Development of a Bayesian forecasting method for warfarin dose individualisation
Daniel F. B. Wright, Stephen B. Duffull
School of Pharmacy, University of Otago, PO Box 56, Dunedin, New Zealand
Objectives: Warfarin is the most commonly prescribed oral anticoagulant worldwide. It is a difficult drug to dose accurately due to a large inter-individual variability in response and a narrow therapeutic range. Warfarin response is routinely monitored using the International Normalised Ratio (INR), a measure of blood clotting time. If the INR falls below 2, the patient is at risk of clotting while INRs above 4.5 carry an increased risk of major bleeding events. In addition, there is a delay between a change in the dosing regimen and achievement of the steady state INR which means that monitoring is often confounded by non-steady state conditions. Not surprisingly, warfarin dose individualisation constitutes a major challenge for clinicians, with reports suggesting that patients achieve therapeutic INRs only 50-60% of the time [1-4]. A simple tool for individualising warfarin dosages will therefore have significant benefits for healthcare.
The aim of this study was to develop a Bayesian dose individualisation tool for warfarin. This was incorporated into the freely available software TCIWorks (http://www.tciworks.info/) for use in the clinic.
Methods: All PKPD models of warfarin in the medical literature were identified and evaluated against two warfarin pharmacokinetic and pharmacodynamic datasets. The model with the best external validity was used to develop an optimal design for Bayesian parameter control. The performance of this design was evaluated using simulation-estimation techniques. Finally, the model was implemented in TCIWorks.
Results: A recently published warfarin KPD model was found to provide the best fit for the two external datasets [5]. Optimal sampling days within the first 14 days of therapy were found to be days 3,4,5,11,12,13 and 14. Simulations and parameter estimations suggested that the design will provide stable estimates of warfarin clearance and EC50. A single patient example showed the potential clinical utility of the method in TCIWorks.
Conclusions: A Bayesian dose individualisation tool for warfarin was developed. Further research to assess the predictive performance of the tool in warfarin patients is underway.
References:
[1] Tan ES et al. Am J Ger Pharmacol. 2007, 5(3):232-5.
[2] The Stroke Prevention in Atrial Fibrillation Investigators. Lancet. 1996;348:633-8.
[3] van Walraven C et al. Chest. 2006;129:1155-66.
[4] Rose AJ et al. J Thromb Haemost. 2009;7:94-101.
[5] Hamberg AK et al. Clin Pharmacol Ther. 2010;87(6):727-34.
