声辐射力脉冲成像和无创预测模型对肝硬化患者并发食道胃底静脉曲张的预测价值

Value of acoustic radiation force impulse elastography and noninvasive prediction model in the prediction of esophagealgastro varices in patients with cirrhosis

目的探讨声辐射力脉冲成像(ARFI)及无创性预测模型对肝硬化患者并发食道胃底静脉曲张严重程度的预测价值。方法回顾性分析392例肝硬化住院患者,常规行实验室参数、胃镜及B超检测,同时B超引导下行ARFI值测定,并根据相关实验室参数计算出不同无创性预测模型值。单因素和多因素分析不同程度食道胃底静脉曲张肝硬化患者的实验室参数,根据ROC曲线确定各个预测方法的截断值,计算其灵敏度、特异度、阳性预测值、阴性预测值,并通过单独和联合比较各个预测方法的准确度以评价最佳预测方案。结果多因素回归分析显示低风险组和高风险组食道胃底静脉曲张的肝硬化患者在脾长径、血小板计数、ARFI值及无创性预测模型方面差异有统计学意义(P<0.01);ROC曲线分析显示:ARFI值和无创性预测模型对肝硬化患者并发食道胃底静脉曲张严重程度的预测均有较高的灵敏度和特异度。ARFI与其他无创性预测模型联合使用,更有助于提高预测肝硬化患者食道胃底静脉曲张严重程度的敏感度。结论 ARFI值可有效预测肝硬化患者并发食道胃底静脉曲张严重程度,当ARFI值>3.18 m/s时考虑肝硬化患者存在高风险的食道胃底静脉曲张,应对患者采取相应措施进行干预和治疗。

Objective To evaluate the prediction value of acoustic radiation force impulse（ ARFI） elastography and noninvasive prediction model for esophagealgastro varices（ HEGVs） in patients with cirrhosis. Methods To retrospectively study 392 inpatients with cirrhosis for routine laboratory parameters,gastroscopy and detection of B ultrasound; meanwhile,ARFI determination was performed with B ultrasound＇s guidance,and according to the related laboratory parameters,different noninvasive prediction models were calculated. Single factor and multi factor analysis were carried out on the laboratory parameters of different degrees of esophageal and gastric varices in patients with cirrhosis. According to ROC curve to determine the cut-off value of each prediction method,and to calculate the sensitivity,specificity,positive predictive value,negative predictive value. And through the individual test and combined tests to compare the accuracy of each prediction method,so as to evaluate the best prediction scheme. Results Multivariate analysis showed that spleen size,platelet count,ARFI,noninvasive prediction model in patients with HEGVs between in low risk group and in high risk group had differences and the differences were statistically significant（ P 0. 01）. ROC curve analysis showed that ARFI value and non invasive prediction model had higher sensitivity and specificity in predicting the severity of esophageal gastric varices in patients with cirrhosis. A combination of ARFI with these noninvasive prediction models achieved a higher sensitivity for the prediction of HEGVs. Conclusion ARFI can accurately identify the severity of HEGVs in patients with cirrhosis. And the patients with ARFI 3. 18 m / s should be considered for endoscopic examinations or appropriate prophylactic treatment.