Potential clinical and economic outcomes of CYP2C9 and VKORC1 genotype-guided dosing in patients starting warfarin therapy.

The US Food and Drug Administration has updated the label information for warfarin to encourage the use of genetic information before initiating treatment with the drug. We used decision-tree modeling to simulate the outcomes of CYP2C9 and vitamin K epoxide reductase complex 1 (VKORC1) genotype-guided dosing in patients in whom warfarin therapy is to be initiated. The inputs for the model were derived from the literature. The incremental costs per unit outcome improved (ICERs) were US$347,059 per quality-adjusted life-year (QALY) gained, $170,192 per adverse event averted, and $1,106,250 per life saved. The outcomes of 10,000 Monte Carlo simulations demonstrate that the ICER per QALY gained was >$50,000 62.1% of the time. ICER was sensitive to baseline international normalized ratio (INR) control, reduction in out-of-range INRs by genotype-guided dosing, and genotyping cost. In conclusion, genotype-guided dosing for warfarin therapy does not appear to be cost-effective, with the potential ICER per QALY being >$50,000. Lowering the genotyping cost, improving effectiveness of INR control of the genotype-guided dosing algorithm, and applying the algorithm in practice sites with high out-of-range INRs would improve the cost-effectiveness of the dosing algorithm.