【作者】 周锐；

【导师】 周元平；

【作者基本信息】 南方医科大学 ， 内科学（专业学位）， 2014， 博士

【摘要】 目的：对慢加急性乙型肝炎肝衰竭患者的病历进行回顾性分析,收集患者的基本临床资料和临床检测指标,研究下列问题：(1)慢加急性乙型肝炎肝衰竭患者合并肝性脑病的影响因素；(2)慢加急性乙型肝炎肝衰竭患者合并自发性腹膜炎的影响因素；(3)慢加急性乙型肝炎肝衰竭患者合并肝肾综合征的影响因素；(4)慢加急性乙型肝炎肝衰竭患者合并肺部感染的影响因素；(5)慢加急性乙型肝炎肝衰竭患者合并真菌感染的影响因素；(6)慢加急性乙型肝炎肝衰竭患者合并上消化道出血的影响因素；(7)慢加急性乙型肝炎肝衰竭患者合并电解质紊乱的影响因素；(8)建立基于动态变化指标的慢加急性乙型肝炎肝衰竭患者预后预测模型。方法：(1)慢加急性乙型肝炎肝衰竭患者各并发症的影响因素分析方法：830例慢加急性乙型肝炎肝衰竭患者均为2004年1月至2010年12月在我院住院治疗的患者。诊断符合2012年制定的《肝衰竭诊治指南》。收集患者基础临床资料及诊断慢加急性乙型肝炎肝衰竭时的基线临床检测指标。基础临床资料包括性别、年龄(岁)、家族史、病程中发生肝性脑病、自发性腹膜炎、肝肾综合征、肺部感染、真菌感染、上消化道出血、电解质紊乱情况,基线临床检测指标包括白蛋白(g/l)、球蛋白(g/l)、A/G、总胆红素(μmol/L)、直接胆红素(μmol/L)、总胆红素/直接胆红素比值、ALT(I/L)、AST(U/L)、ALT/AST、Y-谷氨酰转移酶(U/L)、碱性磷酸酶(U/L)、胆固醇Immol/L)、胆碱酯酶(U/L)、血钾(mmol/L)、血钠(mmol/L)、肌酐(μmol/L)、凝血酶原活动度(%)、国际标准化比值(INR)、甲胎蛋白(ng/ml)、HBVDNA(IU/ml)、HBeAg定性、白细胞计数(*109/L)、中性粒细胞比率(%)、血红蛋白(g/l)和血小板水平(*109/L)。为减少误差,计量资料中的总胆红素、直接胆红素、ALT、AST、胆碱酯酶、凝血酶原活动度、甲胎蛋白、HBV DNA水平转化以10为底的对数结果进行分析。统计学分析用统计学软件SPSS13.0进行。计量资料采用t检验或者非参数检验,计数资料采用x2检验或Fisher’s精确检验。二分类多因素logistic回归分析时,计数资料赋值为1或者2,1表示发生阳性事件。(2)基于动态变化指标的慢加急性乙型肝炎肝衰竭患者预后预测模型的建立：714例慢加急性乙型肝炎肝衰竭患者均为2004年1月至2010年12月在我院住院治疗大于7天的患者。诊断符合2012年制定的《肝衰竭诊治指南》。通过病历收集714例患者的基础临床资料及诊断慢加急性乙型肝炎肝衰竭时基线及第7天的临床检测指标。基础临床资料包括性别、年龄(岁)、家族史,基线及第7天临床检测指标包括白蛋白(g/l)、球蛋白(g/1)、A/G、总胆红素(μmol/L)、直接胆红素(μmol/L)、总胆红素/直接胆红素比值、ALT(U/L)、 AST(U/L)、ALT/AST、Y-谷氨酰转移酶(U/L)、碱性磷酸酶(U/L)、胆固醇(mmol/L)、胆碱酯酶(U/L)、血钾(mmol/L)、血钠(mmol/L)、肌酐(μmol/L)、凝血酶原活动度(%)、国际标准化比值(INR)、甲胎蛋白(ng/ml)、HBVDNA(IU/ml) HBeAg定性、白细胞计数(*109/L)、中性粒细胞比率(%)、血红蛋白(g/l)和血小板水平(*109/L)。为减少误差,计量资料中的总胆红素、直接胆红素、ALT、AST、胆碱酯酶、凝血酶原活动度、甲胎蛋白、HBV DNA水平等指标转化为自然对数,计算出基线与第7天除甲胎蛋白、HBVDNA水平外各临床检测指标的差值。死亡赋值为0,截尾为1,男性赋值为1,女性为2,HBeAg阳性赋值为1,阴性为2,并对所有患者进行随访,直至患者发生死亡或研究结束(研究截止日为2011年1月10日)。生存时间定义为入院时间至死亡或失访、终止检查、肝移植手术日、确诊恶性肿瘤的时间。将患者按照85%：15%的比例随机分成2组：模型建立组与模型检测组,病例数分别为607例、107例。COX比例风险回归模型分析得出模型建立组患者的预后影响因素,建立预后评估模型。根据预后评估模型,算出模型检测组患者预后评分,并算出MELD评分值,通过受试者工作特征(ROC)曲线评估本模型与MELD对慢加急性乙型肝炎肝衰竭预后预测的准确性。结果：(1)830例慢加急性乙型肝炎肝衰竭患者年龄(40.7±12.1)岁,其中男性721例(86.9%),女性109例(13.1%)。慢加急性乙型肝炎肝衰竭患者中共出现肝性脑病患者278例(33.5%),其中男性244例(29.4%),女性34例(4.1%),肝性脑病发病率在男女性别间的差异无统计学意义(x2=0.298,P=0.585)。肝性脑病患者年龄42(17)岁,显著高于非肝性脑病患者的38.5(14)岁,差异有统计学意义(Z=-4.46,P=0.000)。单因素分析结果显示：年龄、总胆红素与直接胆红素比值、胆固醇、血钠、凝血酶原活动度、国际标准化比值(INR)、甲胎蛋白、白细胞计数、中性粒细胞比率和血小板水平在发生肝性脑病和未发生肝性脑病的两组慢加急性乙型肝炎肝衰竭患者的差异有统计学意义。Logistic回归分析结果显示,年龄、总胆红素/直接胆红素比值、凝血酶原活动度、国际标准化比值(INR)、甲胎蛋白、HBV-DNA、白细胞计数、中性粒细胞比率对肝性脑病的发生存在重要的影响。(2)830例慢加急性乙型肝炎肝衰竭患者中共出现自发性腹膜炎患者593例(71.4%),其中男性244例(61.3%),女性34例(11.1%),自发性腹膜炎发病率在男女性别间的差异无统计学意义(x2=1.942,P=0.163)。自发性腹膜炎患者年龄40(16)岁,显著高于非自发性腹膜炎患者的37(13)岁,差异有统计学意义(Z=-2.63,P=0.009)。单因素分析结果显示：年龄、白蛋白、A/G、总胆红素、直接胆红素、ALT、AST、ALT/AST、Y-谷氨酰转移酶、碱性磷酸酶、胆固醇、胆碱酯酶、血钾、凝血酶原活动度、国际标准化比值(INR)、HBV-DNA、白细胞计数、中性粒细胞比率、血红蛋白和血小板水平在发生自发性腹膜炎和未发生自发性腹膜炎的两组肝衰竭患者中的差异有统计学意义。