Colonoscopy is not a perfect test. It misses a substantial number of neoplastic lesions and has some risk of missing cancer. Nearly all work on detection during colonoscopy has focused on the withdrawal phase of the examination. This randomized, controlled trial will compare the additional effect on the rate of adenoma detection of mucosal inspection during colonoscope insertion, with inspection during instrument withdrawal, in patients undergoing colonoscopy for colorectal cancer screening or surveillance.

Background: Colonoscopy is not a perfect test. It misses a substantial number of neoplastic lesions and has some risk of missing cancer. Nearly all work on detection during colonoscopy has focused on the withdrawal phase of the examination. Thus, colonoscopy is typically performed by rapidly passing the instrument through the loops and bends of the colon in order to reach the tip of the cecum, and then performing a slow withdrawal in which the tip of the instrument is systematically deflected, and the mucosa is careful cleaned and suctioned, to expose all of the colonic mucosa for viewing.

Many experienced colonoscopists recognize that small polyps seen incidentally but not removed during insertion are sometimes quite difficult to find during withdrawal. The reason for this observation is probably because the colon is in a very different anatomical conformation during endoscope insertion and withdrawal. During insertion, the colon is in its natural conformation in which the sigmoid and transverse colon has several sharp bends or flexures, and the overall length has not yet been shortened. In this phase, the colon is often significantly stretched because of the formation of loops and bends in the colonoscope. This greatly affects the conformation of the colonic wall visualized proximal to the instrument tip. During withdrawal, the colon is shortened and pleated over the colonoscope, with successive regions of the colon being inspected as they slip off the end of the instrument. Thus, segments of visualized colon are often much straighter during withdrawal than during insertion. The insertion and withdrawal phases, therefore, expose somewhat different sections of the mucosal surface to the colonoscope and inspection on insertion and withdrawal are, quite possibly, complementary.

Aims: This randomized, controlled trial will compare the additional effect on the rate of adenoma detection of mucosal inspection during colonoscope insertion, with inspection during instrument withdrawal, in patients undergoing colonoscopy for colorectal cancer screening or surveillance.

Study procedure: In this study, we plan to investigate whether a specified interval of inspection during insertion can increase overall adenoma detection. We will conduct a randomized controlled trial, in which patients will be randomized to have all of the inspection performed during the withdrawal phase (as is usual care) versus having several minutes of examination specifically devoted to inspection during insertion.

Inclusion Criteria:

• Age ≥ 50 years
• Patients undergoing colonoscopy for screening or surveillance indications

Exclusion Criteria:

• Previous surgical resection of all or part of the colon.
• Inability to give informed consent.
• Ulcerative colitis or Crohn's disease.
• Polyposis syndrome or Lynch syndrome (HNPCC)
• Any comorbid condition which the investigator deems would put the patient at increased risk from a slightly prolonged procedure