Suzanne Bull always half expected that she'd get cancer. After all, she lived in Marin County, California, where breast cancer rates are among the highest in the country. Still, she was determined to do whatever she could to protect herself. She ate right and exercised, and every year, she went into San Francisco to get a mammogram.

Last year, when Bull was 54, she got the news she'd been dreading. An ultrasensitive digital mammogram showed a suspicious spot on her left breast. A biopsy confirmed it was cancer. Fortunately, the surgeon told her, it had been caught early: She had ductal carcinoma in situ, or DCIS, which meant that the cancer was still confined to a single milk duct. And it might well stay there, he added, since DCIS generally doesn't become invasive. That all sounded great, Bull recalls, until the surgeon told her that there was no way to know whether her cancer would turn out to be the lazy, nonthreatening type of DCIS or the potentially invasive kind. She needed a lumpectomy, he told her, and should also consider undergoing radiation and taking the drug tamoxifen.

Bull agonized over the decision for two weeks but in the end went ahead with the lumpectomy and radiation. "I had to do everything I could to stop this disease," she says. With two clean mammograms behind her, Bull feels lucky. "I'm just glad I had access to digital mammography," she says. "It finds things so much earlier."

It's hard to believe, but some researchers wouldn't call Bull lucky at all. They say that yearly mammograms are not nearly as effective at reducing the risk of dying of breast cancer as most women think, and that mammography leads many women to get unnecessary treatment -- especially those diagnosed with DCIS. The problem is bigger than just mammography: They say the prostate-specific antigen (PSA) test may do men more harm than good if they don't already have symptoms of prostate cancer. And they have similarly grim things to say about other widely used cancer screening tests.

Their view stands in stark contrast to the message being put out by groups like the American Cancer Society and even the federal government, which say that finding and treating tumors as early as possible is the surest way to avoid a cancer death. But a growing group of scientific heretics -- published in highly respected medical journals, working at some of the most august institutions -- strongly believe that it's time to rethink our whole approach to cancer screening.

That's because screening tests pick up many small cancers that would never have caused any symptoms. "Screening for cancer means that tens of thousands of patients who never would have become sick are diagnosed with this disease," says H. Gilbert Welch, MD, codirector of the Outcomes Group at the Veterans Affairs Medical Center in White River Junction, Vermont, and a leading expert in cancer screening. "Once they're diagnosed, almost everybody gets treated -- and we know that treatment can cause harm." Tamoxifen for breast cancer can trigger life-threatening clots in the lungs, for instance. Surgery for prostate cancer leaves 60 percent of men unable to have an erection. For that matter, some of the screening tests themselves carry risks: Up to 5 out of every 1,000 people who get a colonoscopy have a serious complication, such as a colon perforation or major bleeding.

Most people diagnosed with cancer undoubtedly see these risks as the price they must pay to avoid dying of cancer. "The reality is not so simple," says Dr. Welch. Screening tests are very good at catching tumors that would never bother us, he notes, but they're actually pretty bad at catching the fastest-growing and most deadly cancers in time to cure them. The bottom line, says researcher Floyd Fowler, Jr., PhD, president of the Boston-based nonprofit Foundation for Informed Medical Decision Making: "Screening's power to cut your risk of dying has been wildly overinflated."

How Cancer Can Fool a Screening Test

The idea that getting tested for cancer might be useless or even harmful may strike you as completely wrongheaded. After all, smaller cancers are easier to cut out. They're also less likely to have metastasized, or spread to other parts of the body -- and metastasis is generally what makes cancer deadly. Sure, it's possible for a tumor to kill without metastasizing: A brain tumor, for example, can cause devastating harm when it grows big enough to squeeze healthy tissue inside the skull. But most cancers threaten life only after a few cells break free and travel through the bloodstream or lymph fluid to set up shop in another part of the body. Once that's happened, a surgeon can no longer cure a patient by removing the tumor. And even powerful chemotherapy drugs are often unable to kill every last errant cell.

Physicians used to think that a tumor needed to get to a certain size before it would spread. But that's not necessarily so, says Barnett S. Kramer, MD, associate director for disease prevention at the National Institutes of Health. "Some tumors spread extremely early," he says. They begin metastasizing when they consist of only a few million cells, which sounds like a lot but is smaller than the period at the end of this sentence -- too small to detect with most screening tests. By the time this kind of cancer is big enough to be seen on a mammogram or other test, it's already sent seeds to other parts of the body.

The flip side of this problem is that many screening tests do a great job at catching cancers that would never have caused problems and could simply have been left alone. This notion violates most of what we think we know about cancer, says Dr. Kramer, because most of what we know is based on the tumors that cause harm. If you think of all the different varieties of cancer as making up an iceberg, cancers that cause symptoms represent only the part of the berg above the waterline. For most of human history, these were the only tumors we knew anything about: the breast cancer that had grown big enough to feel, the lung cancer that was causing shortness of breath.

Screening allows us to look under the water, at the tumors that haven't yet become symptomatic. We assume they will eventually cause symptoms, but increasing evidence suggests that's not always the case. Evidence from autopsies, for instance: In one study, postmortem exams showed that nearly 9 percent of women of all ages who died of any cause other than breast cancer had undiagnosed DCIS. Among women from Denmark, where mammography is not as common as it is here, a whopping 39 percent of middle-aged women who died of other causes had undetected breast cancers. Similarly, says outcomes researcher Dr. Welch, a 1989 study found that 60 percent of men over age 60 have undetected prostate cancer -- yet only about 3 percent of deaths in men are due to prostate cancer.

So screening tests raise red flags about cancers destined to loll about quietly, causing no problems. But there's more. They also blare the alarm about cancers that would actually go away on their own -- because, in fact, some cancers simply disappear.

Brandon Connor, now age seven, was suspected of having cancer even before he was born. It had been a difficult pregnancy, and Brandon's mother, Kristin, then 35 and a lawyer in Atlanta, was undergoing regular ultrasounds. One of the tests picked up what looked like a tumor on Brandon's spine. Doctors made a tentative diagnosis of neuroblastoma, a nervous system cancer.

Neuroblastoma comes in two forms, one of which is deadly. But there was no way of knowing if Brandon's tumor was indeed a neuroblastoma, much less whether it was dangerous, without doing a biopsy, and its location made that risky. The Connors opted instead to keep a close watch to see if the cancer grew; the doctors said Brandon's tumor should regress within his first year if it was going to. It didn't, and by the time Brandon was two years old, he'd undergone more than a dozen MRI scans.

Finally, the doctors advised the Connors to go ahead with surgery. The day before the operation, though, the surgeon ordered one last imaging test. The neuroblastoma was gone. "We couldn't believe it," says his mother. Today, physicians know that many neuroblastomas regress on their own during infancy or early childhood.

"People kept telling us, 'Thank God they found it on the ultrasound,'" Kristin Connor says. Looking back on the years of worry, she adds, "In hindsight, I'd say it was more like a curse."

The Damage Screening Can Do

Forget the fact that unnecessary therapies for cancer are a tremendous drain on our health care budget, already strained to the breaking point. "Many oncologists would probably tell you that they've had patients who suffered serious side effects, even death, from treatment that they might not have needed," says William C. Black, MD, a professor of radiology at Dartmouth-Hitchcock Medical Center. No one intentionally prescribes unnecessary treatment, of course. But it's often difficult to know if a patient really needs to be treated, so the tendency is to be aggressive, just in case.

Treatment can exact a profound toll. Take the case of George Brown. At 75, Brown was still a practicing lawyer in Denver last year when he was diagnosed with prostate cancer. His doctor prescribed Lupron to block production of testosterone (which many prostate tumors need in order to grow). "I didn't realize that Lupron was chemical castration," says Brown. "I was extremely depressed. I was having hot and cold flashes. I cried at everything." Radiation therapy damaged his rectum and left him with little control of his bladder or bowels. He is now facing another round of a different testosterone-blocking drug.

Despite his troubles, Brown believes his care was lifesaving. And there's no way to know in any particular case. But the fact is that most men diagnosed with this cancer have invasive therapy, even though statistics say that many men could safely choose "watchful waiting": getting PSA tests to monitor the cancer and treating it only if it begins to grow rapidly.

Does Screening Save Lives?

For many people, even serious side effects like the ones Brown suffered would be worth putting up with if the treatment reduced their risk of dying of cancer. That's the point of getting screened, isn't it? Yet only one cancer screening test, the venerable Pap smear, has truly slashed the risk of death. Between 1955 and 1992, according to the American Cancer Society, Pap smears cut the death rate for cervical cancer by 74 percent, and deaths have continued to decline each year.

