术前预测肝脏切除术后肝衰发生的风险因素的初步分析  

Preoperative predictors and risk factors of liver failure post hepatectomy

作　　者：苏松[1] 陈川[1] 李波[1] 贺凯[1] 冯春红[1] 张孟愈[1] 夏先明[1] 

机构地区：[1]四川省泸州医学院附属医院肝胆外科,四川泸州646000

出　　处：《现代预防医学》2014年第17期3244-3248,共5页Modern Preventive Medicine

基　　金：四川省泸州市科技局重点资助项目(2010-S-14(5/16))

摘　　要：目的探索常用临床检测指标中预测肝脏切除术后肝衰发生的风险因素。方法回顾性分析99例资料齐全的肝切除术患者的临床资料,并将其是否发生肝衰分为肝衰组和非肝衰组,采用t检验、χ2检验、秩和检验比较等统计方法逐一比较两组间各项常规血清学,以及生理学指标,筛选存在统计学差异的指标,即为风险因素,同时我们采用logistic回归分析尝试建立肝衰预测模型,并与Child、MELD等临床常用肝功评估系统进行ROC曲线比较。结果根据选定标准,99例患者中有15例发生肝衰,肝衰组与非肝衰组有15个指标存在统计学差异,其分别为:病变性质,乙肝表面抗原,天门冬氨酸氨基转移酶、直接胆红素、总胆汁酸、碱性磷酸酶、白细胞、血小板、血清白蛋白、胆碱酯酶、α-L-岩藻糖苷酶、凝血酶原时间、凝血酶原时间的国际标准化比率、凝血酶原时间活动度、活化部分凝血酶时间。根据得到的风险因素,我们初次建立了术前预测术后肝衰的风险模型P=1/{1+e-[0.691×ln(AFP)+2.397×ln(AFU)+1.337×ln(SF)-13.152]},在回顾性分析中,该模型ROC曲线下面积为0.948,明显高于Child评分(0.723)和Meld评分(0.819)(均P<0.05)。结论我们通过统计学研究发现多个术前判断肝脏功能以及预测肝衰发生的临床指标,这些指标的发现为建立适合国人的术前肝脏储备功能评估系统做了一定的铺垫。Objective To investigate the risk factor predicting the occurrence of liver failure post hepatectomy. Methods 99 cases of hepatectomy were enrolling in this retrospective study. All cases were divided into liver failure group（LF） and non liver failure group（NLF） according to whether suffered the liver failure. The physiological and routine serum markers were compared respectively using t test, χ^2test, rank sum test and the statistical different marker was defined as the risk factor. Then we tried to establish the liver failure prognosis model using logistic regression analysis and compared to Child and MELD system by Receptor Operation Curve（ROC）. Results Among all the cases underwent hepatectomy, 15 cases suffered the liver failure in accordance with the criteria.Between the LF and NLF groups, 15 markers showed statistical significance, which included lesion nature, hepatitis B virus surface antigen, aspartate aminotransferase, conjugated bilirubin, total bile acids, alkaline phosphatase, white blood cells, platelets, serum albumin, cholinesterase, alpha-L-fucosidase, prothrombin time, international normalized ratio of prothrombin time, prothrombin time activity, activated partial thromboplastin time. After that, the liver failure prognosis model： P=1/{1＋e-[0.691×ln（AFP） ＋ 2.397×ln（AFU） ＋1.337×ln（SF）-13.152]} was established using logistic regression analysis. Then the new model was compared to Child and MELD system by ROC curve, and Area Under Curve（AUC） of new model was 0.948, which was statistically higher than Child（0.723） as well as Meld（0.819）. Conclusion The markers we found via statistical analysis to predict the occurrence of liver failure after hepatectomy, could contribute to build a preoperative liver reserved function analysis system suitable for natives.

关 键 词：肝脏切除术 肝功能衰竭 并发症 肝脏储备功能 预测因素 

分 类 号：R61[医药卫生—外科学]