Portal de Periódicos da CAPES

302 Diagnostic Yield of Capsule Endoscopy vs. Colonoscopy in Patients With Melena and a Negative EGD

2011; Elsevier BV; Volume: 73; Issue: 4 Linguagem: Inglês

10.1016/j.gie.2011.03.042

1097-6779

Eugene Zolotarevsky, Amit G. Singal, Nitya Prabhakar, Akbar K. Waljee, Jason J. Grove, Jeff Costanzo, Ashish R. Shah, Darren M. Brenner, Jason Baker, Laurel Fisher,

Gastrointestinal disorders and treatments

EGD is the first step in the evaluation of patients with melena. Although colonoscopy has traditionally been the next test in patients without a bleeding source on EGD, video capsule endoscopy (VCE) is being used with increasing frequency. To compare the diagnostic yield of VCE and colonoscopy for identifying hemorrhagic lesions in patients with melena and an unremarkable EGD. We retrospectively reviewed the records of all patients at the University of Michigan who presented with melena between 2001-2009. Patients with a positive EGD (severe esophagitis, ulcerations, angioectasias, GAVE, varices, portal hypertensive gastropathy, bleeding tumors, or visualized active bleeding) were excluded. Patient demographics, clinical history, and labs at the time of presentation were recorded from the electronic medical record. Endoscopic data were extracted from endoscopy reports and an endoscopic database. A positive colonoscopy or VCE was defined as the presence of angioectasias, ulcerations, or an actively bleeding lesion. The yield of colonoscopy and capsule endoscopy was compared using Fisher's exact test. Among 987 patients who presented with melena, 305 (30.9%) had a negative EGD. Mean age of EGD-negative patients was 60.3 ± 0.9 years; 53.1% were males, and 82.2% were Caucasian. Of the 150 (49.2%) EGD-negative patients who underwent colonoscopy, 22 (14.7%) had right-sided colonic lesions and 9 (6.0%) had left-sided colonic lesions. Of the 119 patients without a bleeding source on colonoscopy, 44 underwent VCE and 22 (50%) had small bowel hemorrhagic lesions. Two EGD-negative patients with a positive colonoscopy (both with angioectasias) and eleven EGD-negative patients without prior colonoscopy also underwent VCE. Five (38.5%) of these 13 patients had a hemorrhagic lesion identified on VCE. Overall, VCE had a higher rate of identifying hemorrhagic lesions than colonoscopy (47.4% vs. 20.7%, p<0.001). Patients who had colonoscopy and VCE had the VCE performed an average of 12.7 ± 2.4 days after EGD compared to 1.9 ± 1.2 days in patients who had VCE without colonoscopy (p<0.001). Delayed VCE deployment was associated with a trend toward decreased yield for identifying hemorrhagic lesions, but this did not reach statistical significance (50.0% vs. 36.4%, p=0.51). In patients with melena and negative EGD, VCE has a higher yield than colonoscopy for identifying hemorrhagic lesions. Delayed VCE deployment may decrease the likelihood of identifying hemorrhagic lesions. Prospective trials comparing VCE and colonoscopy are needed to better determine if VCE should be the initial study in patients with melena and a negative EGD.