Zhou Yu-Jie, Zheng Ji-Na, Zhou Yi-Fan, Han Yi-Jing, Zou Tian-Tian, Liu Wen-Yue, Braddock Martin, Shi Ke-Qing, Wang Xiao-Dong, Zheng Ming-Hua
Departments of aHepatology, Liver Research Center bEndocrinology, the First Affiliated Hospital of Wenzhou Medical University cSchool of the First Clinical Medical Sciences dSchool of the Second Clinical Medical Sciences, Wenzhou Medical University, Wenzhou, China eInstitute of Hepatology, Wenzhou Medical University, Wenzhou, China fGlobal Medicines Development, AstraZeneca R&D, Loughborough, UK.
Eur J Gastroenterol Hepatol. 2017 Oct;29(10):1166-1173. doi: 10.1097/MEG.0000000000000943.
Upper gastrointestinal bleeding (UGIB) is a complication with a high mortality rate in critically ill patients presenting with cirrhosis. Today, there exist few accurate scoring models specifically designed for mortality risk assessment in critically ill cirrhotic patients with upper gastrointestinal bleeding (CICGIB). Our aim was to develop and evaluate a novel nomogram-based model specific for CICGIB.
Overall, 540 consecutive CICGIB patients were enrolled. On the basis of Cox regression analyses, the nomogram was constructed to estimate the probability of 30-day, 90-day, 270-day, and 1-year survival. An upper gastrointestinal bleeding-chronic liver failure-sequential organ failure assessment (UGIB-CLIF-SOFA) score was derived from the nomogram. Performance assessment and internal validation of the model were performed using Harrell's concordance index (C-index), calibration plot, and bootstrap sample procedures. UGIB-CLIF-SOFA was also compared with other prognostic models, such as CLIF-SOFA and model for end-stage liver disease, using C-indices.
Eight independent factors derived from Cox analysis (including bilirubin, creatinine, international normalized ratio, sodium, albumin, mean artery pressure, vasopressin used, and hematocrit decrease>10%) were assembled into the nomogram and the UGIB-CLIF-SOFA score. The calibration plots showed optimal agreement between nomogram prediction and actual observation. The C-index of the nomogram using bootstrap (0.729; 95% confidence interval: 0.689-0.766) was higher than that of the other models for predicting survival of CICGIB.
We have developed and internally validated a novel nomogram and an easy-to-use scoring system that accurately predicts the mortality probability of CICGIB on the basis of eight easy-to-obtain parameters. External validation is now warranted in future clinical studies.
上消化道出血（UGIB）是肝硬化危重症患者的一种并发症，死亡率很高。目前，专门用于评估肝硬化危重症患者上消化道出血（CICGIB）死亡风险的准确评分模型很少。我们的目的是开发并评估一种针对CICGIB的新型列线图模型。
总共纳入了540例连续性CICGIB患者。基于Cox回归分析构建列线图，以估计30天、90天、270天和1年生存率的概率。从列线图中得出上消化道出血-慢性肝衰竭-序贯器官衰竭评估（UGIB-CLIF-SOFA）评分。使用Harrell一致性指数（C指数）、校准图和自助抽样程序对模型进行性能评估和内部验证。还使用C指数将UGIB-CLIF-SOFA与其他预后模型（如CLIF-SOFA和终末期肝病模型）进行比较。
将Cox分析得出的8个独立因素（包括胆红素、肌酐、国际标准化比值、钠、白蛋白、平均动脉压、使用的血管加压素和血细胞比容下降>10%）纳入列线图和UGIB-CLIF-SOFA评分。校准图显示列线图预测与实际观察之间具有最佳一致性。使用自助抽样的列线图C指数（0.729；95%置信区间：0.689-0.766）高于其他预测CICGIB生存率的模型。
我们开发并内部验证了一种新型列线图和一个易于使用的评分系统，该系统基于8个易于获取的参数准确预测CICGIB的死亡概率。未来的临床研究有必要进行外部验证。