Logistic回归分析结果显示,胆碱酯酶、血红蛋白对自发性腹膜炎的发生存在重要的影响。(3)830例慢加急性乙型肝炎肝衰竭患者中共出现肝肾综合征患者90例(10.84%),其中男性76例(9.15%),女性14例(1.69%),肝肾综合征发病率在男女性别间的差异无统计学意义(x2=0.52,P=0.471)。肝肾综合征患者年龄45(15)岁,显著高于非肝肾综合征患者的39(15)岁,差异有统计学意义(Z=-4.06,P=0.000)。单因素分析结果显示：年龄、白蛋白、总胆红素、直接胆红素、总胆红素/直接胆红素比值、ALT、ALT/AST、胆固醇、胆碱酯酶、血钠、肌酐、凝血酶原活动度、国际标准化比值(INR)、甲胎蛋白、白细胞计数、中性粒细胞比率、血红蛋白和血小板水平在发生肝肾综合征和未发生肝肾综合征的两组肝衰竭患者中的差异有统计学意义。Logistic回归分析结果显示,年龄、肌酐、白细胞计数对肝肾综合征的发生存在重要的影响。(4)830例慢加急性乙型肝炎肝衰竭患者中共出现肺部感染患者138例(16.6%),其中男性120例(14.5%),女性18例(2.1%),肺部感染发病率在男女性别间的差异无统计学意义(x2=0.001,P=0.973)。肺部感染患者年龄45(16)岁,显著高于非肺部感染患者的39(15)岁,差异有统计学意义(Z=-4.854,P=0.000)。单因素分析结果显示：年龄、白蛋白、A/G、总胆红素、总胆红素/直接胆红素比值、ALT、AST、ALT/AST、胆固醇、胆碱酯酶、血钾、血钠、凝血酶原活动度、甲胎蛋白、HBV-DNA、白细胞计数、中性粒细胞比率、血红蛋白和血小板水平在发生肺部感染和未发生肺部感染的两组肝衰竭患者中的差异有统计学意义。Logistic回归分析结果显示,年龄、甲胎蛋白、白细胞计数、血红蛋白对肺部感染的发生存在重要的影响。(5)830例慢加急性乙型肝炎肝衰竭患者中共出现真菌感染患者68例(8.2%),其中男性58例(7.0%),女性10例(1.2%),真菌感染发病率在男女性别间的差异无统计学意义(x2=0.161,P=0.688)。真菌感染患者年龄46.5(18)岁,显著高于非真菌感染患者的39(15)岁,差异有统计学意义(Z=-4.424,P=0.000)。单因素分析结果显示：年龄、白蛋白、A/G、白细胞计数、血红蛋白和血小板水平在发生真菌感染和未发生真菌感染的两组肝衰竭患者中的差异有统计学意义。Logistic回归分析结果显示,年龄、白细胞计数对真菌感染的发生存在重要的影响。(6)830例慢加急性乙型肝炎肝衰竭患者中共出现上消化道出血患者83例(10.0%),其中男性70例(8.4%),女性13例(1.6%),上消化道出血发病率在男女性别间的差异无统计学意义(x2=0.518,P=0.472)。上消化道出血患者年龄42(19)岁,显著高于非上消化道出血患者的39(15)岁,差异有统计学意义(Z=-2.317,P=0.021)。单因素分析结果显示：年龄、Y-谷氨酰转移酶、碱性磷酸酶、胆固醇、胆碱酯酶、血钠、肌酐、凝血酶原活动度、国际标准化比值(INR)、白细胞计数、血红蛋白和血小板水平在发生上消化道出血和未发生上消化道出血的两组肝衰竭患者中的差异有统计学意义。Logistic回归分析结果显示,年龄、凝血酶原活动度、血红蛋白对上消化道出血的发生存在重要的影响。(7)830例慢加急性乙型肝炎肝衰竭患者中共出现电解质紊乱患者570例(68.68%),其中男性120例(59.04%),女性80例(9.64%),电解质紊乱发病率在男女性别间的差异无统计学意义(x2=1.299,P=0.254)。电解质紊乱患者年龄40(15)岁,显著高于非电解质紊乱患者的37(15)岁,差异有统计学意义(Z=-3.785,P=0.000)。单因素分析结果显示：年龄、白蛋白、A/G、总胆红素、直接胆红素、ALT、ALT/AST.胆固醇、胆碱酯酶、血钾、血钠、凝血酶原活动度、国际标准化比值(INR)、甲胎蛋白、白细胞计数、中性粒细胞比率、血红蛋白在发生电解质紊乱和未发生电解质紊乱的两组肝衰竭患者中的差异有统计学意义。Logistic回归分析结果显示,白蛋白、ALT/AST、血钾、血钠、甲胎蛋白对电解质紊乱的发生存在重要的影响。(8)714例慢加急性乙型肝炎肝衰竭患者年龄39(15)岁,其中男性622例(87.1%),女性92例(12.9%)；模型建立组患者年龄40(16)岁,其中男性529例(87.1%),女性78例(12.9%)；模型检测组患者年龄38(14)岁,其中男性93例(86.9%),女性14例(13.1%)。至随访结束,714例慢加急性乙型肝炎肝衰竭患者共死亡193例(27.0%),失访、存活或肝移植等521例(73.0%)；模型建立组患者共死亡162例(26.7%),失访、存活或肝移植等445例(73.3%)；模型检测组患者共死亡31例(29.0%),失访、存活或肝移植等76例(71%)。中位随访时间为334d,最长为2558d。模型建立组患者COX比例风险回归模型分析结果得出慢加急性乙型肝炎肝衰竭患者预后影响因素有：白蛋白、ALT/AST比值、凝血酶原活动度、甲胎蛋白、HBVDNA水平、中性粒细胞比率、肌酐差值、凝血酶原活动度差值、中性粒细胞比率差值、血小板水平差值。个体预后指数：PI=0.052*ALB(g/l)-0.203*(ALT/AST)-1.189*LN(PTA%)-0.161*LN[AFP(ng/ml)]-0.06-*LN[HBVDNA(IU/ml)]+0.036*(GR%)-0.003*ΔCR(μmol/l)+1.119*ΔLN(PTA%)-0.028*(ΔGR%)+0.005*Δ[PLT(*109/l)]。绘制受试者工作特征曲线(receiver operating characteristic,ROC)分析结果显示本研究预测模型曲线下面积为79.1%,P=0.000,具有很好的预测价值。MELD预测模型曲线下面积为57.9%,P=0.199。本研究模型评分对应的临界值为-4.98,其灵敏度、特异度分别为80.6%、72.4%。结论：(1)慢加急性乙型肝炎肝衰竭患者合并肝性脑病的危险因素有年龄增大、总胆红素/直接胆红素比值升高、凝血酶原活动度降低、国际标准化比值(INR)升高、甲胎蛋白降低、HBV-DNA高、白细胞计数升高、中性粒细胞比率升高。