However, there is now clear evidence that colonoscopy by surgeons is less effective and less cost effective on average than is colonoscopy by gastroenterologists. First, all studies of interval cancers organized by specialty show that the patients of general surgeons and primary care physicians who do colonoscopy are more likely to develop interval cancers (Gastroenterology 1997;112:17-23; 2010;139:1128-37; 2011;140:65-72; Clin. Gastroenterol. Hepatol. 2010;8:275-9). Surgeons are also more likely to recommend colonoscopies at shorter-than-recommended intervals (Ann. Intern. Med. 2004;141:264-71; CA Cancer J. Clin. 2006;56:143-59) and are more likely to utilize anesthesiologists to sedate patients (Clin. Gastroenterol. Hepatol. 2012;10:58-64). The evidence indicates that colonoscopy that is performed by gastroenterologists is more likely to detect cancer, is less likely to be overused, and is less likely to utilize expensive forms of sedation.
Why is average performance better by gastroenterologists than by general surgeons? There are a number of possible reasons, including differences in personality. Although we don’t know what the reasons for the differences between specialties are, there is of course concern that it relates to differences in training. Perhaps this comes down to the fact that gastroenterology trainees do more endoscopic procedures than general surgeon trainees do. Or, it could be that gastroenterologists receive more instruction in withdrawal technique or lesion recognition skills.
Many practicing endoscopists are under the impression that precancerous lesions in the colon are obvious, when the reality is that certain lesions – such as the serrated lesions and the flat and depressed lesions – can be remarkably subtle. Are surgeons being trained to recognize the full spectrum of precancerous lesions?
With respect to the numbers of procedures required in training, there is evidence that it takes far more than 50 colonoscopies for gastroenterology trainees to reach the point at which they are considered competent. However, in my opinion, we should worry less about the number of cases trainees perform and more about whether or not they can achieve quality targets for colonoscopy and endoscopy by the time they have completed training.
The two targets that are most important for colonoscopy are the cecal intubation rate and the adenoma detection rate (Gastrointest. Endosc. 2006;63[suppl]:S16-28). I doubt that trainees in either surgery or gastroenterology are currently getting their adenoma detection rates measured on a consistent basis.
Many practicing endoscopists are still not making these measurements, though the quality mandate is rapidly becoming widely accepted within gastroenterology. If we made quality measurements during training, our trainees would learn the importance of measuring quality on an ongoing basis once they become practicing endoscopists.
I understand that surgeons are the only colonoscopists available in some areas of the United States, especially in rural communities and I do not think surgeons are going to stop doing colonoscopy on any widespread basis. However, general surgeons can’t escape the consistent evidence that their average colonoscopy quality is lower than that of gastroenterologists. We should not accept low quality when it’s been proven to result in bad outcomes.
Dr. Rex is Chancellor’s Professor and a distinguished professor of medicine at Indiana University, Indianapolis, as well as director of endoscopy at Indiana University Hospital.