Are physicians doing too much colonoscopy? A national survey of colorectal surveillance after polypectomy.

Background: Increasing use of colonoscopy for colorectal cancer screening and surveillance of colorectal adenomas after polypectomy has given rise to concerns about the availability of endoscopic resources in the United States. Guidelines recommend surveillance after polypectomy at 3 to 5 years for a small adenoma, and follow-up is not advised for hyperplastic polyps. The intensity of physicians' surveillance is largely unstudied.

Objective: To survey practicing gastroenterologists and general surgeons about their perceived need for the frequency of surveillance after polypectomy, to compare survey responses to practice guidelines, and to identify factors influencing their recommendations for surveillance.

Design: Survey study conducted by the National Cancer Institute.

Setting: A nationally representative study of physicians in the United States.

Participants: 349 gastroenterologists and 316 general surgeons.

Measurements: Questionnaires mailed in 1999 and 2000 assessed physicians' recommendations for surveillance after polypectomy in asymptomatic, average-risk patients.

Results: Response rates were 83%. Among gastroenterologists (317 of 349) and surgeons (125 of 316) who perform screening colonoscopy, 24% (95% CI, 19.3% to 28.7%) of gastroenterologists and 54% (CI, 44.9% to 62.5%) of surgeons recommend surveillance for a hyperplastic polyp. For a small adenoma, most physicians recommended surveillance colonoscopy and more than 50% recommended examinations every 3 years or more often. Physicians indicated that published evidence was very influential in their practice (83% [CI, 78.8% to 87.2%] of gastroenterologists and 78% [CI, 72.5% to 86.8%] of surgeons). By contrast, only half of respondents reported that guidelines were very influential.

Limitations: The study was based on physicians' self-reported practice patterns. Results may overestimate or underestimate the performance of surveillance colonoscopy.

Conclusions: Some surveillance colonoscopy seems to be inappropriately performed and in excess of guidelines, particularly for hyperplastic polyps and low-risk lesions such as a small adenoma. These results suggest unnecessary demand for endoscopic resources.