Objective: To determine whether a complaint of visible rectal bleeding that is elicited by a screening review of systems merits investigation and to assess the accuracy of a defined protocol to evaluate bleeding.
Design: Prospective cohort study.
Setting: Primary care clinics in a veterans medical center.
Patients: We used an 8-item review of systems to identify 297 individuals with visible rectal bleeding; 201 (68%) of these individuals completed a specified protocol consisting of double-contrast barium enema (DCBE) examination, rigid sigmoidoscopy, and follow-up visit after 6 to 12 months. Ten years later we verified the diagnosis in 131 (93%) of 141 patients whose initial evaluation suggested no cause, or a benign anorectal cause, of bleeding.
Main outcome measures: Final diagnoses after 2 and 10 years; sensitivity and specificity of symptoms, DCBE, and rigid sigmoidoscopy.
Results: We diagnosed serious disease in 48 (24%) of the 201 patients; 26 had polyps, 9 had inflammatory bowel disease, and 13 (6.5%) had colon cancer. Symptoms did not predict the diagnosis. Neither DCBE nor rigid sigmoidoscopy alone was sufficiently sensitive to be used alone, but the combination of DCBE and rigid sigmoidoscopy had a sensitivity of 0.96 and a specificity of 0.76 for the diagnosis of polyps, cancer, or inflammatory bowel disease.
Conclusions: Self-reported rectal bleeding detected by means of a review of systems was associated with a high likelihood of important pathology. Physicians should ask all adults about visible rectal bleeding and should visualize the entire colon in those who report bleeding.