Systematic literature review and pooled analyses of risk factors for finding adenomas at surveillance colonoscopy

Background and study aim: Colorectal cancer (CRC) screening guidelines recommend surveillance after polypectomy. There is variation in the surveillance intervals that are being advised. This variation also affects adherence. Surveillance intervals need to be based on risk factors at index. We therefore aimed to systematically review risk factors of adenoma findings at surveillance colonoscopy. Methods: A systematic literature search was performed up to September 2009. Studies that reported on follow-up colonoscopy findings with stratification for index characteristics were included. Pooled relative risks (RR) were calculated using random effects models, and heterogeneity was determined by means of the I(2)-statistic. Results: A total of 27 studies met the inclusion criteria. The most important risk factors for adenoma findings were the presence on index colonoscopy of the following: advanced adenomas (RR: 1.81), >= 3 adenomas (RR: 1.64), size >= 10 mm (RR: 1.66), and age >= 60 years (RR: 1.65). The presence of villous adenomas, high grade dysplasia, proximal adenomas, and male gender were associated with less profound increases in RR. Marked variation in study design and substantial heterogeneity between studies was observed. Conclusions: Convincing evidence exists that patients with advanced adenomas, >= 3 adenomas, adenomas >= 10 mm, or age >= 60 years have an increased risk of adenoma recurrence. The evidence for other baseline findings for an increased risk of adenoma recurrence is inconclusive. Marked variation and consistently lower RRs in studies of medium or low quality emphasize the necessity for well performed and well reported studies. Given the high impact of surveillance on patients and service providers, there is need for further assessment of the risk(s) of adenoma recurrence.