Introduction Modern colonoscopic surveillance follows clear evidence based guidelines, and clinical decisions appear straight forward. However, the guidelines do not define what should be done in cases with sub-optimal preparation.

Methods We retrospectively reviewed all adenoma and post carcinoma surveillance colonoscopies over 5 years (2006–2011) with a further year follow up looking for incident cancers.

Polyp site, number, size, and colonoscopy completion rates were recorded.

Two cleaning regimes were used: standard - Fleet; second line and in older patients/renal failure - KleanPrep, picolax and senna. Preparation was graded by the performing colonoscopist as satisfactory, sub-optimal or unsatisfactory.

Results 2176 patients underwent 2649 surveillance colonoscopies. Average patient age was 68.8 years. Mean follow up was 3.4 years.

3758 polyps were identified in 1539 procedures; 525 polyps were 1cm or larger. There was a marked trend towards lower polyp detection with worsening bowel preparation (p = 0.056 Chi²) with a marked reduction in mean number of polyps particularly on the right side with worsening preparation.

12 colorectal cancers were detected in this population. The overall cancer rate in this high risk population was 1 in 181 patients (95% C.I. 103 – 350).

73 patients with sub-optimal or unsatisfactory preparation had a repeat colonoscopy. In 64% the preparation was satisfactory, 28% were sub-optimal, and 8% had an un-satisfactory examination.

Conclusion 15% of procedures in our surveillance population have sub-optimal or un-satisfactory bowel preparation, making the interpretation of the clinical guidelines difficult.

Patients who have sub-optimal or unsatisfactory preparation have a high proportion of further sub-optimal or unsatisfactory procedures. Endoscopy units should have a strategy for improving this.

In patients with sub-optimal or unsatisfactory bowel preparation there is a significant reduction in overall polyp detection which is particularly marked in the right colon. This does not appear to be the case with larger polyps.

In patients with sub-optimal or unsatisfactory preparation undergoing a complete colonoscopy, shorter surveillance intervals should be considered taking other patient related factors into account.

Disclosure of Interest None Declared.