Background: Colonoscopy is the most sensitive screening and surveillance test for colorectal cancer and precancerous polyps, but it is also associated with risks of bleeding, perforation, and death. To maximize benefits and minimize risks from colonoscopy, national consensus guidelines have established recommended follow-up intervals for repeat colonoscopy based on findings and family history. A previous survey of endoscopists suggested that many were repeating colonoscopies earlier than guidelines recommended. Krist and colleagues compared the recommended intervals in contemporaneous and current (2006) guidelines with repeat colonoscopy recommendations from endoscopists that were received by primary care physicians.

The Study: The study sample consisted of 1,006 randomly selected adults 50 to 70 years of age who visited one of 10 primary care practices in the District of Columbia, Mary-land, and Virginia between February 2005 and August 2006. Chart reviews collected information on age, sex, family history of colorectal cancer, previous colonoscopies (screening, surveillance, or diagnostic), and recommendations for follow-up testing. Patients were included in the study if the chart documented any previous colonoscopy that was accompanied by a written communication of the findings from the endoscopist to the primary care physician. Patients with a history of colorectal cancer or inflammatory bowel disease and patients with a finding of high-risk or more than 10 adenomas were excluded from the study because appropriate intervals for repeat colonoscopy could not be determined without additional historical data.

Most of the colonoscopies were performed by gastroenterologists (86.2 percent); other endoscopists included general surgeons (5.4 percent), colorectal surgeons (3.3 percent), and primary care physicians (0.9 percent). In 33.5 percent of cases, the communication to the primary care physician did not provide guidance on follow-up testing. Of the remaining 743 patients in whom retesting was discussed, 663 were recommended to have a repeat colonoscopy. These recommendations were compared with intervals recommended in the current consensus guideline (see accompanying table).

Results: Overall, endoscopists' recommendations for the timing of repeat colonoscopy were only consistent with contemporaneous guidelines 39.2 percent of the time; most endoscopists recommended shorter intervals than those in the guidelines. Although current guidelines recommend a repeat colonoscopy 10 years after a normal examination or a finding of hyperplastic polyps, the average retesting intervals recommended by the endoscopists in the study were 7.8 and 5.8 years, respectively. Extrapolating these findings to the general U.S. population, the authors estimated that 2.85 million unnecessary colonoscopies would be performed over the next 10 years, leading to $3.4 billion in direct medical costs, 14,299 serious complications, and 142 deaths.

Conclusion: The authors conclude that endoscopists commonly recommend intervals for repeat colonoscopy that are substantially shorter than those in current guidelines. They also conclude that communication between endoscopists and primary care physicians is often limited by the absence of recommendations for retesting or explicit rationales for advocating shorter retesting intervals.

editor's note: Fewer than one half of eligible patients in the United States are up-to-date on colorectal cancer screening.¹ Therefore, family physicians have appropriately focused their efforts on overcoming barriers to screening,² trusting that subspecialists who perform colonoscopies are aware of relevant follow-up guidelines. Unfortunately, the study by Krist and colleagues shows that procedural expertise does not necessarily correlate with knowledge about how to use that procedure to provide the greatest benefit to patients. Because relying exclusively on endoscopists' recommendations could likely result in many unnecessary colonoscopies and increase the potential for serious harms, family physicians should be prepared to “question authority” when communication about retesting intervals is unclear or inconsistent with evidence-based practice.—k.l.