3种华法林剂量预测模型与汉族人群的维持剂量相关性评价

目的：评价3种不同华法林剂量预测模型对汉族人群华法林每日维持剂量预测的准确性。方法：选取86例行VKORC1、CYP2C9基因多态性检测，华法林抗凝治疗且INR达标患者，收集患者平均每日华法林维持剂量、人口学信息及临床情况等；通过比较患者的实际剂量与预测剂量的差异，评价Gage模型、国际华法林药物基因组协会（IWPC）和药物作用靶点基因检测技术指南（指南）等推荐的华法林剂量预测模型的准确性。结果：Gage模型的预测剂量稍小于患者实际剂量，但没有统计学差异（3.81±0.95）mg vs.（4.19±1.19） mg，P=0.199]，预测准确率61.6%，而IWPC、指南模型的平均预测剂量均低于患者实际剂量（3.48±0.85）mg，P=0.025和（3.38±0.66）mg，P=0.007）]；预测准确率分别为48.8%和41.8%。结论：GAGE模型更适合威海地区汉族人群，但需大样本研究进一步验证其准确性。

A correlation evaluation of three Warfarin dosing algorithms models with maintenance dosage for the Han population

OBJECTIVE To evaluate the correlation of three different Warfarin dosing algorithms models with warfarin maintenance dosage of Han population in Weihai.METHODS VKORC1 and CYP2C9 gene of 86 cases were detected,the maintenance dosage of Warfarin,demographic information and clinical setting were achieved.The correlations of results of three genotypes predictive models and stable warfarin dose were compared.RESULTS The Gage model was slightly less than the maintenance dosage of patients,but the difference was no statistical significant (3.81±0.95) mg vs. (4.19±1.19) mg,P=0.199].The IWPC and the guideline models were lower than the the maintenance dosage (3.48±0.85) mg,P=0.025;(3.38±0.66) mg,P=0.007)].CONCLUSION The Gage model is more suitable for people of Han population than others,but it need further study to verify its accuracy.