(2)慢加急性乙型肝炎肝衰竭患者合并自发性腹膜炎的危险因素有胆碱酯酶降低、血红蛋白降低。(3)慢加急性乙型肝炎肝衰竭患者合并肝肾综合征的危险因素有年龄增大、肌酐升高、白细胞计数升高。(4)慢加急性乙型肝炎肝衰竭患者合并肺部感染的危险因素有年龄增大、甲胎蛋白降低、白细胞计数升高、血红蛋白降低。(5)慢加急性乙型肝炎肝衰竭患者合并真菌感染的危险因素有年龄升高、白细胞计数升高。(6)慢加急性乙型肝炎肝衰竭患者合并上消化道出血的危险因素有年龄增大、凝血酶原活动度下降、血红蛋白下降。(7)慢加急性乙型肝炎肝衰竭患者合并电解质紊乱的危险因素有白蛋白降低、ALT/AST降低、血钾下降、血钠下降、甲胎蛋白降低。(8)影响慢加急性乙型肝炎肝衰竭患者预后的因素有白蛋白、ALT/AST比值、凝血酶原活动度、甲胎蛋白、HBVDNA水平、中性粒细胞比率、肌酐差值、凝血酶原活动度差值、中性粒细胞比率差值、血小板水平差值。本研究预后模型具有很好的预测价值,评分PI>-4.98预后差。更多还原

【Abstract】 Objective:Analysed the medical record of acute-on-chronic hepatitis B liver failure patients retrospectively, collected the baseline demographic, clinical, biochemical and virological features, the following issues were studied:(1) The influencing factors of hepatic encephalopathy on acute-on-chronic hepatitis B liver failure patients;(2) The influencing factors of spontaneous bacterial peritonitis on acute-on-chronic hepatitis B liver failure patients;(3) The influencing factors of hepatorenal syndrome on acute-on-chronic hepatitis B liver failure patients;(4) The influencing factors of pulmonary infection on acute-on-chronic hepatitis B liver failure patients;(5) The influencing factors of fungal infections on acute-on-chronic hepatitis B liver failure patients;(6) The influencing factors of upper gastrointestinal bleeding on acute-on-chronic hepatitis B liver failure patients;(7) The influencing factors of electrolyte imbalance on acute-on-chronic hepatitis B liver failure patients;(8) Developed the predictive model on the basic of dynamic change of index in acute-on-chronic hepatitis B liver failure patients. Method:(1) To study the influencing factors of the complications in acute-on-chronic hepatitis B liver failure patients:eight hundred and thirty acute-on-chronic hepatitis B liver failure patients were hospitalized in the department of liver diseases of Mengchao Hepatobiliary Hospital of Fujian Medical University during January2004to December2010, and the diagnosis acute-on-chronic hepatitis B liver failure was followed by The Prevention and Treatment Proposal of Liver Failure established in2012. Collected the baseline demographic, clinical, biochemical and virological features, inclued:gende, age(year), family history, whether hepatic encephalopathy occured or not, whether spontaneous bacterial peritonitis occured or not, whether hepatorenal syndrome occured or not, whether pulmonary infection occured or not, whether fungal infections occured or not, whether upper gastrointestinal bleeding occured or not, whether electrolyte imbalance occured or not, serum albumin (ALB)(g/1), serum globulin (GLB)(g/1), A/G ratio, serum total bilirubin (TBIL)(mol/L), serum direct bilirubin (DBIL)(mol/L), TBIL/DBIL ratio, serum alanine aminotransferase (ALT)(U/L), serum aspartate aminotransferase (AST)(U/L), ALT/AST ratio, serum gamma-glutamyl transferase (GGT)(U/L), serum alkaline phosphatase (ALP)(U/L), total serum cholesterol(CHO)(mmol/L), serum cholinesterase(CHE)(U/L), serum potassium(k+)(mmol/L), serum sodium (Na+)(mmol/L), serum creatinine (CR)(mol/L), prothrombin activity (PTA)(%), international normalized ratio (INR), surem alpha fetoprotein(AFP)(ng/ml), HBVDNA(IU/ml), hepatitis B e antigen(HBeAg) qualitative, white blood cell count (WBC)(*109/L), neutral granulocyte ratio(GR)(%), hemoglobin(Hb)(g/1), platelet count(PLT)(*109/L). For reducing the deviations, the measurement data inclued TBIL, DBIL, ALT, AST, CHE, PTA, AFP, HBVDNA were transformed logarithm with base10. Statistical analysis of data were performed by nonparametric test, Chi-square test, t test, etrietest,Ch-squaretest, and multiple binary logistic regression analysis using the SPSS v.13.0statistical package. Enumeration data were calculated as1or2,1stand for positive events.(2) Developed the predictive model on the basic of dynamic change of index in acute-on-chronic hepatitis B liver failure patients:seven hundred and fourteen acute-on-chronic hepatitis B liver failure patients were hospitalized more than seven days in the department of liver diseases of Mengchao Hepatobiliary Hospital of Fujian Medical University during January2004to December2010, and the diagnosis acute-on-chronic hepatitis B liver failure was followed by The Prevention and Treatment Proposal of Liver Failure established in2012. Collected the baseline and7th day’s demographic, clinical, biochemical and virological features, inclued:gende, age(year), family history, whether hepatic encephalopathy occured or not, whether spontaneous bacterial peritonitis occured or not, whether hepatorenal syndrome occured or not, whether pulmonary infection occured or not, whether fungal infections occured or not, whether upper gastrointestinal bleeding occured or not, whether electrolyte imbalance occured or not, ALB (g/1), serum globulin GLB (g/1), A/G ratio, TBIL (mol/L), DBIL(mol/L), TBIL/DBIL ratio, ALT (U/L), AST (U/L), ALT/AST ratio, GGT (U/L), ALP (U/L), CHO (mmol/L), CHE (U/L), k+(mmol/L), Na+(mmol/L), CR(mol/L), PTA (%), INR, AFP(ng/ml), HBVDNA(IU/ml), HBeAg qualitative, WBC (*109/L), GR(%), Hb (g/1), PLT(*109/L). For reducing the deviations, the measurement data inclued TBIL, DBIL, ALT, AST, CHE, PTA, AFP, HBVDNA were transformed to natural logarithm. Calculated the difference of measurement data between the baseline and7th day’s excerpt the AFP, HBVDNA.0stand for death,1stand for cansored