华法林抗凝质量影响因素的Meta分析与预测模型

Factors affecting the anticoagulation quality of warfarin: a meta-analysisand prediction model

目的 通过构建华法林抗凝质量预测模型,提高华法林抗凝效果。方法 计算机检索国内外数据库,对建库之初至2022年5月1日之间的所有关于华法林抗凝质量影响因素的研究进行Meta分析,筛选影响因素,构建华法林抗凝质量的预测模型。结果 共获得3390篇文献,最终纳入文献17篇,包含66 119名患者,Meta分析结果显示:年龄、糖尿病、心力衰竭、慢性阻塞性肺疾病、肝脏疾病、肾脏疾病是治疗范围内时间百分比(TTR)达标的影响因素。以此构建的华法林抗凝质量预测模型为Logit(P)=-0.819+0.015X_(年龄)-0.146X_(糖尿病)-0.180X_(心衰)-0.209X_(慢阻肺)-0.511X_(肝脏疾病)-0.464X_(肾脏疾病),ROC曲线下面积为0.807,95%置信区间为(0.715,0.900),最佳临界值为Logit(P)=-1.238、敏感性为0.829、特异性为0.714。结论 基于Meta分析构建的华法林抗凝质量预测模型能有效预测患者TTR是否达标,预测效能中等,但仍需大量数据验证。

Objective To establish the prediction model of warfarin anticoagulation quality,and to improve anticoagulation of warfarin.Methods We searched domestic and foreign databases,conducted a meta-analysis of all relevant studies on factors affecting the anticoagulation quality of warfarin from the beginning of the database to May 1st,2022.We also screened the influencing factors,and constructed a prediction model of anticoagulation quality.Results Totally 3390 articles were retrieved from the databases,and 17 studies were included,involving 66119 patients.The meta analysis showed that age,diabetes mellitus(DM),heart failure(HF),chronic obstructive pulmonary disease(COPD),liver disease(LD),and kidney disease(KD)were the risk factors forthe standard time within therapeutic range.The prediction model of warfarin anticoagulation quality was Logit(P)=-0.819+0.015X_(Age)-0.146X_(DM)-0.180X_(HF)-0.209X_(COPD)-0.511X_(LD)-0.464X_(KD).The AUC of ROC was 0.807,the 95%confidence interval was(0.715,0.900),the best critical value was Logit(P)=-1.238,the sensitivity was 0.829,and the specificity was 0.714.Conclusion The prediction model of warfarin anticoagulation quality based on meta-analysis can predict whether patients’TTR reaches the standard with moderate prediction efficiency,which still needs to be verified with more data.