Scholarlink

Background <p>Acute upper gastrointestinal bleeding (AUGIB) is a life-threatening condition with substantial mortality, and early, accurate risk stratification is essential to improve outcomes. The endothelial activation and stress index (EASIX) is a novel composite marker reflecting endothelial injury and systemic stress; however, its prognostic value in AUGIB remains uncertain. This study aimed to evaluate the predictive value of EASIX for in-hospital adverse outcomes in patients with AUGIB and to develop a prediction model integrating clinical variables.</p> Methods <p>In this single-center retrospective cohort study, we consecutively enrolled 483 patients hospitalized with AUGIB between January and December 2023. Clinical data were extracted from electronic medical records, and log₂(EASIX) was calculated at admission. Candidate predictors were selected using least absolute shrinkage and selection operator (LASSO) regression to reduce overfitting, and Firth penalized logistic regression models were developed to predict in-hospital mortality and ICU admission in view of the relatively low event rate. Model performance was evaluated using receiver operating characteristic (ROC) analysis, calibration plots, and decision curve analysis, with internal validation by bootstrap resampling (B = 1000). Subgroup and sensitivity analyses were conducted to assess robustness.</p> Results <p>In-hospital mortality occurred in 55 patients (11.39%) and ICU admission in 34 (7.04%). Log<sub>2</sub>(EASIX) levels were significantly higher in non-survivors compared with survivors (<i>P</i> &lt; 0.001) and remained independently associated with in-hospital mortality after adjustment (OR = 1.39, 95% CI 1.10–1.78). The LASSO-refit model demonstrated excellent discrimination for mortality (AUC = 0.924) and ICU admission (AUC = 0.894), outperforming conventional risk scores. After multivariable adjustment, the association between log<sub>2</sub>(EASIX) and ICU admission was attenuated and became statistically non-significant. Findings were generally consistent across subgroup and sensitivity analyses.</p> Conclusions <p>Admission EASIX independently predicts in-hospital mortality in AUGIB, and an EASIX-integrated model provides excellent discrimination and clinical utility. Although EASIX shows some ability to identify ICU admission risk, its predictive effect appears more sensitive to multivariable adjustment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Endothelial activation and stress index predicts adverse outcomes in acute upper gastrointestinal bleeding: a retrospective study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月17日

查看全文
  • Chenao Zhang,
  • Junyan Wang,
  • Yu Wang,
  • Jianming Xu,
  • Qiao Mei

摘要

Background

Acute upper gastrointestinal bleeding (AUGIB) is a life-threatening condition with substantial mortality, and early, accurate risk stratification is essential to improve outcomes. The endothelial activation and stress index (EASIX) is a novel composite marker reflecting endothelial injury and systemic stress; however, its prognostic value in AUGIB remains uncertain. This study aimed to evaluate the predictive value of EASIX for in-hospital adverse outcomes in patients with AUGIB and to develop a prediction model integrating clinical variables.

Methods

In this single-center retrospective cohort study, we consecutively enrolled 483 patients hospitalized with AUGIB between January and December 2023. Clinical data were extracted from electronic medical records, and log₂(EASIX) was calculated at admission. Candidate predictors were selected using least absolute shrinkage and selection operator (LASSO) regression to reduce overfitting, and Firth penalized logistic regression models were developed to predict in-hospital mortality and ICU admission in view of the relatively low event rate. Model performance was evaluated using receiver operating characteristic (ROC) analysis, calibration plots, and decision curve analysis, with internal validation by bootstrap resampling (B = 1000). Subgroup and sensitivity analyses were conducted to assess robustness.

Results

In-hospital mortality occurred in 55 patients (11.39%) and ICU admission in 34 (7.04%). Log2(EASIX) levels were significantly higher in non-survivors compared with survivors (P < 0.001) and remained independently associated with in-hospital mortality after adjustment (OR = 1.39, 95% CI 1.10–1.78). The LASSO-refit model demonstrated excellent discrimination for mortality (AUC = 0.924) and ICU admission (AUC = 0.894), outperforming conventional risk scores. After multivariable adjustment, the association between log2(EASIX) and ICU admission was attenuated and became statistically non-significant. Findings were generally consistent across subgroup and sensitivity analyses.

Conclusions

Admission EASIX independently predicts in-hospital mortality in AUGIB, and an EASIX-integrated model provides excellent discrimination and clinical utility. Although EASIX shows some ability to identify ICU admission risk, its predictive effect appears more sensitive to multivariable adjustment.