Prospective study of the frequency and size distribution of polyps missed by colonoscopy.

An important determinant in interpreting the results of colorectal polyp chemoprevention trials is the rate of polyps missed during colonscopic examination. We prospectively examined 90 patients by tandem colonoscopy performed by two alternating examiners. In 69 (76.7%) patients, 221 neoplastic lesions were documented histologically. Of a total of 58 lesions detected in 31 patients, no neoplastic lesion greater than or equal to 10 mm in size was missed; 16% of diminutive (less than or equal to 5 mm) neoplastic polyps and 12.3% of medium-sized (6-9 mm) neoplastic polyps were missed by the first examiner. We conclude that an experienced colonoscopist will miss about 15% of colorectal neoplastic polyps less than 10 mm in size in the setting of adequate bowel preparation. Large (greater than or equal to 10 mm) polyps were rarely missed, however, with the "miss" rate in our study equal to 0, with a 95% confidence limit of 4.64%.