MELD score, child-pugh score, and decreased albumin as risk factors for gastric variceal bleeding.

Background/aims: Ruptured esophagogastric varices are commonly associated with bleeding in patients with portal hypertension. However, the prediction of esophageal variceal bleeding is not matched by means of predicting gastric variceal bleeding. The present study aim is to elucidate risk factors for gastric variceal bleeding.

Methodology: Twelve patients with gastric variceal bleeding and 18 patients receiving preventive treatment for gastric varices were included in the study.

Results: The Child-Pugh (8.0 +/- 0.9 vs. 5.5 +/- 0.3; p = 0.0025) and Model for end-stage liver disease (MELD) (10.6 +/- 2.7 vs. 4.0 +/- 0.9; p = 0.0095) scores were significantly higher for patients with bleeding than for those receiving preventive treatment. Serum albumin concentration was significantly lower in bleeding than in preventive treatment cases, as determined by univariate (2.9 +/- 0.2 vs. 3.7 +/- 0.1 mg/dL; p < 0.0001) and multivariate analyses of serological data (odds ratio, 0.02, 95% confidence interval, 0.001-0.479; p = 0.0144).

Conclusions: The Child-Pugh and MELD scores were significantly higher for patients with gastric variceal bleeding than for those receiving preventive treatment, and multivariate analysis revealed that serum albumin was significantly lower in patients with gastric variceal bleeding. Control of serum albumin is important in preventing gastric variceal bleeding.