基于多维度指标构建肝硬化病人食管胃静脉曲张破裂出血后门静脉血栓风险列线图模型及验证

杨悦军; 郑华银; 江逸锋

doi:10.13898/j.cnki.issn.1000-2200.2022.11.013

基于多维度指标构建肝硬化病人食管胃静脉曲张破裂出血后门静脉血栓风险列线图模型及验证

Construction and validation of a nomogram model for portal vein thrombosis risk after esophagogastric variceal rupture and bleeding in patients with cirrhosis based on multi-dimensional indicators

摘要

摘要:
目的基于多维度指标构建肝硬化病人食管胃静脉曲张破裂出血后门静脉血栓（PVT）风险的列线图模型，并进行验证。
方法选取肝硬化伴食管胃底静脉曲张破裂出血病人448例，按照3∶1的比例分为研究队列336例和检验队列112例，2队列再根据是否发生PVT分为血栓组和无血栓组，采用单因素分析和多因素回归分析筛选研究队列肝硬化病人食管胃静脉曲张破裂出血后发生PVT主要风险因素，采用RStudio软件中的rms程序包构建列线图并进行验证。
结果研究队列中77例（22.92%）发生PVT，其中出血后最快形成PVT者用时2周，49.35%（38/77）在1年内发生PVT，79.22%（61/77）在2年内合并PVT。检验队列中26例（23.21%）发生PVT，出血后最快形成PVT者用时12 d，50%（13/26）在1年内发生PVT，80.77%（21/26）在2年内合并PVT。单因素分析结果显示，研究队列PVT组与无PVT组肝功能Child-Pugh分级、门静脉内径、血小板计数、D-二聚体、P-选择素、三酰甘油差异均有统计学意义（P < 0.05~P < 0.01）。多因素logistic回归分析结果显示，肝功能Child-Pugh分级、门静脉内径、血小板计数、D-二聚体、P-选择素、三酰甘油均为肝硬化病人食管胃静脉曲张破裂出血后PVT的风险因素（P < 0.01）。ROC曲线分析结果显示，风险列线图预测模型在研究队列的曲线下面积（AUC）为0.953（95%CI：0.931~0.976），灵敏度为81.8%，特异度为94.2%；风险列线图预测模型在检验队列的AUC为0.944（95%CI：0.904~0.984），灵敏度为98.9%，特异度为81.4%。
结论肝功能Child-Pugh分级、门静脉内径、血小板计数、D-二聚体、P-选择素、三酰甘油均为肝硬化病人食管胃静脉曲张破裂出血PVT的风险因素。基于上述指标建立的列线图预测模型具有良好的诊断效能，可为临床制定针对性治疗方案提供参考。

Abstract:
ObjectiveTo construct and validate a nomograph model for portal vein thrombosis(PVT) risk after esophagogastric variceal rupture and bleeding in patients with cirrhosis based on multi-dimensional indicators.
MethodsA total of 448 patients with liver cirrhosis complicated with esophagogastric variceal rupture and bleeding were selected and divided into the study cohort(n=336) and test cohort(n=112) at the ratio of 3:1.The two cohorts were subdivided into the thrombus group and non-thrombus group according to whether PVT occurrence.The main risk factors of PVT in the study cohort after esophagogastric variceal rupture and bleeding in patients with cirrhosis were screened by univariate analysis and multivariate regression analysis, and a nomograph model was constructed using rms package in RStudio software and verified.
ResultsIn the study cohort, 77 cases(22.92%) developed PVT, in which the fastest PVT formation after bleeding took two weeks, 49.35%(38/77) developed PVT within one year, and 79.22%(61/77) combined with PVT within two years.In the test cohort, 26 cases(23.21%) developed PVT, in which the fastest PVT formation after bleeding took 12 days, 50%(13/26) developed PVT within 1 year, and 80.77%(21/26)combined PVT within 2 years.Univariate analysis showed that there were significant differences in Child-Pugh liver function grade, portal vein diameter, platelet count, D-dimer, P-selectin and triacylglycerol between PVT group and non-PVT group(P < 0.05 to P < 0.01).Multivariate logistic regression analysis showed that Child-Pugh liver function grade, portal vein diameter, platelet count, D-dimer, P-selectin and triacylglycerol were the risk factors for PVT after esophagogastric variceal rupture and bleeding in patients with cirrhosis(P < 0.01).ROC curve analysis showed that the area under the curve(AUC) of the risk nomogram prediction model in the study cohort was 0.953(95%CI: 0.931-0.976), the sensitivity was 81.8%, and the specificity was 94.2%.ROC curve analysis showed that the AUC of the risk nomogram prediction model in the test cohort was 0.944(95%CI: 0.904-0.984), the sensitivity was 98.9%, and the specificity was 81.4%.
ConclusionsChild-Pugh liver function grade, portal vein diameter, platelet count, D-dimer, P-selectin and triacylglycerol are the risk factors of PVT in patients with liver cirrhosis complicated with esophagogastric variceal rupture and bleeding.The nomogram prediction model based on the above indicators has good diagnostic performance, and can provide a reference for development of targeted treatment plans in clinic.

HTML全文

参考文献(24)

施引文献

资源附件(0)