The aims of this study were to assess the clinical observation of outpatient who showed hematochezia, and to determine whether specific clinical symptoms associated with hematochezia were predictive of important gastrointestinal pathology.

Prospective study was carried out from July 1998 to July 1999 with sixty-five outpatients(35 males and 30 females with mean age, 43±11 years) who had no evidence of recent bleeding. Patients were interviewed by questionnaires about the amount and frequency of bleeding, change in bowel habits, weight loss, usage of aspirin/NSAIDs, and family history, prior gastrointestinal pathologic illness before colonoscopy. Based on this information, endoscopist were asked to predict whether the bleeding was from a benign perianal or other lesion. Important gastrointestinal pathology was defined as carcinoma, adenomas more than 1cm, active ulcerative colitis, and active tuberculosis by colonoscopy.

Colonoscopic findings were as follows : 27 cases of benign anorectal lesion ; 16 cases of polyps, 10 cases of normal ; 8 cases of acute colitis and nonspecific colitis ; 7 cases of coloerctal cancer, 7 cases of ulcerative colitis and intestinal tuberculosis ; and other cases. Important gastrointestinal pathology was 17 cases. Variables including duration, type and frequency of bleeding, weight loss and change in bowel habit did not predict the colonoscopic diagnosis. Of the 35 patients diagnosed clinically by endoscopist to begin anorectal lesion alone, 18 patients were found to have benign anorectal lesion, 2 patients had cancer, 2 patient had polyp(bigger than 1cm), and 1 patient had ulcerative colitis.

In outpatients with hematochezia, the incidence of colon cancer was 10.8%. Clinicians were unable to distinguish significant colonic lesions by history. Therefore accurate diagnostic workup is needed for this group of patients.