But no other test has had such a powerful effect. The PSA test has been widely used in the United States since the late 1980s, but it's not clear that it's had a big impact on the death rate for prostate cancer. Between 1975 and 2005, the latest year for which statistics are available, the death rate dropped from 31 per 100,000 men to 24.6. That's a real decline, but many experts doubt that PSA testing deserves all the credit -- especially given what happened during a "natural experiment" in Seattle and the state of Connecticut in the late 1980s.

Medicare patients in Seattle were five times more likely than those in Connecticut to get PSA testing between 1988 and 1990 and were also more likely to have surgery and radiation for prostate cancer. But when researchers followed up through 1997, they found the Seattle men were just as likely to die of prostate cancer.

"Prostate screening seems to make sense," says Nortin M. Hadler, MD, a professor of medicine at the University of North Carolina at Chapel Hill and the author of Worried Sick: A Prescription for Health in an Over-treated America. "If only it worked."

Mammograms also offer a smaller benefit than many patients -- and doctors -- assume. Mammography's effectiveness has been hotly debated, but a carefully conducted 2005 analysis suggests it cuts the risk of dying of breast cancer by 15 percent, says the NIH's Kramer. That means a 60-year-old who gets regular mammograms shaves her risk of dying of the disease in the next decade from 7 per 1,000 to 6 per 1,000

As for colonoscopy: It allows the doctor to remove polyps, growths that can turn into cancer. The best estimates suggest that colonoscopy can cut the risk of death from colon cancer by as much as 60 percent. (We don't know for sure if it reduces the risk of death, because those studies haven't been done.) Sixty percent sounds great, until you realize that the chances of dying of colon cancer aren't all that big to start with. The average woman has a 2.1 percent risk of dying of colorectal cancer. (So of all the things that can kill her, this will be the culprit about 2.1 percent of the time.) The average man's risk is a little higher, about 2.3 percent. Knocking a 2.3 percent risk down by 60 percent means it drops to 0.9 percent -- a benefit, yes, but not necessarily big enough to outweigh all other considerations.

To Screen or Not to Screen

The fact is, there's no single answer. It depends on many factors, including how old you are, what other diseases you have, and what you value most in terms of your health. Dennis Fryback, PhD, is a former member of the U.S. Preventive Services Task Force, a group of experts convened by the federal government to make recommendations about screening. The task force recommends colonoscopy every ten years for people between the ages of 50 and 75, yet the 61-year-old Fryback has concluded it does not make sense for him to get screened.

He came to that decision in part because he has no family history of colon cancer. If he did, his chances of getting it would increase, and so would the odds he'd benefit from the test. He also knows that getting the exam requires at least a day of taking laxatives to clean out the colon and then facing the possibility of a perforation from the procedure, a risk that goes up with age. He balanced the possible reduction in his chances of dying of colon cancer against his other health problems. He had a heart attack last year and suspects he will die of heart disease before a colon polyp has a chance to kill him.

Given his circumstances, Fryback figures, colonoscopy "is like an expensive lottery ticket. I might get some extra time, but chances are much better that I won't get anything. It's like paying, say, $5 to have a very long-shot chance at a few hundred dollars."

When looking at his odds, Fryback has an advantage: He's an expert in medical decision making. Most of us, of course, are much less familiar with medical statistics, but there are tools to help average patients come to a decision that's right for them. Called patient decision aids, these tools come in the form of brochures, videos, and Web-based interactive programs; some include interviews with cancer survivors and people considering getting screened, who discuss their own decisions. Patients can sometimes take them home to study at their own pace.

Decision aids aren't widely available yet, but some insurance companies and a handful of medical centers offer them. Suzanne Bull used a patient decision aid DVD before opting to undergo radiation treatment for her breast cancer. "Watching it was the best thing I did," she says.

Eventually, researchers and doctors hope, better screening tests will be able to distinguish between cancers that need to be treated and those that don't. But until then, many experts believe, the decision to get screened should rest on an individual's values and his or her ability to handle uncertainty. "We have come to fear dying from disease more than dying at the hands of overzealous doctors," says Dartmouth's Dr. Black. The fact is, both are risks when we get screened for cancer.