HEALTH MATTERS : COLON CANCER PREVENTION / THE NEW YORK TIMES EDITORIAL ( RECOMMENDED READING )

August 16, 2010

Colon Cancer Prevention

A provocative new analysis suggests that the colonoscopies many people get to detect and eliminate colon cancer before it can kill them may be no more effective than a less invasive (and cheaper) procedure would be. The analysis, more thought-provoking than conclusive, illustrates the kind of evidence scientists will need to develop as leaders of the American health care system struggle to define which treatments are more effective — and more cost-effective — than others.

Over the past two decades, colonoscopies have become the preferred method for trying to detect and remove cancers and precancerous polyps in the colon. They have largely supplanted their main rival, sigmoidoscopies. The sigmoidoscopy looks at only part of the colon; the colonoscopy inspects the whole length. For many patients and doctors, it seems only common sense that the more comprehensive exam is the safest choice.

Yet there is scant evidence to support that belief. A recent commentary by two Columbia University experts in the Journal of the American Medical Association suggested that colonoscopies may be no better at preventing cancer deaths than sigmoidoscopies, which are easier on the patient and have fewer complications.

Three studies conducted abroad over the past two years found that colonoscopies did indeed reduce colorectal cancer mortality or incidence — by essentially the same amount that sigmoidoscopies did in separate studies. The colonoscopies achieved their effect by eliminating only growths found in the part of the colon that sigmoidoscopies can also reach and identify, not by eliminating growths deeper inside.

Experts have been scrambling for explanations. Their theories include: a failure by many doctors abroad, who are often less experienced than American specialists, to inspect the entire colon; a failure by many patients to cleanse their bowels sufficiently; and molecular or anatomical differences that might make growths in the distant reaches of the colon more difficult to spot.

It would seem premature for patients to abandon colonoscopies based on these studies. The critical need now is for researchers to determine how well colonoscopies perform in the United States and to identify how to improve their effectiveness. Meanwhile, those who have shunned colonoscopies as too arduous may prefer a sigmoidoscopy. Either technique can substantially reduce the risk of colon cancer.