Screening cuts colon cancer deaths.

People whose stools are tested regularly for traces of blood are less likely to die of colorectal cancer than those who don't submit samples for testing, an 18-year Minnesota study finds.

Starting in the late 1970s, researchers enrolled 46,551 healthy volunteers, age 50 to 80, into three roughly equal groups to evaluate such screening. One group submitted fecal samples from three consecutive stools once a year, another group did so every other year, and the third control group submitted none. When a stool sample showed traces of blood, the volunteer underwent a colonoscopy--in which a doctor visually checks the colon for precancerous lesions or tumors. Patients were then treated accordingly, says study coauthor John H. Bond, a gastroenterologist at the University of Minnesota in Minneapolis and the Minneapolis Veterans Affairs Medical Center.

Overall death rates among the three groups were similar, but 177 people in the control group died of colorectal cancer compared with 148 of those whose stools were analyzed biennially and 121 in the annual group, the researchers report in the March 3 JOURNAL of the NATIONAL CANCER INSTITUTE.

Those getting annual stool tests were diagnosed with about half as many advanced colorectal cancers as the controls.

"We didn't know if screening would be helpful," Bond says. "This reconfirms the validity of this approach."