Abstract: AIM: To evaluate the efficacy and safety of warfarin dosage algorithms in clinical application by the use of Meta-analysis.METHODS: Cochrane library (1995-2013),Pubmed (1995-2013), EMBASE (1995-2013), CNKI (1995-2013), CBM(1995-2013) and Wanfang Data were searched by computer for investigating the prospective studies . The quality of included studies and data adoptions were assessed by Cochrane system. A Meta analysis for included Study was performed with The Cochrane Collaboration's Review Manager 5.1.RESULTS: Nine prospective studies including 5 in English and 4 in Chinese were included(n＝3238).The results of Meta analysis showed that: (1) warfarin dosage algorithms could decrease the risk of adverse reactions in the course of warfarin treatment[RR=0.56,95%CI(0.44,0.77),P<0.01]. The results had statistical difference and sensitivity analysis showed that the analysis results were stable and reliable. (2)The mean times to reach stable dosage:5 studies involved have heterogenicity(P<0.01,I²=97%).The mean times to reach stable dosage in the individual group were shorter than those in the control after excluding two studies of low quality[SMD=-0.97,95%CI(-4.36,3.68),P<0.01],and had statistical difference . (3) The adjusted dosage in the individual group was less than that in the control and had statistical difference[SMD=-0.74,95%CI(-1.23,-0.26),P=0.003].CONCLUSION: Warfarin dosage algorithms could improve the efficacy of anticoagulant therapy and decrease the risk of thrombocmbolism and hemorrhagic events.

Key words: warfarin, genetic polymorphism, Meta-analysis