摘要目的:评估终末期肝病模型联合血清钠评分（MELD-Na）、慢性肝衰竭-序贯器官衰竭评估模型简化评分（CLIF-C OFs）、中国重型乙型肝炎（乙肝）研究组乙肝相关的慢加急性肝衰竭预后评分（COSSH-ACLFs）、中性粒细胞/淋巴细胞比值（NLR）评分系统在乙型肝炎病毒相关慢加急性肝衰竭（HBV-ACLF）患者中的应用价值。方法:回顾性收集2010年7月至2018年7月天津市第二人民医院收治的163例HBV-ACLF患者的临床资料（性别、年龄、疾病分期）以及实验室检查指标〔丙氨酸转氨酶（ALT）、天冬氨酸转氨酶（AST）、总胆红素（TBil）、白蛋白（ALB）、血尿素氮（BUN）、肌酐（Cr）、血钠（Na）、凝血酶原活动度（PTA）、国际标准化比值（INR）、中性粒细胞计数（NEU）、淋巴细胞计数（LYM）〕，以入院8周为观察时间节点，根据预后不同将患者分为存活组（90例）和死亡组（73例）。比较死亡组和存活组的MELD-Na、CLIF-C OFs、COSSH-ACLFs评分及NLR差异，并行Logistic回归分析，确定HBV-ACLF的独立危险因素。行倾向匹配分析，验证独立危险因素的准确性，绘制受试者工作特征曲线（ROC），判断独立危险因素的临床价值。结果:存活组和死亡组患者入院时性别、疾病分期、ALB、BUN、Cr、Na、NEU水平比较差异均无统计学意义（均 P>0.05）；存活组年龄、ALT、AST、TBil、INR水平均明显低于死亡组〔年龄（岁）：43.00（34.00，53.00）比50.00（42.50，55.00），ALT（U/L）：252.90（61.43，613.33）比359.10（115.15，784.70），AST（U/L）：146.15（90.88，449.30）比237.80（109.00，635.05），TBil（μmol/L）：265.10（183.10，347.60）比307.50（229.90，405.55），INR：2.13（1.91，2.46）比2.29（2.02，2.94）〕，PTA和LYM水平均明显高于死亡组〔PTA（%）：34.00（28.00，38.00）比31.00（24.00，36.00），LYM（×10 9/L）：1.37（0.72，1.79）比0.85（0.51，1.39），均 P<0.05〕；与死亡组比较，存活组MELD-Na、CLIF-C OFs、COSSH-ACLFs评分及NLR均更低〔MELD-Na评分（分）：17.99（16.60，19.63）比19.16（17.43，20.80），CLIF-C OFs评分（分）：9.00（8.00，9.00）比9.00（9.00，10.00），COSSH-ACLFs评分（分）：4.87（4.63，5.48）比5.47（5.07，5.80），NLR：2.86（2.21，5.19）比4.38（2.54，8.46），均 P<0.05〕。Logistic回归分析显示，CLIF-C OFs评分〔优势比（ OR）＝0.532，95%可信区间（95% CI）为0.380～0.744， P<0.05〕、NLR（ OR＝0.901，95% CI为0.835～0.972， P<0.05）是判断HBV-ACLF患者8周预后的独立危险因素。倾向性匹配后，59例存活者和59例死亡者数据匹配成功，两组年龄、性别、疾病分期、ALT、AST、TBil、ALB、BUN、Cr、Na、PTA、INR、NEU比较差异均无统计学意义（均 P>0.05），入院基线资料中仅LYM间比较差异具有统计学意义〔×10 9/L：1.35（0.74，1.73）比0.81（0.51，1.30）， P<0.05〕；存活组的CLIF-C OFs、COSSH-ACLFs评分及NLR均明显低于死亡组〔CLIF-C OFs评分（分）：9.00（8.00，9.00）比9.00（8.00，10.00），COSSH-ACLFs评分（分）：4.99（4.69，5.64）比5.34（5.03，5.81），NLR：2.85（2.21，5.72）比4.38（2.47，10.20），均 P<0.05〕；CLIF-C OFs评分（ OR＝0.593，95% CI为0.401～0.878， P<0.05）、NLR（ OR＝0.914，95% CI为0.842～0.993， P<0.05）仍是判断HBV-ACLF患者预后的独立危险因素。CLIF-C OFs评分≥9分预测HBV-ACLF患者8周预后的敏感度为76.7%，特异度为48.9%，AUC为0.662；NLR≥3.14预测HBV-ACLF患者8周预后的敏感度为67.1%，特异度为56.7%，AUC为0.623。CLIF-C OFs评分、NLR串联试验诊断的方式可提高预测HBV-ACLF患者8周预后的特异度至77.8%。 结论:CLIF-C OFs评分、NLR是影响HBV-ACLF临床结局的独立危险因素，对判断HBV-ACLF患者的预后有更高的临床价值，两者联合应用将更有利于判断HBV-ACLF患者的预后。

abstractsObjective:To verify the accuracy of the model for end-stage liver disease-sodium (MELD-Na), chronic liver failure consortium organ failure score (CLIF-C OFs), Chinese Group on the Study of Severe Hepatitis B-Acute-on-chronic Liver Failure score (COSSH-ACLFs) and neutrophil-to-lymphocyte ratio (NLR) scoring systems in patients with hepatitis B virus related acute-on-chronic liver failure (HBV-ACLF) and to explore its value in clinical application.Methods:The clinical data (gender, age, disease stage) and laboratory indicators [alanine transferase (ALT), glutamate transaminase (AST), total bilirubin (TBil), albumin (ALB), blood urea nitrogen (BUN), creatinine (Cr), serum sodium (Na), prothrombin activity (PTA), international standardized ratio (INR), neutrophils count (NEU) and lymphocytes count (LYM)] of 163 patients with HBV-ACLF from July 2010 to July 2018 in Tianjin Second People's Hospital were retrospectively analyzed. After 8 weeks of admission, the patients were divided into death group (90 cases) and survival group (73 cases) according to survival