Semi-empirical Anticoagulation Model (SAM): Dose prediction during warfarin therapy

Area of Science:

• Pharmacology
• Biomedical Engineering
• Clinical Medicine

Background:

• Oral anticoagulant therapy management is crucial for patient quality of life.
• Existing computerized dosing algorithms have limitations leading to inaccurate warfarin dosing.
• Individual patient sensitivity significantly impacts warfarin response.

Purpose of the Study:

• To propose a novel warfarin dose prediction algorithm incorporating individual patient sensitivity analysis.
• To enhance the accuracy of warfarin dosing recommendations by accounting for patient-specific factors.
• To improve the management of oral anticoagulant therapy for conditions like venous thromboembolism.

Main Methods:

• Development of a warfarin dose prediction algorithm based on the Semi-empirical Anticoagulation Model (SAM) extended with stochastic process analysis.
• Retrospective observational study involving 1796 patients over one year.
• Comparison of algorithm-suggested warfarin doses with clinically administered doses using Percentage Bland-Altman Analysis.

Main Results:

• The proposed algorithm demonstrated an average accuracy of (3.24±25.80)%.
• This accuracy is significantly better than other existing algorithms in the literature, which showed an average accuracy of (5.73±60.9)%.
• The algorithm effectively assesses the probability of events affecting the International Normalized Ratio (INR) and therapy relationship.

Conclusions:

• The novel algorithm provides more flexible and precise warfarin dosing adjustments.
• Improved accuracy leads to an International Normalized Ratio (INR) closer to the target range with reduced variability.
• Enhanced warfarin management improves patient safety and quality of life for patients with venous thromboembolism.