Colonoscopy in acute lower gastrointestinal bleeding

Introduction: Colonoscopy is the procedure of choice for complete evaluation of the colon. In acute lower GI bleeding, the goal of colonoscopy is to identify the bleeding site and utilize endoscopic therapy when appropriate. Aim: To assess the role of colonoscopy in acute lower GI bleeding. Methods: The Bleeding Team data base which has been created to prospectively collect information on all patients in our institution with gastrointestinal hemorrhage was used to identify pts presenting with hematochezia between 1988-1996. This information included clinical presentation, colonoscopic assessment, additional tests, surgery, and outcome. Only the initial urgent colonoscopy was evaluated. A bleeding site was definitively identified if active bleeding, bleeding stigmata, or blood limited to a segment were identified. Results: 666 bleeding events involved in 524 pts. Mean age was 71±14 yrs (range 20-100; F:283, M:383). Colonoscopy was performed within 12 hrs. Oral polyethylene glycol lavage was used in 613 (92%), tap water enema in 14 (2%) and fleet enema in 24 (3.6%). Bowel preparation was reported as good in 605 (91%) and poor in 49 (7%). 6% of polyethylene glycol preps were poor. Cecal intubation was achieved in 98%. A definitive diagnosis was achieved in 280 (42%) pts. The main endoscopic diagnoses are presented in the following table: Diagnosis Number (%) Definitive (%) Endo Tx (% Surgery (%) Rebleed (%) Diverticula 236 (35) 24(10) 4(2) 21 (9) 33(14) Angioectasia 59 (9) 51(86) 51(86) 11 (19) 7(12) Tumor 50 (7.5) 49 (98) 0 45 (90) 1 (2) Polypectomy 27 (4) 26(96) 23(85) 1 (3.7) 0 Radiation 20 (3) 15(75) 7(35) 0 3(15) Dieulafoy's 4 (0.6) 4(100) 4(100) 0 1 (25) (Endo Tx: endoscopic therapy; definitive: definitive diagnosis; rebleed: rate) Surgery was performed in 101 pts (19%), for tumor in 47, diverticula in 21 and angioectasia in 11 pts. Rebleeding rates post-operative and post-endoscopic therapy were 2% and 5% respectively. Of 3 perforations (0.45%), 1 pts required surgery, and 2 were related to endoscopic therapy (Nd-YAG laser, polypectomy). Conclusions: 1. The diagnostic yield of urgent colonoscopy in acute lower GI bleeding is high. 2. Colonoscopy and endoscopic therapy are safe and have a low bleeding rate. 3. There are a limited number of diagnoses for which definitive endoscopic therapy can be applied. 4. Polyethylene glycol lavage is an effective bowel preparation prior to colonoscopy in acute GI bleeding.