Introduction PR bleeding is a common indication for endoscopy. Other risk factors for sinister pathology include increasing age, anaemia, change in bowel habit and family history of colorectal cancer. In our centre there is debate about whether a flexible sigmoidoscopy, vs colonoscopy, is sufficient for young patients with PR bleeding alone. If sigmoidoscopy is sufficient it would reduce the risk patients are exposed to by full colonoscopy and the workload on the endoscopy unit. We ultimately aim to design a protocol for how to investigate PR bleeding.

Methods Retrospective review of all lower GI endoscopies done for either PR bleeding alone or in combination with another indication in 2008–2010. We reviewed patient age, indications and findings. Age groups were divided into <45 years or ≥45 years. Indications were divided into PR bleeding alone or plus another indication.

Results 1492 procedures were done in this period. 15 were abandoned. 17 of 199 (8.5%) procedures performed in people under 45 years for PR bleeding alone found polyps. The histology showed 10 metaplastic polyps, 1×15 mm rectal low grade villous adenoma, 1×12 mm sigmoid low grade tubulovillous adenoma, 1×2 mm sigmoid low grade tubular adenoma, 1 prolapsed haemorrhoid, 1 polypoid ganglioneuroma, 1×3 mm splenic polyp (not retrieved for histology) and 1×2 mm sigmoid polyp not removed given current GI bleed.

Conclusion No patients in the low risk group had cancer. Only two patients (1%) had large (>10 mm) polyps (low grade dysplasia, completely excised), both within reach of a flexible sigmoidoscope. No patients in the younger age group with PR bleeding as the sole indication would have had significant pathology (large polyp or cancer) missed due to having a flexible sigmoidoscopy rather than colonoscopy. This suggests that a protocol for this group could be implemented to prevent unnecessary tests with the associated incumbent risks. Further review of a larger cohort is required to ensure that this strategy does not expose patients to an unacceptable risk of missing significant pathology.

Competing interests None declared.