MELD-Na、MELD及Child-Pugh评分系统预测人工肝治疗的肝衰竭预后价值的探讨

摘要

Objective: To evaluate the value of MELD-Na scoring system, MELD scoring system and Child-Pugh scoring system for liver failure with artificial liver support. Methods: The values of MELD-Na scoring system, MELD scoring system and Child-Pngh scoring system were evaluated using receiver operating characteristic (ROC) curves. Results: The area under curve (AUC) values generated by the ROC curves for Child-Pugh score were higher (AUC=0.794) than those of ME LD-Na score (AUC=0.724) and MELD score (AUC=0.664) respectively. The eutoff scores of three systems were 10.5, 24.8, 26.4 respectively, which could discriminate higher and lower mortality accurately. There were no significant statistic differences in predictive values of three systems for different liver failure(sub-acute liver failure and chronic-on-acute liver failure). But the Child-Pugh scoring system was the best for prediction of the chronic liver failure. Conclusion: MELD-Na scoring system,Child-Pugh scoring system and MELD scoring system can predict the prognosis of liver failure, in which the Child-Pugh scoring system was the best.%目的:评价MELD-Na、终末期肝病模型(MELD)及Child-Pugsh评分系统对预测人工肝治疗的肝衰竭预后的价值.方法:使用受试者工作特征(ROC)曲线对Child-Push、MELD-Na及MELD评分系统的诊断价值进行评估.结果:ROC曲线评估Child-Push,MELD,MELD-Na评分系统预测预后的能力,显示Child-PUsh评分系统的曲线下面积(AUC)为0.794,显著高于MELD(Auc为0.664)和MELD-Na(AUC为0.724),差异均有统计学意义,其各自的临界值(cut-off值)分别是10.5、24.8、26.4.对于不同肝衰竭分型3种评分系统的比较发现亚急性和慢加急性肝衰竭3种评分系统的预测性差异无统计学意义,而慢性肝衰竭的Child-Push评分系统的预测性更高.结论:Child-Push、MELD和MELD-Na评分系统对于肝衰竭的诊断均有较好的预测性,Child-PIlsh评分系统对肝衰竭预后判断的价值更高.