events,1stand for male,2stand for female,1stand for HBeAg positive,2stand for HBeAg negative. Followed-up all the patients till the patients dead or the study finished on January10,2011. The survival time was from hospitalized day to the day that the patients were dead, Lost of follow up, accepted the liver transplant operation and were diagnosed as malignant tumor. All the patients were devided to two groups according to the ratio of85%to15%:607patients on the model developed group and107patients on the model validity group. COX proportional hazards regression analyses identified the influencing factors on the patients of model developed group, developed the statistical model. According to the model, calculated the scores of our model and Model for end-stage liver disease(MELD), evaluated the validity of the two model through the Receiver operating characteristic curve(ROC).Results:(1) The age of830acute-on-chronic hepatitis B liver failure patients were (40.7±12.1) years,721patients were male(86.9%),109patients were female (13.1%), Hepatic encephalopathy occured on278acute-on-chronic hepatitis B liver failure patients(33.5%),244patients of them were male(29.4%),34patients of them were female(4.1%), there was no significant difference between the male and female patients(x2=0.298,P=0.585). Among all the acute-on-chronic hepatitis B liver failure patients, the age of hepatic encephalopathy patients were42(17) years, the age of non-hepatic encephalopathy patients were38.5(14) years, there was significant difference between the two kinds of patients(Z=-4.46, P=0.000). The results of univariate analysis is that there was significant difference between hepatic encephalopathy patients and non-hepatic encephalopathy patients on the index of age, TBIL/DBIL ratio, CHO, Na+, PTA, INR, AFP, WBC, GR, PLT. The results of logistic is that age, TBIL/DBIL ratio, PTA, INR, AFP, WBC, GR, HBVDNA are the influencing factors of hepatic encephalopathy on acute-on-chronic hepatitis B liver failure patients.(2) Spontaneous bacterial peritonitis occured on593acute-on-chronic hepatitis B liver failure patients(71.4%),244patients of them were male(61.3%),34patients of them were female(11.1%), there was no significant difference between the male and female patients(x2=1.942,P=0.163). Among all the acute-on-chronic hepatitis B liver failure patients, the age of spontaneous bacterial peritonitis patients were40(16) years, the age of non-spontaneous bacterial peritonitis patients were37(13) years, there was significant difference between the two kinds of patients (Z=-4.46, P=0.000). The results of univariate analysis is that there was significant difference between spontaneous bacterial peritonitis patients and non-spontaneous bacterial peritonitis patients on the index of age, ALB, A/G, TBIL, DBIL, ALT, AST, ALT/AST ratio, GGT, ALP, CHO, CHE, K+, PTA, INR, WBC, GR, Hb, PLT, HBVDNA. The results of logistic is that CHE, Hb are the influencing factors of spontaneous bacterial peritonitis on acute-on-chronic hepatitis B liver failure patients.