status. The MELD-Na, CLIF-C OFs, COSSH-ACLFs scores and NLR of death group and survival group were compared, and a multivariate Logistic regression analysis was used to analyze the independent risk factors for HBV-ACLF. Propensity score analysis was used to demonstrate the accuracy of the method and receiver operating characteristic curve (ROC) was used to analyze the diagnostic value of the independent risk factors.Results:There were no significant differences in gender, disease stage, ALB, BUN, Cr, Na, NEU on admission between the two groups (all P > 0.05). The age [years old: 43.00 (34.00, 53.00) vs. 50.00 (42.50, 55.00)] and serum levels of ALT [U/L: 252.90 (61.43, 613.33) vs. 359.10 (115.15, 784.70)], AST [U/L: 146.15 (90.88, 449.30) vs. 237.80 (109.00, 635.05)], TBil [μmol/L: 265.10 (183.10, 347.60) vs. 307.50 (229.90, 405.55)] and INR [2.13 (1.91, 2.46) vs. 2.29 (2.02, 2.94)] in survival group were lower than those in death group and the PTA [%: 34.00 (28.00, 38.00) vs. 31.00 (24.00, 36.00)] and LYM [×10 9/L: 1.37 (0.72, 1.79) vs. 0.85 (0.51, 1.39)] levels were significantly higher than those in death group (both P < 0.05). The MELD-Na [17.99 (16.60, 19.63) vs. 19.16 (17.43, 20.80)], CLIF-C OFs [9.00 (8.00, 9.00) vs. 9.00 (9.00, 10.00)], COSSH-ACLFs [4.87 (4.63, 5.48) vs. 5.47 (5.07, 5.80)] and NLR [2.86 (2.21, 5.19) vs. 4.38 (2.54, 8.46)] were lower in survival group than those of the death group (all P < 0.05). Logistic regression analysis showed that CLIF-C OFs [odds ratio ( OR) = 0.532, 95% confidence interval (95% CI) was 0.380-0.744, P < 0.05] and NLR ( OR = 0.901, 95% CI was 0.835-0.972, P < 0.05) were the independent risk factors for the prognosis of HBV-ACLF. After propensity score matching, the data of 59 cases in each group were successfully matched, there were no significant differences in age, gender, disease stage, ALT, AST, TBil, ALB, BUN, Cr, Na, PTA, INR and NEU between the two groups (all P > 0.05), and statistically significant difference in the baseline LYM [×10 9/L: 1.35 (0.74, 1.73) vs. 0.81 (0.51, 1.30)] were found between the survival group and the death group. The CLIF-C OFs, COSSH-ACLFs scores and NLR were lower in survival group compared with those of the death group [CLIF-C OFs: 9.00 (8.00, 9.00) vs. 9.00 (8.00, 10.00), COSSH-ACLFs: 4.99 (4.69, 5.64) vs. 5.34 (5.03, 5.81), NLR: 2.85 (2.21, 5.72) vs. 4.38 (2.47, 10.20), all P < 0.05] and CLIF-C OFs ( OR = 0.593, 95% CI was 0.401-0.878, P < 0.05) and NLR ( OR = 0.593, 95% CI was 0.401-0.878, P < 0.05) were still as the independent risk factors for the prognosis of HBV-ACLF. The sensitivity of CLIF-C OFs≥9 and NLR≥3.14 to forecast the 8-week clinical outcome of HBV-ACLF patients were 76.7% and 67.1%, the specificity were 48.9% and 56.7%, and AUC were 0.662 and 0.623. CLIF-C OFs was combined with NLR to increase the specificity of forecasting the 8-week clinical outcome of HBV-ACLF patients to 77.8%. Conclusions:CLIF-C OFs and NLR scores are independent risk factors affecting the clinical outcome of HBV-ACLF, and have better clinical value in predicting the prognosis of HBV-ACLF. Combined application of the two scores will be more beneficial to the prognosis of HBV-ACLF.