Prostate cancer offers a good example. According to the U.S. Preventive Services Task Force (USPSTF), "A large majority of the men who are being treated would do well without treatment." Yet in the U.S., 90% of these men opt for treatment. “We have a culture of treating cancers aggressively, but we know all those men don't need to be treated," Kramer says. Another example is the most feared form of skin cancer: melanoma. Melanoma rates have gone up since the late 1980s. Most of the increase is in early cancers detected as skin-exam screening became more common. But late-stage melanoma cases didn't decline, Kramer says. Neither did the death rate. Brawley agrees. "I cannot quote a study showing that melanoma screening definitely saves lives," he says. "We cure some melanomas that don't need to be cured. “Nobody wants to live with cancer. Nobody wants to be over diagnosed or over treated. It's just that we want -- need -- answers that medical science does not yet have. "What we really need is a 21st century definition of cancer so we can move away from 20th century screening and diagnosis using a 1840s definition of cancer," Brawley says. There are limits to science, says Susan G. Fisher, PhD, professor and chair of public health sciences at the University of Rochester, N.Y. "People are uncomfortable and concerned because they think we are recommending less screening in some groups," she says. "The message for the public is that science is hard. As we get more and more information we get smarter about our advice. The most recent evidence says that in groups at low risk, we are creating more problems than benefits with early screening." The USPSTF recommends routine screening -- that is, for people at normal risk -- for only three cancers: