antagonist for long-term oral anticoagulant therapy, with over 10 million patients in the US and Europe.
• Maintaining patients within their International Normalized Ratio (INR) therapeutic range is crucial for effective anticoagulation and minimizing risks.

Purpose of the Study:

• To retrospectively validate a dynamic statistical model designed to provide warfarin dosage suggestions.
• To assess the model's accuracy in predicting future INR values and optimal warfarin doses for patients.

Main Methods:

• Validation on a cohort of 553 patients with a mean Time in Therapeutic Range (TTR) of 83%.
• The predictive model incorporates INR history, warfarin dosage, and biological noise, with parameters that individualize over time.
• Model performance was assessed by comparing its dosage suggestions to observed warfarin intake.

Main Results:

• The model achieved a median absolute error of 0.53 mg/day in predicting warfarin dosage.
• In over 70% of cases with INR measurements outside the therapeutic range, the model suggested a more appropriate warfarin dose than the patient's actual intake.
• The model demonstrated accuracy in predicting future INR values and optimal warfarin doses.

Conclusions:

• The validated dynamic statistical model shows potential to further increase the Time in Therapeutic Range (TTR) beyond the current 83%.
• Implementing this dosing algorithm could lead to improved anticoagulation management and patient outcomes.