Researchers found that when they had 17 doctors view images taken during what's known as capsule endoscopy, the doctors caught fewer than half of the small intestine abnormalities on display.

The researchers say the findings send a message to doctors reading images from the capsule cameras: Slow down.

During capsule endoscopy, the patient swallows a pill-sized video camera that takes images of the small intestine. The images are transmitted to a small recording device the patient wears for about 8 hours. The camera itself is expelled in the stool in anywhere from 12 to 48 hours.

Capsule endoscopy is most often used to find the source of "obscure" bleeding in the digestive tract. That refers to bleeding that could not be diagnosed after a standard colonoscopy or an upper endoscopy, in which a scope is passed into the esophagus, stomach and upper part of the small intestine.

The problem is, the scopes used for colonoscopy and upper endoscopy cannot reach the middle of the small intestine -- which, despite its name, is actually the longest part of the digestive tract. If it were uncoiled, the small intestine would extend to about 14 feet.

So capsule endoscopy, which was approved in the U.S. in 2000, offers a way to see what's going on there.

But surprisingly little has been known about how doctors fare when reading the images from capsule endoscopy, according to Dr. Eric Goldberg, the senior researcher on the new study.

That's important because a single capsule endoscopy produces about 50,000 images for doctors to read, which makes the chances of missing something high.

"We're looking at 50,000 images in about 30 minutes to an hour-and-a-half," said Goldberg, a gastroenterologist and assistant professor at the University of Maryland School of Medicine in Baltimore. "So the question is, are we looking at them too quickly?"

For their study, published in the American Journal of Gastroenterology, Goldberg's team had 17 capsule endoscopists read 24 "clips," 18 of which showed small intestine abnormalities, such as blood vessel lesions that can be a source of bleeding.

Doctors can choose from a range of "reading modes" when looking at capsule endoscopy images. The modes vary based on how many images are displayed at once, and how long they are on view.

In this study, the doctors looked at images in the four most commonly used reading modes.

With three of the reading modes, Goldberg's team found, the doctors detected between 43 percent and 47 percent of abnormalities. With the fourth, the detection rate was only 26 percent.

"This was a real eye-opener," Goldberg told Reuters Health.

"I think the big message here," he added, "is that we need to be reading much more slowly."

There are other ways to view the mid portion of the small intestine. A longer scope known as an enteroscope can be threaded into the area. And scopes have the advantage of allowing doctors not only to view the digestive tract, but also to take tissue samples or treat abnormalities. A blood vessel lesion, for example, can be cauterized to stop the bleeding.

But the advantage of capsule endoscopy, Goldberg said, is that "it's a non-invasive, safe test."

And studies looking at "diagnostic yield" -- that is, how often a test produces an actual diagnosis -- have suggested that capsule endoscopy does it about 70 percent of the time. That's better than what has been seen with enteroscopes or with X-rays, which can catch only certain sources of bleeding, like tumors in the small intestine wall.

Capsule endoscopy, which costs around $1,000, is mainly used when the standard ways of finding GI bleeding fail. But it may also be used in certain other cases, such as to help diagnose Crohn's disease, an immune system disorder that damages the lining of the intestines.

The procedure is considered low-risk, but there is a chance of the capsule becoming lodged in the digestive tract. That's thought to happen about one percent of the time; in some cases, the capsule has to be removed with a scope or through surgery.

If a person has had capsule endoscopy with normal results but still has symptoms, Goldberg said his advice is to have a second doctor read the stored images.

At his center, Goldberg said, it's standard now for a second doctor to look at capsule endoscopy images, to help avoid misses.

A limit of the current study is that it included only 17 doctors, all from the Baltimore area. But Goldberg said he thinks the findings could apply more broadly; the endoscopists had a range of experience, and included doctors from university medical centers and community practices.

It's likely, Goldberg said, that doctors elsewhere would perform similarly.

SOURCE: bit.ly/xUnlR2 American Journal of Gastroenterology, online January 10, 2012.

$INS01; Line LNY Insave:- TI line name (Map report)

• A Permanent Economic Emergency