Colon cancer kills more people each year in this country than automobile accidents. Most drivers now accept that seatbelts save lives, but it has been harder to get them to Ïbuckle-upÓ for colonoscopy to reduce deaths from colon cancer. Less than 50% of those eligible have the procedure done. Reasons are varied, but the need for vigorous bowel preparation, intravenous sedation and a day lost from work are often invoked.
Recent advances in CT technology, as published in the New England Journal of Medicine, raise good questions as to whether virtual colonoscopy (VC) can safely and effectively increase the numbers of those being screened. What is new
is the software that allows three-dimensional Ïfly-throughsÓ of the colon (). This results in rather striking increases in sensitivity and specificity compared to previous studies of this technology. For larger polyps, VC compares favorably to the traditional gold standard of optical colonoscopy.
Why not then move immediately to this technology with all available speed? There are some concerns in the details.
Most drivers now accept that seatbelts save lives, but it has been harder to get them to "buckle-up" for colonoscopy to reduce deaths from colon cancer.
First, the prep still exists. For the scanner to effectively differentiate stool from polyp, most of the stool must be removed. Any remaining stool is tagged with oral contrast to permit electronic cleansing during post-processing. If a polyp is found, optical colonoscopy is required to remove it, often requiring a second prep. A lot of energy is being expended by GE Medical and others to solve the Ïprep problem,Ó but this solution is likely four to five years away. Of all of the barriers to accepting VC as a replacement for optical colonoscopy, this one is probably the Ïdeal breaker.Ó If solved, the balance could shift toward VC as the initial screening step.
Secondly, we need to see these results duplicated by others. The New England Journal of Medicine study demonstrated consistent readings and a short learning curve among different radiologists using the same software for VC interpretation. This suggests that if the software were widely available, community hospitals (where most of the screening in this country is done) could gain acceptable proficiency relatively quickly.