Introduction The latest (2010) British Society of Gastroenterology (BSG) guidelines on colonoscopic surveillance in colitis (1) propose surveillance intervals based on risk of dysplasia determined by endoscopic and histological findings using pan-colonic chromoendoscopy. The 2010 surveillance strategy is projected to be significantly more cost-effective than previous BSG guidelines (2002). Our unit employs a two-tiered strategy using high-risk features including and additional to those of the BSG 2010 guidelines. We aim to compare the different surveillance guidelines (BSG 2002, BSG 2010 and two tier strategy) with respect to cost of surveillance and detection of endoscopically visible dysplasia.

Methods Patients who had a surveillance colonoscopy for colitis between 2003 and 2008 and had an endoscopic resection of a dysplastic lesion were identified from the endoscopic database. Colonoscopy and histopathology reports and clinical notes were reviewed. Surveillance intervals, based on the findings of predysplasia colonoscopy, were predicted using each of the three strategies. The cost of colonoscopy was based on national tariff cost of £476. If the colonoscopy at which dysplasia was detected was within the predicted surveillance interval, the surveillance strategy was deemed to have been successful at detecting dysplasia.

Results Twenty-six patients meeting inclusion criteria were identified. 16 (60%) patients were male. The median age was 66 years (IQR 47–70 years) and the median duration of disease 23 years (IQR 17–33 years). The calculated cost of surveillance per year for the 26 patients was £7933 for BSG 2002 guidelines, £5522.6 for BSG 2010 guidelines and £11 107 for two-tiered strategy. Using BSG 2002 guidelines, 22 of 26 patients (85%) would have a surveillance colonoscopy that would have detected dysplasia, 12 of 26 (46%) according to BSG 2010 guidelines and 24 of 26 patients (92%) according to two-tiered strategy.

Conclusion BSG 2010 guidelines are less costly than BSG 2002 guidelines and two-tiered strategy. However, using the surveillance intervals based on BSG 2010 guidelines would not have detected as many patients with dysplasia as BSG 2002 guidelines or two-tiered strategy. However, as the outcome of low-grade dysplasia in colitis is not certain, it is unclear whether that would have altered patient outcomes. Prospective audit of patient outcomes using 2010 surveillance intervals is needed to assess the effect of delayed detection of dysplastic lesions in ulcerative colitis.