Predicting early rebleeding and mortality after endoscopic hemostasis of esophagogastric varices: Diagnostic performance of aspartate aminotransferase-to-platelet ratio index and model for end-stage liver disease-Na score.

ABSTRACT

BACKGROUND: Endoscopic variceal ligation and sclerotherapy are recommended for esophagogastric variceal bleeding (EGVB) in cirrhosis but can be complicated by early rebleeding and death. This study aimed to identify noninvasive markers accurately predicting early rebleeding and mortality after endoscopic hemostasis for EGVB. METHODS: Among 116 patients with endoscopically confirmed EGVB and endoscopic hemostasis, various noninvasive markers were calculated, and their predictive accuracy was compared by receiver-operating characteristic curve analysis. Endpoints included 5-day rebleeding, 5-day mortality, 6-week rebleeding, and 6-week mortality. RESULTS: The median age was 63 years. Child-Pugh class B and C patients accounted for 40.5% and 34.5%, respectively. Only the aspartate aminotransferase-to-platelet ratio index (APRI) significantly predicted 5-day rebleeding, with an area under the curve (AUC) of 0.777 (95% confidence interval [CI] 0.537-1). The model for end-stage liver disease-Na (MELD-Na) score showed good predictive accuracy for 5-day mortality (AUC 0.839, 95% CI 0.681-0.997), 6-week rebleeding (AUC 0.797, 95% CI 0.663-0.932), and 6-week mortality (AUC 0.888, 95% CI 0.797-0.979). CONCLUSIONS: Patients with cirrhosis with a high APRI and MELD-Na score were at high risk of early rebleeding and death after EGVB. Allocating appropriate monitoring and care for those patients is necessary.