"The recommendations on this are all over the map, and that's because it's a close call," Welch said. "Different people look at the numbers and come to different conclusions. There is the potential to help some people and hurt some people."
In his own medical practice, Welch said, "I try to stay away from this one. If someone asks me, I explain the risks and the benefits."
Given the unknowns, it's entirely appropriate for a man to decide on having a PSA test without advice from a physician, Welch said.
But Dr. Judd Moul, director of the Duke University Prostate Center, after reading the new study, said, "This is not going to change my mind on the issue of screening." Moul recommends an initial PSA test at age 40, with follow-up if necessary.
"We now do have the potential for over-detection, but we still have 30,000 men a year dying from prostate cancer, and the best way to prevent deaths from prostate cancer is still screening," Moul said.
Both Moul and Welch agree that the PSA test is decidedly imperfect because it can't distinguish between the majority of prostate cancers that grow so slowly that they are no danger to a man's life and the fast-growing minority that are potentially fatal and require decisive treatment.
"Right now we don't have that magic biomarker," Moul said. "So I think it is more important to try to reduce the number of deaths than to worry about over-detection."
"I wouldn't expect one to come soon," Welch said, referring to a specific test for virulent prostate cancer. "So we will continue to find a lot of it, and a lot of it that doesn't matter."
Learn about prostate cancer from the U.S. National Cancer Institute.
SOURCES: H. Gilbert Welch, M.D., profe
All rights reserved