Michelle Fay Cortez, "Colon-Cancer Test Misses 95% of Cases, Study
Shows", Bloomberg, January 18, 2005,

A single fecal blood test that doctors often use to screen for colon
cancer misses 95 percent of malignancies and lesions likely to become
cancerous, according to a study in the Jan. 18 Annals of Internal

Even when the tests are positive, about one in four doctors don't
follow up with a colonoscopy, considered the benchmark test for finding
colon cancer, a second study found. An at-home version of the fecal
blood test that involves taking six separate samples spotted less than
a quarter of cancers.

``These two studies discredit the office-based single-sample screening
test for occult blood while simultaneously showing that it is common
practice,'' wrote Harold C. Sox, editor of the journal, in an
editorial, calling the results ``shocking.'' ``Taken together, they
send a strong message to primary-care physicians to re-examine their
colorectal cancer screening practices.''

Accurate tests for colon cancer are essential because it is curable
when caught early. About 56,000 Americans a year die from colon cancer,
making it the second leading cause of cancer death in men and women,
and 106,000 are diagnosed with it annually.

Researchers gave 2,665 healthy men a single blood test in their
doctor's office and a six-sample at-home test, followed later by a
colonoscopy. Of the 284 people with colorectal cancer or large polyps
likely become cancer, the single office test spotted the disease 4.9
percent of the time and the six-sample home test was positive in 23.9
percent of cases.

The U.S. Department of Veterans Affairs funded the study, led by Judith
F. Collins from the Portland Veterans Affairs Medical Center. Thirteen
VA medical centers nationwide participated.

Separate surveys of 1,147 primary-care doctors and 11,365 adults aged
50 and older found about 30 percent of colon cancer screening was done
using only the test performed in the doctor's office. Ninety percent of
doctors said they use the office test at least once a month, according
to the study led by Marion R. Nadel, from the U.S. Centers for Disease
Control and Prevention.

Doctors ``may be concerned that the patient will not comply with the
home-sampling routine, reasoning that it's better to get one sample
than none at all,'' Sox wrote. ``It is time to break this habit.''