• Accuracy of the initial versus the stable warfarin dose, measured as mean absolute difference in initial versus stable dose
  

• Time to stable dose
  
• The frequency of subtherapeutic and supratherapeutic international normalized ratio (INR) measurements
  
• The fraction of population overdosed and underdosed at warfarin initiation
  

Warfarin is the mainstay of therapy in preventing venous thromboembolism (VTE), pulmonary embolism (PE), and subsequent morbidity and mortality. Despite its proven efficacy in reducing the advent of clot formation, patient-specific warfarin dosing is difficult to predict, with the initial dose regimen often resulting in supra- and subtherapeutic anticoagulation, and subsequently increasing patients' risk of bleeding or embolism. It has been shown that interpatient warfarin dosing variability is due in part to genetic variations found in the cytochrome P450 2C9 metabolism pathway (CYP2C9), as well as proteins involved in the coagulation cascade, most importantly vitamin K epoxide reductase complex subunit 1 (VKORC1). A recent retrospective study has shown that these two genes in addition to several clinical/demographic factors account for greater than 58% of the variation in patient-specific warfarin doses. However, there have been no studies documenting prospective use of this information in selecting an initial warfarin dose. Hypothesis: Stable therapeutic management of warfarin therapy can be more precisely achieved when patients are prospectively dosed based on a pharmacogenetic-guided dosing equation compared to usual care. Aim a: To determine if pharmacogenetic-guided dosing of warfarin is superior to usual care, when defined as the accuracy of the initial versus the stable warfarin dose. This will be assessed as the mean absolute difference in initial versus stable dose. Aim b: To determine if a stable warfarin dose is more quickly achieved using the pharmacogenetic-guided dosing equation compared to usual care. This will be assessed as time to stable dose. Aim c: To determine if pharmacogenetic-guided dosing is superior to usual care in terms of overdosing and underdosing patients. This will be assessed as the fraction of the population overdosed and the fraction of population underdosed by the two methods. We propose to evaluate a pharmacogenetic-guided dosing approach compared to usual care in the initiation of warfarin management. The selected pharmacogenetic-guided equation is a validated equation that includes both genetic and clinical factors and is relatively easy to incorporate into current clinical practice. Patients newly initiating warfarin therapy in a hospital setting will be randomized to receive either pharmacogenetic-guided or usual care, with follow-up anticoagulation management services provided by the University of Florida Health System Anticoagulation Clinic. Prospectively determining patients' stable dose has important clinical implications in today's management of warfarin therapy. Being able to predetermine a patient's stable dose upon initiation of therapy has the potential to decrease the time spent in supra- and subtherapeutic anticoagulation and reduce the number of clinic visits required to achieve a stable dose. Therefore we propose this study to test if using genotype data in determining the initial warfarin dose is more effective than usual care in predicting stable dose. If we can document the value of such an approach, this will provide the level of evidence needed to translate pharmacogenetic-guided dosing of warfarin into clinical practice.