论文部分内容阅读
目的:探讨终末期肝病模型(model for end-stage liver disease, MELD)联合高密度脂蛋白胆固醇(high density lipoprotein-cholesterol，HDL-C)评估乙型肝炎病毒相关慢加急性肝衰竭(hepatitis B virus-related acute-on-chronic liver failure, HBV-ACLF)短期预后的临床价值。方法:纳入2015年12月至2018年12月在河南省人民医院住院的182例HBV-ACLF患者。回顾性分析患者入院24 h内HDL-C、总胆红素、国际标准化比值、肌酐等临床资料及预后情况，并进行MELD评分。应用二元logistic回归分析影响HBV-ACLF患者90 d预后的独立危险因素，通过受试者操作特征曲线和MedCalc 15.2软件分析MELD、HDL-C和MELD联合HDL-C三种评分模型对短期预后的评估价值，应用Kaplan-Meier法绘制生存曲线。结果:182例患者住院和出院后90 d随访期间，死亡60例，存活122例。生存组MELD评分为21(19, 24)分，低于死亡组的29(25, 34)分，而生存组HDL-C水平高于死亡组[0.3 (0.1, 0.6) mmol/L比0.2(0.1, 0.5) mmol/L]，差异均有统计学意义(n Z＝-6.290、-4.087，均n P<0.01)。多因素logistic回归分析显示，MELD评分、HDL-C水平是影响HBV-ACLF患者短期死亡的独立相关因素[比值比(odds ratio，n OR)＝1.432，95%可信区间(confidence interval, n CI)1.271～1.613；n OR＝0.584，95%n CI 0.487～0.700；均n P<0.01]。MELD、HDL-C、MELD联合HDL-C三种评分模型的受试者操作特征曲线下面积分别为0.775、0.782、0.878。MELD联合HDL-C评分分别优于MELD、HDL-C评分，差异均有统计学意义(n Z＝3.944、3.104，均n P<0.01)。MELD联合HDL-C评分模型的约登指数为0.72时，其最佳临界值为24.69。MELD联合HDL-C评分≥24.69分患者的生存率较MELD联合HDL-C评分<24.69分患者的生存率低，差异有统计学意义(n χ2＝142.900，n P<0.01)。n 结论:MELD、HDL-C和MELD联合HDL-C评分模型均能较好地评估HBV-ACLF患者的短期预后，其中MELD联合HDL-C评分模型的临床评估价值更高。“,”Objective:To evaluate the short-term prognostic value of model for end-stage liver disease (MELD) combined with high density lipoprotein-cholesterol (HDL-C) in patients with hepatitis B virus-related acute-on-chronic liver failure (HBV-ACLF).Methods:From December 2015 to December 2018, 182 patients with HBV-ACLF who were treated in Henan Provincial People′s Hospital were included. Prognosis and clinical data including HDL-C, total bilirubin, international standardized ratio (INR), creatinine of patients within 24 hours after admission were collected and analyzed retrospectively.The values of MELD were calculated. The binary logistic regression analysis was used to analyze the independent risk factors affecting 90-day mortality in HBV-ACLF patients.The receiver operator characteristic curve (ROC) and MedCalc 15.2 software were used to assess the predictive value of MELD, HDL-C and MELD-HDL-C model for prognosis. Kaplan-Meier survival curve was performed to analyze the prognosis of patients in different groups.Results:Sixty patients were divided into the death group and 122 patients were divided into the survival group according to the prognosis during hospitalization and 90 days after discharge. The MELD score of patients in the survival group was 21(19, 24), which was significantly lower than that in the death group (29(25, 34)), and the HDL-C value of patients in the survival group was significantly higher than that in the death group (0.3 (0.1, 0.6) mmol/L n vs 0.2(0.1, 0.5) mmol/L). The differences were both statistically significant (n Z=-6.290 and -4.087, respectively, both n P<0.01). Multivariate logistic regression analysis showed that MELD score and HDL-C value were the independent risk factors for 90-day mortality in patients with HBV-ACLF(odds ratio (n OR)=1.432, 95% confidence interval (n CI)1.271-1.613; n OR=0.584, 95%n CI 0.487-0.700, respectively; both n P<0.01). Areas under the ROC of MELD, HDL-C and MELD-HDL-C scoring models were 0.775, 0.782 and 0.878, respectively. MELD-HDL-C scoring model was superior to both MELD and HDL-C , and the differences were both statistically significant (n Z=3.944 and 3.104, respectively, both n P<0.01). When the MELD-HDL-C Youden′s index was set at 0.72, the optimal threshold was 24.69. Patients with MELD-HDL-C score≥24.69 had lower survival rate than patients with MELD-HDL-C score<24.69, and the difference was statistically significant (n χ2=142.900, n P<0.01).n Conclusion:MELD, HDL-C and MELD-HDL-C scoring systems could predict the short-term prognosis in patients with HBV-ACLF, and the predictive value of MELD-HDL-C has the superiority.