(3) Hepatorenal syndrome occured on90acute-on-chronic hepatitis B liver failure patients(10.84%),76patients of them were male(9.15%),14patients of them were female(9.15%), there was no significant difference between the male and female patients(x2=0.52,P=0.471). Among all the acute-on-chronic hepatitis B liver failure patients, the age of hepatorenal syndrome patients were45(15) years, the age of non-hepatorenal syndrome patients were39(15) years, there was significant difference between the two kinds of patients(Z=-4.46,P=0.000). The results of univariate analysis is that there was significant difference between hepatorenal syndrome patients and non-hepatorenal syndrome patients on the index of age, ALB, TBIL, DBIL, ALT, ALT/AST ratio, CHO, CHE, Na+, CR, PTA, INR, AFP, WBC, GR, Hb, PLT. The results of logistic is that age, CR, WBC are the influencing factors of hepatorenal syndrome on acute-on-chronic hepatitis B liver failure patients.(4) Pulmonary infection occured on138acute-on-chronic hepatitis B liver failure patients(16.6%),120patients of them were male(14.5%),18patients of them were female(2.1%), there was no significant difference between the male and female patients(x2=0.001,P=0.973). Among all the acute-on-chronic hepatitis B liver failure patients, the age of pulmonary infection patients were45(16) years, the age of non-pulmonary infection patients were39(15) years, there was significant difference between the two kinds of patients(Z=-4.854, P=0.000). The results of univariate analysis is that there was significant difference between pulmonary infection patients and non-pulmonary infection patients on the index of age, ALB, A/G, TBIL, TBIL/DBIL ratio, ALT, AST, ALT/AST ratio, CHO, CHE, K+, Na+, PTA, AFP, WBC, GR, Hb, PLT. The results of logistic is that age, AFP, WBC, Hb are the influencing factors of pulmonary infection on acute-on-chronic hepatitis B liver failure patients.