Background & Aims
Esophageal variceal bleeding is a life-threatening condition with a reported mortality rate of 20% even after hemostasis. This makes medical resource allocation and explaining the condition to patients and their families challenging after initial treatment. We aimed to create and validate a concise scoring system, the Hospital Outcome Prediction following Endoscopic Variceal Ligation (HOPE-EVL) score, designed to predict in-hospital mortality after EVL hemostasis for esophageal variceal bleeding.
Methods
Data from a 13-year study involving 46 Japanese institutions were split into development (initial 7 years) and validation (last 6 years) cohorts. We included cases that required hospitalization for esophageal variceal bleeding and underwent successful hemostasis with EVL on emergency admission. The outcome measured was in-hospital mortality. Our study collected potential predictive factors, including demographic and clinical characteristics. We also gathered comprehensive data at the time of visit, encompassing medical histories, routine medication usage, vital signs, and laboratory data. Variable selection from the development cohort used least absolute shrinkage and selection operator (LASSO) regression with in-hospital mortality as the response variable. We developed a simple scoring system, the HOPE-EVL score, using logistic regression, and then evaluated its discrimination and calibration. Additionally, we compared the HOPE-EVL score with commonly used prognostic scores, such as the MELD score and Child-Pugh score.
Results
The study included 980 patients: 536 in the development cohort and 444 in the validation cohort. In-hospital mortality was 13.6% and 10.1% for the respective cohorts. The scoring system used five variables: systolic blood pressure (<80mmHg: 2 points), Glasgow Coma Scale (≤12: 1 point), total bilirubin (≥5 mg/dL or ≥85.5 µmol/L: 1 point), creatinine (≥1.5 mg/dL or ≥132.6 µmol/L: 1 point), and albumin (<2.8 g/dL or <28 g/L: 1 point). The risk groups (low: 0–1, middle: 2–3, high: ≥4) corresponded to observed and predicted mortality probabilities of 2.7% and 2.8%, 25.6% and 27.4%, and 73.3% and 74.7%, respectively. In the validation cohort, the HOPE-EVL score demonstrated excellent discrimination ability (area under the curve [AUC]: 0.890, 95% confidence interval [CI]: 0.850–0.930) compared to the MELD score (AUC: 0.853, 95% CI: 0.794–0.912), Child–Pugh score (AUC: 0.798, 95% CI: 0.727–0.869).
Conclusions
The HOPE-EVL score practically and effectively predicts in-hospital mortality. This score could facilitate the appropriate allocation of resources and effective communication with patients and their families.

Receiver operating characteristic curves comparing the HOPE-EVL, MELD, and Child–Pugh scores and the age between the development and validation cohorts
The mean observed and predicted probabilities of in-hospital mortality stratified by the HOPE-EVL score for each risk group