"We know the test has found many, many, many, many more cancers," said Dr. Peter Albertsen, chief of the division of urology at the University of Connecticut Health Center and a professor of surgery at UConn. "Where the controversy comes in, however, is that the subsequent decline in mortality — in other words, how many men die from this disease — is relatively modest."

Experts believe the disparity in part reflects the fact that while screening can identify aggressive cancers, it also finds those that would never have progressed or progressed so slowly they never would have killed a man. That means some men receive treatment — which can lead to urinary, sexual or bowel problems — for cancers that might not have caused problems if not detected.

A randomized study released last year, conducted in Europe, found that to avoid one prostate cancer death, 1,410 men would have to be screened and 48 men would need to be treated.

"If you're that one guy, you think this is wonderful, and every man who undergoes surgery assumes they are that one guy," Albertsen said. "The problem is the public health folks are saying, 'Wait a minute, you can't all be that one guy because 47 of you are not that one guy.'"

Tests are not sophisticated enough to tell who might fall into which category, he said.

There is no question that screening and treatment has had an impact, Albertsen said; the question is how much.

In its guidelines, the cancer society emphasized the importance of telling patients about the benefits and drawbacks of screening and reaching an informed decision about whether to proceed.

"The abundance of data now available regarding prostate cancer early detection supports neither a clear mandate to screen nor a proscription against screening," the American Cancer Society Prostate Cancer Advisory Committee wrote in the new guidelines.

Doubts About Exam

But the group did raise doubts about the benefits of the digital rectal exam, citing a lack of evidence about its benefits and concerns that it keeps men from getting screened.

There have only been three case control studies that examined digital rectal exams alone, without the PSA, and only one found a statistically significant benefit, according to the cancer society. The sensitivity of the exam depends on the individual performing it, the group wrote.

Even under optimal circumstances, the cancer society committee wrote, "the added value of performing a periodic DRE is likely to be quite low."

Still, the group noted that the digital rectal exam can detect some high-grade cancers in men whose PSA test results would not raise concern. It also suggested that the exam may be helpful in some circumstances, such as in evaluating men whose PSA test results fall into a range where a biopsy may or may not be warranted.

The exam is not particularly comfortable, and for some men, particularly men of color, it may be seen as culturally inappropriate, said Salner, who was a member of the committee that wrote the guidelines.

"It has implications of them being kind of violated in some ways," Salner said. "Particularly if another man is doing it."

Which makes it particularly tricky to reach the men at the greatest risk from prostate cancer, Salner said. African-American men have higher rates of both getting prostate cancer and dying from it.

"That's the very population that we really need to try and screen," he said.

In more than five years of talking about prostate cancer with clients at his shop, Shallimar Barber and Beauty Salon, Davis has found the men to be receptive. Some tell him about relatives who had prostate cancer. One mentioned a 37-year-old nephew who died form it.

Davis took a course at Hartford Hospital and participated in a program aimed at getting barbers to talk to men about the disease. He holds annual screening events at his salon in conjunction with Hartford Hospital and St. Francis Hospital and Medical Center.

Some men say they fear what the exam might find and what treatment might cost if it does find cancer. But the biggest concern, he said, is what he calls the "invasiveness" of the test.