(5) Fungal infections occured on68acute-on-chronic hepatitis B liver failure patients(8.2%),58patients of them were male(7.0%),10patients of them were female(1.2%), there was no significant difference between the male and female patients(x2=0.161,P=0.688). Among all the acute-on-chronic hepatitis B liver failure patients, the age of fungal infections patients were46.5(18) years, the age of non-fungal infections patients were39(15) years, there was significant difference between the two kinds of patients(Z=-4.424, P=0.000). The results of univariate analysis is that there was significant difference between fungal infections patients and non-fungal infections patients on the index of age, ALB, A/G, WBC, Hb, PLT. The results of logistic is that age, WBC are the influencing factors of fungal infections on acute-on-chronic hepatitis B liver failure patients.(6) Upper gastrointestinal bleeding occured on83acute-on-chronic hepatitis B liver failure patients(10.0%),70patients of them were male(8.4%),13patients of them were female(1.62%), there was no significant difference between the male and female patients(x2=0.518,P=0.472). Among all the acute-on-chronic hepatitis B liver failure patients, the age of upper gastrointestinal bleeding patients were42(19) years, the age of non-upper gastrointestinal bleeding patients were39(15) years, there was significant difference between the two kinds of patients(Z=-2.317, P=0.021). The results of univariate analysis is that there was significant difference between upper gastrointestinal bleeding patients and non-upper gastrointestinal bleeding patients on the index of age, GGT, ALP, CHO, CHE, Na+, CR, PTA, INR, WBC, Hb, PLT. The results of logistic is that age, PTA, Hb are the influencing factors of upper gastrointestinal bleeding on acute-on-chronic hepatitis B liver failure patients.(7) Electrolyte imbalance occured on570acute-on-chronic hepatitis B liver failure patients(68.68%),490patients of them were male(59.04%),80patients of them were female(9.64%), there was no significant difference between the male and female patients(χ2=1.299, P=0.254). Among all the acute-on-chronic hepatitis B liver failure patients, the age of electrolyte imbalance patients were40(15) years, the age of non-electrolyte imbalance patients were37(15) years, there was significant difference between the two kinds of patients(Z=-3.785, P=0.000). The results of univariate analysis is that there was significant difference between electrolyte imbalance patients and non-electrolyte imbalance patients on the index of age, ALB, A/G ratio, TBIL, DBIL, ALT, ALT/AST ratio, CHO, CHE, K+, Na+, PTA, INR, AFP, WBC, GR, Hb. The results of logistic is that ALB, ALT/AST ratio, K+, Na+,AFP are the influencing factors of electrolyte imbalance on acute-on-chronic hepatitis B liver failure patients.