Secondary Logo

Journal Logo

Abstracts: ACCEPTED: GI BLEEDING

Acute Overt Gastrointestinal Bleeding: Etiologies, Interventions, and Outcomes in an Urban Health System

589

Niu, Bolin MD; Tarangelo, Nicholas P. MD; Shah, Mital MD; Ahmed, Monjur MD

Author Information
American Journal of Gastroenterology: October 2017 - Volume 112 - Issue - p S317
  • Free

Introduction: Overt gastrointestinal bleeding (GIB) is defined as hematemesis, melena, and/or hematochezia. Endoscopy is often the first-line in diagnosis and management. However, interventional radiology (IR) and surgery may be necessary as well. We compare the characteristics of patients who received inpatient endoscopic treatment, IR embolization, or surgery versus those who did not.

Methods: The electronic medical data from Thomas Jefferson University Hospital and Methodist Hospital in Philadelphia related to overt GIB from 1/1/2014 to 12/31/2015 were queried using ICD-10 codes. 1257 records were reviewed and 773 patients were included. Exclusion criteria included age<18 and occult GIB. T-test was used to compare characteristics of those who underwent inpatient procedures and those who did not.

Results: Of 773 patients with overt GIB, the average age was 67.8 years. Patients were 55.4% male and 44.6% female. In-hospital mortality was 3.4%, with 69.2% of deaths related to variceal bleeding. Of 82 variceal bleeds (65 esophageal, 10 gastric, 7 both), band ligation was performed in 40 patients, all with esophageal varices (EV). Of 10 patients with gastric varices (GV), transjugular intrahepatic portosystemic shunt (TIPS) was performed in 8. In terms of endoscopic procedures, 89.4% of patients received endoscopic evaluation and 15.7% of patients underwent endoscopic therapy. Most common etiologies of bleeding were: diverticular (112), variceal (82), duodenal ulcer (81), and gastric ulcer (80). Most common therapies included band ligation (40), endoclip (37), and cautery (31). For non-endoscopic procedures performed in 5.7% (44) of patients, the most common were IR angiogram (16), surgery (15), and IR embolization (13). Arteriovenous malformations were the most common reason for IR angiogram, while IR embolization was performed most often for upper GI ulcers (Table 1). A comparison of patient characteristics shows endoscopic therapy and surgery were performed in significantly younger patients (Table 2). IR embolization was performed in those with significantly lower hemoglobin on admission.

Table
Table:
Table. Etiology of Bleeding Requiring Non-endoscopic Interventions
Table
Table:
Table. Comparison of Patient Characteristics by Procedure Versus No Procedure

Conclusion: Overall in-hospital mortality from overt GIB is low, with most deaths resulting from variceal bleeding. TIPS was instrumental in therapy of GV bleeds. While most patients received endoscopic evaluation, most did not require endoscopic therapy likely due to successful medical therapy. IR and surgery also have important roles in a small number of cases.

© The American College of Gastroenterology 2017. All Rights Reserved.