ACCURACY OF AAR, APRI, FIB-4 AND MELD SCORES AS PREDICTORS OF A VARICEAL SOURCE AMONG CIRRHOTIC PATIENTS PRESENTING WITH UPPER GASTROINTESTINAL BLEEDING

Aims In patients with cirrhosis admitted with upper gastrointestinal bleeding (UGB), identification of a variceal source prior to endoscopy would be clinically useful. We determined the accuracy of aspartate aminotransferase to alanine aminotransferase ratio (AAR), aspartate aminotransferase to platelet ratio index (APRI), fibrosis-4 index (FIB-4) and model for end-stage liver disease (MELD) in identifying a variceal origin of UGB in cirrhotic patients.

Methods Cirrhotic patients undergoing upper endoscopy for UGB between 7/2019 and 10/2021 were retrospectively reviewed. Diagnostic accuracies were assessed using the area-under-receiver-operating-characteristic-curve (AUROC) with 95% confidence intervals (CIs).

Results Overall, 59 patients (74.6% males, 61.3±11.9 years) with cirrhosis of different etiologies (alcohol=22, viral hepatitis=18, non-alcoholic steatohepatitis=6, other=13) were included. Among them, 45 (76.3%) had esophageal varices (EVs) and in 39 (66.1%) varices were identified as the culprit of UGB. The most common cause of non-variceal bleeding were peptic ulcers (10/59; 16.9%). The AUROCs for predicting the presence of EVs were AAR 0.51 (95%CI: 0.35-0.67), APRI 0.54 (95% CI: 0.35-0.73), FIB-4: 0.47 (0.31-0.64) and MELD 0.42 (95%CI: 0.24-0.59). For identifying a variceal source of UGB, the AUROCS were AAR 0.48 (95%CI: 0.32-0.64), APRI 0.41 (95%CI: 0.24-0.57) FIB-4 0.41 (95%CI: 0.25-0.57) and MELD 0.40 (0.22-0.53).

Conclusions AAR, APRI, FIB-4 and MELD exhibited low performance for predicting the presence of EVs and for identifying a variceal source in patients with cirrhosis admitted with UGB.

Publication History

Article published online:
14 April 2022