(8) The age of714acute-on-chronic hepatitis B liver failure patients were39(15) years,622patients were male(87.1%),92patients were female(12.9%); The age of model developed patients were38(14) years,93patients were male(86.9%),14patients were female(13.1%). Till the follow-up finished, there were193patients died among714acute-on-chronic hepatitis B liver failure patients(27.0%), there were521patients lost of follow up, survived or accept the liver transplant operation(73.0%); there were162patients died among model developed patients(26.7%), there were445patients lost of follow up, survived or accept the liver transplant operation(73.3%); there were31patients died among714acute-on-chronic hepatitis B liver failure patients(29.0%), there were76patients lost of follow up, survived or accept the liver transplant operation(71%). The median follow-up time was334days, and the longest follow-up time was2558days. The results of the COX proportional hazards regression analyses is that ALB, ALT/AST ratio, PTA, AFP, HBVDNA, GR, ACR, ΔPTA, ΔGR, ΔPLT are the influencing factors of prognosis in acute-on-chronic hepatitis B liver failure patients. The personal prognosis index(PI):PI=0.052*ALB(g/l)-0.203*(ALT/AST)-1.189*LN(PTA%)-0.161*LN[AFP(ng/ml)]-0.06*LN[HBVDNA(IU/ml)]+0.036*(GR%)-0.003*ΔCR(μmol/l)+1.119*ΔLN (PTA%)-0.028*(ΔGR%)+0.005*Δ[PLT(*109/l)]The results of ROC is that the area under curve(AUC) of our model is79.1%, P=0.000, meanwhile the AUC of MELD is57.9%, P=0.199. The optimal cutoff of our model was-4.98, the sensitivity was80.6%, the specificity was72.4%.Conclusion:(1) The risk factors of hepatic encephalopathy in acute-on-chronic hepatitis B liver failure patients is higher age, higher TBIL/DBIL ratio, lower PTA, higher INR, lower AFP, higher WBC, higher GR, higher HBVDNA.(2) The risk factors of spontaneous bacterial peritonitis in acute-on-chronic hepatitis B liver failure patients is lower CHE, lower Hb.(3) The risk factors of hepatorenal syndrome in acute-on-chronic hepatitis B liver failure patients is higher age, higher Cr, higher WBC.(4) The risk factors of pulmonary infection in acute-on-chronic hepatitis B liver failure patients is higher age, lower AFP, higher WBC, lower Hb.(5) The risk factors of fungal infections in acute-on-chronic hepatitis B liver failure patients is higher age, higher WBC.(6) The risk factors of upper gastrointestinal bleeding in acute-on-chronic hepatitis B liver failure patients is higher age, lower PTA, lower Hb.(7) The risk factors of electrolyte imbalance in acute-on-chronic hepatitis B liver failure patients is lower ALB, lower ALT/AST ratio, lower K+, lower Na+, lower AFP.(8) The influencing factors of the prognosis in acute-on-chronic hepatitis B liver failure patients is ALB, ALT/AST ratio, PTA, AFP, HBVDNA, GR, ACR, APTA, AGR, ΔPLT. Our model was validated, if the PI of patients is more than-4.98, his prognosis will be bad.更多还原