P4176 - Can the FIB-4 Score Predict Rebleed in Cirrhosis? A Multi-Center, Retrospective Study in the Community Setting

Kaiser Kabir, DO, MS¹, Judy Sheffeh, MD², Ermal Hasalliu, DO², Kathryn Claxton, BA³, Mitchell Oliver, DO², Chase Zeilenga, DO¹, Benjamin Collins-Hamel, DO^2
1Ascension Macomb-Oakland Hospital, Detroit, MI; ²Ascension Macomb-Oakland Hospital, Warren, MI; ³Kansas City University, Warren, MI

Introduction: Multiple studies have attempted to evaluate predictors of varices and first-time variceal hemorrhage. This study aimed to determine if the Fibrosis-4 (FIB-4) score could predict the incidence of rebleed after hospitalization for upper gastrointestinal bleed (GIB) in patients with cirrhosis.

Methods: This was a retrospective study including adults with cirrhosis presenting with GIB at 5 Southeast Michigan community hospitals between 2015-2019. Patients were identified using ICD-10 codes before chart review. Demographics, endoscopic findings, cirrhosis etiology, 7 and 28-day rebleed and in-hospital death data were collected. The occurrence of rebleed at 7 and 28 days compared to the FIB-4 score was analyzed via T-test. Regression analysis looked at the associations of FIB-4 score on in-hospital death with 7 and 28-day rebleed, and type of endoscopic intervention with 7 and 28-day rebleed.

Results: A total of 218 patients were included, with a median age of 66. Eleven patients were in the 7-day rebleed group, with 3 in-hospital deaths, and 25 patients were in the 28-day rebleed group, with 2 in-hospital deaths. A total of 199 underwent upper endoscopy and 96 received intervention. Per regression analysis, the FIB-4 score did not predict the incidence of rebleed at day 7 (CI=0.895-1.055, p-value=0.497) nor at day 28 (CI=0.789-1.096, p-value=0.312). A larger proportion of patients in the 7-day rebleed group had ascites, were on beta-blockers before hospitalization, had albumin levels between 2.4-3.6 g/dL and experienced in-hospital death. There was no association between the type of endoscopic intervention and rebleed. 79 patients were on beta-blockers and, of them, 32 required endoscopic intervention, with 26 for esophageal varices, 7 for arteriovenous malformations, 2 for Dieulafoy’s lesions and 1 for gastric antral vascular ectasia.

Discussion: This study shows that the FIB-4 score cannot be a reliable predictor of rebleed in cirrhosis. The low sample size in the rebleed group is a limitation. A larger proportion of patients on beta-blockers were in the 7-day rebleed group. This may not always mean a failure of beta-blocker therapy as these patients had other GIB sources with possible alternate risk factors. Enhancing our ability to risk stratify patients for rebleed may help to prioritize patient transfer and plan closer outpatient surveillance before trans-intrahepatic portosystemic shunt or transplant, especially for patients initially presenting to centers without these services.


Disclosures:


Kaiser Kabir, DO, MS¹, Judy Sheffeh, MD², Ermal Hasalliu, DO², Kathryn Claxton, BA³, Mitchell Oliver, DO², Chase Zeilenga, DO¹, Benjamin Collins-Hamel, DO². P4176 - Can the FIB-4 Score Predict Rebleed in Cirrhosis? A Multi-Center, Retrospective Study in the Community Setting, ACG 2024 Annual Scientific Meeting Abstracts. Philadelphia, PA: American College of Gastroenterology.