• 中国科学论文统计源期刊
  • 中国科技核心期刊
  • 美国化学文摘(CA)来源期刊
  • 日本科学技术振兴机构数据库(JST)

临床输血与检验 ›› 2024, Vol. 26 ›› Issue (3): 345-352.DOI: 10.3969/j.issn.1671-2587.2024.03.008

• 临床输血 • 上一篇    下一篇

慢加急性肝衰竭血浆置换治疗期间死亡的危险因素探讨:基于PSM法

李佩华1, 苏焕正2, 沙夏珺3, 罗文英1   

  1. 1广东医科大学附属医院检验医学中心,湛江 524000;
    2江门市人民医院重症医学科,江门 529000;
    3江门市人民医院输血科,江门 529000
  • 收稿日期:2023-12-21 出版日期:2024-06-20 发布日期:2024-06-24
  • 通讯作者: 罗文英,主要从事病原微生物致病机制方面研究,(E-mail)luo_wenying@126.com。
  • 作者简介:李佩华,主要从事输血治疗方面研究,(E-mail)tobyhscg@163.com。

Risk Factors for Death during Plasma Exchange Therapy for Subacute and Acute Liver Failure: Based on PSM Method

LI Peihua1, SU Huanzheng2, SHA Xiajun3, LUO Wenying1   

  1. 1Laboratory Medicine Center,Affiliated Hospital of Guangdong Medical University, Zhanjiang 524000;
    2Intensive Care Medicine Department of Jiangmen People's Hospital, Jiangmen 529000;
    3Blood Transfusion Department of Jiangmen People's Hospital, Jiangmen 529000
  • Received:2023-12-21 Online:2024-06-20 Published:2024-06-24

摘要: 目的 基于倾向性评分匹配法(PSM)探讨慢加急性肝衰竭(ACLF)血浆置换治疗期间死亡的危险因素。方法 回顾性分析2019年1月—2023年1月本院收治的196例ACLF患者的临床资料,根据血浆置换治疗期间死亡情况分为死亡组(n=68)、存活组(n=128)。采用1∶1 PSM法匹配两组一般人口学变量。基于匹配后数据采用Cox回归模型分析ACLF血浆置换治疗期间死亡的影响因素。绘制Kaplan-Meier曲线分析影响因素与ACLF血浆置换治疗期间死亡的关系。结果 ACLF血浆置换治疗期间死亡率为34.69%。匹配前两组性别、体质量指数(BMI)、吸烟史、高血压、高脂血症、糖尿病、慢性肾病比较,差异无统计学意义(P>0.05),年龄、饮酒史比较,差异有统计学意义(P<0.05)。PSM后,共成功匹配64对,两组间所有匹配因素比较,差异均无统计学意义(P>0.05),两组间匹配因素分布均衡性好,匹配效果良好。匹配后,死亡组合并肝肾综合征构成比,终末期肝病模型(MELD)评分系统联合血清钠(MELD-Na)评分、红细胞体积分布宽度(RDW)、总胆红素(TBil)、血肌酐(Scr)、国际标准化比值(INR)高于存活组(P<0.05),血钠、凝血酶原活动度(PTA)水平及治疗依从构成比低于存活组(P<0.05)。Cox回归模型经校正混杂因素后显示,MELD-Na评分(HR=2.889,95% CI:1.431~5.836)、合并肝肾综合征(HR=3.048,95% CI:1.332~6.975)、RDW(HR=2.166,95% CI:1.340~3.502)是ACLF血浆置换治疗期间死亡的影响因素(P<0.05)。Log-rank检验显示,高MELD-Na评分、合并肝肾综合征、高RDW患者的死亡率均高于低MELD-Na评分、无肝肾综合征、低RDW患者(P<0.05)。结论 基于PSM法发现MELD-Na评分、合并肝肾综合征、RDW是ACLF血浆置换治疗期间死亡的影响因素。

关键词: 慢加急性肝衰竭, 血浆置换, 死亡, 危险因素, 倾向性评分匹配法

Abstract: Objective To explore the risk factors for death during plasma exchange therapy for subacute and acute liver failure (ACLF) based on propensity score matching method (PSM). Methods The clinical data of 196 patients with ACLF admitted to our hospital from January 2019 to January 2023 were analyzed. According to the death situation during plasma exchange therapy, they were divided into death group (n=68) and survival group (n=128). A 1:1 PSM method was used to match general demographic variables between the two groups; after matching, Cox regression model was applied to analyze the factors influencing mortality during ACLF plasma exchange treatment. Kaplan-Meier curve was drawn to analyze the relationship between influencing factors and death during ACLF plasma exchange. Results The mortality rate during ACLF plasma exchange treatment was 34.69%. Before matching, there were no statistically significant differences between the two groups in terms of gender, BMI, smoking history, hypertension, hyperlipidemia, diabetes, and chronic kidney disease (P>0.05). There were significant differences in age and drinking history (P<0.05). After PSM, 64 pairs were successfully matched, and there were no significant differences in all matched factors between the two groups (P>0.05), indicating good balance and effective matching. After matching, the ratio of the death group with hepatorenal syndrome, the model of end-stage liver disease (MELD) score system combined with serum sodium (MELD-NA) score, red blood cell volume distribution width (RDW), total bilirubin (TBil), blood creatinine (Scr) and international standardized ratio (INR) were higher than those of the survival group (P<0.05). The levels of serum sodium and prothrombin activity (PTA) and the proportion of treatment compliance were lower than those of survival group (P<0.05). Cox regression analysis revealed that MELD-Na scores (HR=2.450, 95% CI:1.483~4.046), concurrent hepatorenal syndrome (HR=2.223, 95% CI:1.496~3.303), RDW (HR=2.912, 95% CI:1.416~5.991), PTA (HR=0.373, 95% CI:0.225~0.620), and treatment compliance (HR=0.284, 95% CI:0.141~0.571) were significant factors affecting mortality during ACLF plasma exchange treatment (P<0.05). Cox regression model after adjusting for confounders showed that MELD-Na score (HR=2.889, 95% CI:1.431~5.836), hepatorenal syndrome (HR=3.048, 95% CI:1.332~6.975), RDW (HR=2.166, 95% CI:1.340~3.502) were the influential factors of death during ACLF plasma exchange (P<0.05). Log-rank test showed that the mortality of patients with high MELD-Na score, hepatorenal syndrome and high RDW were higher than those of patients with low MELD-Na score, no hepatorenal syndrome and low RDW (P<0.05). Conclusion Based on PSM, MELD-Na score, hepatorenal syndrome, RDW are found to be influential factors for death during ACLF plasma exchange treatment.

Key words: Subacute and acute liver failure, Plasma exchange, Death Risk factors, Propensity score matching method

中图分类号: