Artificial bowel markers: a novel method for measuring the accuracy of colonoscopy.

Purpose: Colonoscopy is the "gold standard" for assessing colonic mucosal abnormalities. An important component of this is complete examination to the cecum. However, the ability to detect abnormalities is equally important and has received less attention. This study was designed to assess the accuracy of detection of artificial bowel markers as surrogates for small polyps.

Methods: Patients were randomly assigned to receive between zero and four of each of two types of marker. Markers used were 5-mmx5-mm pieces of latex-free rubber or 12-mmx1.5-mm metallic bowel clips. These were placed on insertion of the colonoscope. At the limit of the colonoscopy, a second blinded endoscopist performed the extubation. Six endoscopists took part in the study. Data regarding the number, type, and position of the markers on insertion and withdrawal were recorded as were insertion and withdrawal times.

Results: A total of 179 markers (85 clips, 94 rubbers) were placed in 44 patients. The cecal intubation rate was 91 percent. Median intubation time was 20 minutes and withdrawal time was 15 minutes; 139 markers (77.7 percent) were detected on withdrawal (clips, 76.5 percent; rubbers, 78.7 percent). There was no correlation between insertion or withdrawal times and marker detection. Detection rates varied between endoscopists (71-87 percent), but there was no correlation with individual cecal intubation rates (P=0.96). Markers placed at flexures were missed more often (P=0.008).

Conclusions: Artificial markers provide a novel method for assessing colonoscopic accuracy. The results obtained with this method are closely concordant with results from other studies. The technique is potentially useful in training, audit, and future research.