This was a retrospective observational study performed in the ED of a tertiary university hospital visited by 43,000 patients annually, with approximately 60 patients presenting with AVB. This study was approved by the Institutional Review Board of Wonju Severance Christian Hospital (IRB No. CR321093). The study protocol conformed to the ethical guidelines of the Declaration of Helsinki (1975) and its later amendments. As the study involved retrospective and observational analysis, the requirement for informed consent was waived, and patient records and information were anonymized before analysis. Computerized hospital records were reviewed, and any patients for whom “esophageal varices with hemorrhage (I85.01)” based on the International Classification of Diseases, 10th revision coding, was used as a discharge code were initially considered for study selection. Included were patients diagnosed with “variceal bleeding” via esophagogastroduodenoscopy after being admitted to the hospital.
A total of 522 patients with upper gastrointestinal bleeding visited the emergency department between January 2019 and December 2020, and 137 patients were retrospectively diagnosed with variceal bleeding according to their medical records. Patients who did not undergo laboratory tests were excluded from the analysis.
Data were collected from a retrospective review of patient electronic medical records performed by emergency physicians blind to study objectives and hypotheses. Data on age, sex, hospitalization period, serum albumin level (reference range: 3.5–5.5 g/dL), d-dimer level (reference range: < 250 ng/mL), and variables required for each scoring system were recorded.
The Modified early warning score (MEWS) is scored based on systolic blood pressure, heart rate, respiratory rate, body temperature, and alertness . The Child–Pugh is scored based on bilirubin and albumin levels, the international normalized ratio (INR), the presence or absence of ascites, and the presence or absence of encephalopathy. The MELD score considers dialysis, serum creatinine, bilirubin, INR, and serum sodium level data . The AIMS65 score is based on the albumin, INR, mental status, presence of shock, and age > 65 years. The GBS score is based on hemoglobin and blood nitrogen urea levels, systolic blood pressure, sex, heart rate, clinical presentations, hepatic disease, and cardiac failure. In addition, the Pre-endoscopy Rockall score is based on age, presence of shock, and comorbidities 
The collected blood samples were sent to the laboratory for analysis (ADVIA 2120i automated hematology analyzer, Siemens Healthcare Diagnostics Manufacturing Limited, Dublin, Ireland), and the coagulation profile (CS-5100 hemostasis system, Sysmex Corp., Kobe, Japan), D-dimer, and albumin were detected with serum biochemical tests (Dimension VISTA 1500, Siemens, Delaware, USA).
The primary endpoint of this study was the efficacy of the DAR in predicting the need for intensive care, long-term hospitalization, transfusion in the ED, and mortality in patients with AVB who visited the ED. The secondary endpoint was the comparative efficacy of the MEWS, MELD, Child–Pugh, GBS, pre-endoscopy Rockall, and AIMS65 scores in predicting the need for intensive care, long-term hospitalization, transfusion, and mortality.
The admission criteria for the ICU are based on priority models . Admission is decided according to clinician judgment, vital signs as objective parameters, and clinical examination and acute onset physical signs when necessary. We categorized patients using the priority model. Priority 1 is considered for patients requiring intensive care and monitoring, such as those on a ventilator or intravenous cardiovascular medications, respiratory failure requiring ventilation after surgery, and those requiring invasive monitoring. Priority 2 is considered for patients who may need immediate treatment at any time during intensive monitoring and chronic disease that can rapidly worsen. Priority 3 is considered for patients with underlying or acute disease, which may require intensive care. Priority 4 is not appropriate for ICU. Long-term hospitalization was defined as hospitalization lasting longer than 14 days [23, 24].
Continuous data are presented as means with standard deviations or medians (interquartile ranges). The normality of data distribution was assessed using the Shapiro–Wilk test. Categorical variables are presented as counts and percentages. Continuous data were analyzed using Student’s t-test or the Mann–Whitney U test, as appropriate. Categorical data were analyzed using a chi-square test or Fisher’s exact test, as appropriate. To assess the predictability of the DAR and MEWS, Child–Pugh, MELD, pre-endoscopy Rockall GBS, and AIMS65 scores, receiver operating characteristic (ROC) curves were constructed using cut-off values determined with the Youden Index . Univariate and multivariate logistic regression analyses were performed to assess the factors contributing to ICU admission. Variables with a p-value of < 0.2 in the univariable logistic regression analysis were entered into the multivariable logistic regression analysis. Statistical significance was set at p < 0.05 . A Kaplan–Meier type plot was constructed to show estimated survival probabilities from after acute variceal bleeding at 30 days. All analyses were performed using SPSS ver. 23 (IBM Corp., New York, NY, USA) and MedCalc Statistical software version 17.5.3 (MedCalc Software, Ostend, Belgium).