Urgent Balloon Assisted Enteroscopy in Acute Overt Suspected Small Bowel Bleeding Is Associated With Decreased Rehospitalization Rate and Increased Bleeding Free Survival

Introduction: While balloon assisted enteroscopy (BAE) has been widely utilized for overt suspected small bowel (SB) bleeding in the inpatient setting, timing of BAE remains under investigation. To date, there is no consensus on the optimal timing to perform BAE in patients with suspected SB bleeding. Previous studies showed that urgent BAE performed within 72 hours has higher diagnostic yield (DY) and therapeutic yield (TY). However, little is known regarding urgent BAE’s impact on long-term outcomes. We aim to assess the outcomes of hospitalized patients with suspected SB bleeding after undergoing BAE at a single tertiary referral academic institution. Methods: Patients hospitalized with suspected SB bleeding who underwent BAE during the admission from 2008-2022 with at least 6 months follow up were identified in a prospectively maintained database. Patients with incomplete follow up or procedure were excluded. Clinical characteristics were collected by chart review. BAE was considered urgent if performed within 72 hours of admission. DY was defined as the proportion of positive tests per the Saurin classification. TY was defined as proportion of BAE with endoscopic therapy performed. Variables were analyzed by Fisher’s Exact Test, Student’s T-Test, or multivariate logistic regression analysis where appropriate. Results: A total of 118 patients met the specified inclusion criteria, of which 63 (53%) underwent urgent BAE. The median follow-up time was 2.8 years (interquartile range 1.37-4.91). Patient outcomes are displayed in Table 1. DY and TY of urgent BAE (DY 60.3%; TY 54%) were similar to those of non-urgent BAE (DY 49.1%; TY 43.6%). Urgent BAE was independently associated with a lower rate of rehospitalization as compared to non-urgent BAE (29% vs 41.3%, P=0.04) with odds ratio of 0.45 (96% confidence interval [CI] 0.2, 0.99). Likewise, urgent BAE was also independently associated with longer time to rebleed as compared to non-urgent BAE (1.85 vs 0.51 years, P< 0.01) with odds ratio of 0.41 (96% -CI 0.15, 0.81). Two patients died from their bleeding – one each after urgent and non-urgent BAE. Conclusion: In the setting of acute suspected SB bleeding, urgent BAE portends better long-term outcomes as compared to non-urgent BAE. Interestingly, both DY and TY were similar between urgent and non-urgent BAE. A larger sample size and additional studies are needed to confirm these findings and to elucidate the underlying association. Table 1. - Clinical Outcomes in Urgent and Nonurgent Balloon-Assisted Enteroscopy Characteristic, N (%) Urgent DBE (N = 63) Non-Urgent DBE (N=55) P-Value Univariate Odds Ratio (95% CI) Multivariate Odds Ratio (95% CI)* Endoscopic Findings Angioectasia 18 (28.6%) 12 (21.8%) NS – – Ulcer 2 (3.2%) 6 (10.9%) Mass 4 (6%) 2 (3.6%) Varices 1 (1.6%) 1 (1.8%) Dieulafoy lesion 1 (1.6%) 1 (1.8%) Polyp 3 (4.8%) 2 (3.6%) Diverticulum 4 (6.3%) 2 (3.6%) Others 2 (3.2%) 1 (1.8%) Therapeutic Interventions APC 18 12 NS – – Endoclip 11 12 Bipolar catheter 3 - Hemostatic powder 1 - Cyanoacrylate glue - 1 Polypectomy 3 2 Diagnostic yield 38 (60.3%) 27 (49.1%) 0.30 -- -- Therapeutic yield 34 (54%) 24 (43.6%) 0.34 -- -- Rehospitalization 18/62 (29%) 27/54 (41.3%) 0.04 0.41 (0.19, 0.95) 0.45 (0.2, 0.99) Time to rebleed (years) 1.85 0.51 < 0.01 0.37 (0.15, 0.87) 0.35 (0.15, 0.81) *Multivariate analysis consisted of the following variables: Ventricular Assist Device, Hemoglobin, INR.