One hundred patients with occult gastrointestinal bleeding (OGIB) (i.e., guaiac-positive stools and/or iron deficiency anemia) were prospectively evaluated with bidirectional endoscopy [upper endoscopy (EGD) and colonoscopy] to determine the origin of occult bleeding. Predetermined criteria were used to prospectively define gastrointestinal bleeding sources. Among the 58 males and 42 females, the median age was 65 yr. Thirty-one percent of the group had gastrointestinal symptoms. Sixty-six percent of the study group were inpatients. Bidirectional endoscopy detected the source of OGIB in 53% of patients, with a positive finding on EGD of 36%, and with colonoscopy, of 26%. In only 9% of patients was a source of OGIB detected on both EGD and colonoscopy. Acid peptic disease accounted for the source of OGIB in 27%, colonic adenomas 14%, angiodysplasia 13%, colorectal carcinoma 6%, and gastric cancer in 1% of patients. The diagnostic yield was significantly higher with EGD than with colonoscopy in patients with anemia and guaiac-positive stools (45% vs. 26%, p less than 0.01). Upper endoscopy directed a change in patient management in 29 patients.
In conclusion: for the patient population described in this study, bidirectional endoscopy determined the source of OGIB in 50%. As expected, colonoscopy resulted in a higher cancer detection rate than EGD--yet EGD detected the origin of OGIB in 68% (36/53) of patients found to have an occult bleeding source, and resulted in a therapeutic initiation or a change in therapy for 30% of all patients.