Objectives: To assess the clinical yield of colonoscoping in patients who present with rectal bleeding, persistent abdominal pain, or change in bowel habits in the absence of bleeding.
Design: A prospective study of colonoscopy patients.
Setting: Three colonoscopy practices in New York City between April 1986 and November 1989.
Patients: Eight hundred sixty-one patients with rectal bleeding, 113 patients with abdominal pain, 154 with change in bowel habits, and 44 patients with both abdominal pain and bowel change.
Main results: Of 861 patients with rectal bleeding, 293 (33.6%) had colonic neoplasia (8.6% cancer and 25% adenomatous polyps). In patients who had colonoscopy because of abdominal pain (n = 113), change in bowel habits (n = 154), or both abdominal pain and bowel change (n = 44), respectively 25 (22.1%), 42 (27.3%), and 10 (22.7%) had colonic neoplasms. If one looks at significant neoplasia (cancer or adenomas > 1 cm), then the findings in rectal bleeders were 14.5%, whereas the abdominal pain, change in bowel habits, and both groups had 7.1%, 7.1%, and 13.6%, respectively. Patients with rectal bleeding were more likely to have multiple adenomas than those with nonbleeding symptoms (p < 0.05).
Conclusions: Patients with persistent nonbleeding GI symptoms, including abdominal pain and change in bowel habits, have almost as high a yield of colorectal neoplasia as those with rectal bleeding.