182

Robert A Enns¹, Alan N Barkun², Yves M Gagnon³, David Armstrong⁴, Jamie C Gregor⁵, Richard N Fedorak⁶, and the RUGBE investigators*

¹Division of Gastroenterology, University of British Columbia, Vancouver, British Columbia; ²Department of Medicine, Division of Gastroenterology, McGill University and the McGill University Health Center, Montreal, Quebec; ³Occam Research & Consulting Inc; ⁴Division of Gastroenterology, McMaster University Medical Centre, Hamilton, Ontario; ⁵Division of Gastroenterology, University of Western Ontario, London, Ontario; ⁶Division of Gastroenterology, University of Alberta, Edmonton, Alberta

A routine ‘second-look’ in patients with non-variceal UGI bleeding is controversial. It refers to the repeat endoscopic evaluation in a stable patient to assess/treat the lesion further. Conflicting reports have been published and recently, a meta-analysis suggested a more positive outcome in select patients.

OBJECTIVE: We evaluated the Registry of Upper Gastrointestinal Bleeding and Endoscopy (RUGBE) to determine: a) How many patients in this Canadian registry underwent a routine second endoscopy and b) How the outcomes (rebleeding, surgery, death and length of stay) compared to those who didn’t undergo a routine second look.

METHODS: Between 06/99 and 04/02, data from 1869 randomly selected patients from 18 centers in Canada with an episode of UGIB who underwent EGD were abstracted and analyzed. All patients who underwent more than one endoscopy were initially evaluated and, subsequently, only those patients who underwent a routine ‘second-look’ endoscopy were studied. Outcomes for these groups were calculated and compared.

RESULTS: A total of 570 patients (30%) underwent a second endoscopy of whom 208 (36.5%) underwent further therapeutic intervention. The most common indications for a second endoscopy were rebleeding (263/570, 46%), clarification of diagnosis (152/570, 27%) and routine second-look (111/570, 19%). In the 111 patients who underwent routine second look, only 19% underwent further endoscopic therapy. In this group only 1 (0.9%) patient died and 4 (3.6%) underwent surgical therapy for recurrent/ongoing bleeding. This data compares favorably to the entire group of patients in RUGBE, where the surgical rate/mortality was 6.5% and 5.4% respectively. The mean length of stay was 8.4 days in the ‘second-look’ category.

CONCLUSION: Second-look endoscopy is practiced in Canada. Although there are numerous confounders in this retrospective analysis, the outcomes in patients for whom routine second-look was done are excellent. Further data analysis, accounting for other confounding variables may support this practice in high risk patients.

NEXT ABSTRACT >