Doctor Draws Fire for Mammography Overdiagnosis Stance
Dr. Gil Welch (Valley News - Ryan Dorgan) Purchase photo reprints »
Dr. Gil Welch, a Dartmouth researcher and author of Overdiagnosed, speaks with the Valley News on Wednesday at his office in Lebanon.(Valley News - Ryan Dorgan) Purchase photo reprints »
Lebanon — Some of the emails Dr. Gilbert Welch has received in recent weeks have been aggressive, even offensive. Turning to his computer last Wednesday, Welch opened a message he received that morning and read it aloud.
“ ‘You are consciously and without conscience working hard to let more women die of breast cancer,’ ” read Welch, a professor of medicine at the Dartmouth Institute for Health Policy and Clinical Practice. “That’s from a doctor.”
Welch pushed away from the computer, his boots pressing against the floor to spin him around in his chair. His eyebrows raised and mouth opened into an expression of disbelief.
That email was an extreme example among the hundreds of messages, some positive and others just the opposite, he’s received since his latest study of mammography was published in The New England Journal of Medicine. The study, which he co-wrote with Dr. Archie Bleyer of Oregon Health and Science University, came out on Thanksgiving Day. The journal article and an accompanying opinion piece he wrote for The New York Times have elicited a strong and divided public reaction from mammographers, oncologists, primary care practitioners, policy makers and patients from around the nation.
In the articles, Welch extended an argument that he has been making for nearly 20 years — that while tests can lead to the early detection of some diseases, there is also a downside known as “overdiagnosis.” Tests are very good at finding something “wrong” with otherwise healthy patients.
In fact, Welch argues, the tests are often too good, identifying abnormalities that pose a minimal health threat and often leading to treatments that patients don’t need. That leads to unnecessary suffering for patients who get a mastectomy or chemotherapy to “cure” a cancer that was never going to cause them problems.
Most cancer specialists do not agree with him, and a few have been especially vehement in their response.
“This is simply malicious nonsense,” Dr. Daniel B. Kopans, a senior breast imager at Massachusetts General Hospital in Boston, told The Los Angeles Times. “It is time to stop blaming mammography screening for ‘over-diagnosis’ and ‘over-treatment’ in an effort to deny women access to screening.”
Days after Welch’s op-ed appeared in The New York Times, nine doctors and a nurse at Memorial Sloan-Kettering Cancer Center wrote a letter to the editor to express how “horrified” they were by Welch’s “repeated attacks on screening mammograms.” They attacked his analysis as flawed and wondered if Welch would “have us return to the era when tumors were discovered only when they were large enough to be felt, meaning more disfiguring surgery, prolonged chemotherapy and lower cure rates?”
Others have been more measured. Groups such as Susan G. Komen for the Cure and the American Cancer Society have acknowledged that screening mammography is not perfect science, but say it nevertheless saves lives.
Criticism of Welch’s study has come from near and far. Physicians practicing at Dartmouth-Hitchcock Medical Center across the street from Welch’s office in Centerra Park are troubled by his conclusions.
“My first reaction was concern and dismay,” said Steven Poplack, director of breast imaging at Norris Cotton Cancer Center.
“I’m concerned that this article will lead patients to not have screening mammography and that will result in a loss of life from breast cancer, lives that otherwise would’ve been saved.”
But Welch has also been praised, both publicly and in personal messages sent directly to him, for offering an alternative viewpoint on a practice that has become firmly entrenched in the way medical professionals approach cancer diagnosis.
His longtime colleague at Dartmouth, radiologist Bill Black, said medical professionals who remain open to Welch’s arguments are coming around to his ideas, however counter-cultural they might be.
“He’s extremely thoughtful and balanced,” Black said. “And he’s a very, very open communicator. He doesn’t have any axe to grind. He’s just trying to explain the situation as he sees it.”
That some of Welch’s views, while still in the minority, are even being acknowledged by groups such as the American Cancer Society and published in The New England Journal of Medicine signal a slow shifting of cultural norms. Some cancer groups, patients and doctors are beginning to question ideas that had once been sacrosanct.
“Dr. Welch’s points are so well-taken,” Judy Norsigian, co-author of Our Bodies, Ourselves, said in a recent discussion of Welch’s study on WBUR, the National Public Radio affiliate in Boston. “We need to bring balance to this discussion.”
Against the Advice
Breast cancer is one of the most common forms of cancer in women, with more than a quarter million new cases expected this year in the United States and resulting in deaths of nearly 40,000 women, according to the National Cancer Institute. It is, arguably, the most important cancer that nonsmoking women should care about, Welch said.
The color pink has become synonymous with breast cancer awareness campaigns and for years, advocacy groups and physicians have promoted regular screening mammograms — in which x-ray pictures are taken of the breast — as a way to detect early stage cancer and treat it.
“We know it does save lives,” said Peter Ames, vice president of health initiatives at the American Cancer Society.
Cancer groups vary in their recommendations. The American Cancer Society says women over the age of 40 should get screened annually. The Centers for Disease Control and Prevention recommends screening mammograms every two years for women between the ages of 50 and 74, adding that women 40 to 49 should talk to their doctors.
Nevertheless, there is a widely held belief that regular screening is a good thing for women of a certain age, bringing far more benefits than potential harms. The earlier cancer is discovered then the faster it can be addressed, proponents reason. Thus, less aggressive treatment will become necessary later in life.
That’s where Welch disagrees.
“(Early diagnosis) is intuitively a very appealing idea,” he said. “It’s sort of hard to argue against. But it can be extremely misleading. In fact, it can be a bad thing to do.”
In their study, Welch and Bleyer concluded that 1.3 million women over the past 30 years had been diagnosed with cancer that never posed a threat to their health and that roughly one-third of breast cancer patients, or 70,000 women per year, got treatment they did not need.
Critics have attacked the study’s methods and pointed to previous studies, randomized trials they say are based on much better science, that suggest screening mammography does indeed lead to lower breast cancer death rates.
Welch acknowledges that the death rate from breast cancer has fallen, but says the primary explanation is because of better treatments, not early diagnosis. Ironically, Welch said, as treatment improves the benefit of screening diminishes.
Welch doesn’t argue against all tests. His concerns do not apply to diagnostic mammograms — the test done in women who have become aware of new breast lump. It is an important test to figure out whether or not the lump is something to worry about, he said.
Screening mammograms are different, done as a matter of routine for otherwise healthy people.
It’s the healthy people he’s concerned about.
e_SDLqThis is our typical thinking, the (test) that finds the most is the best. And that’s led us into a cycle of more overdiagnosis, not less,” he said. “And all I’m saying is we’ve got to pay attention to that balance. We’ve always focused on the one person out of a thousand we can help, and all I’m asking is what happens to the other 999?”
A History of ‘Overdiagnosis’
This is not the first time that Welch has tackled the issue of overdiagnosis.
Last year, he published a book, called Overdiagnosed, that explored the subject. His previous book published in 2004, Should I Be Tested for Cancer?, delved into many of the same themes.
But Welch’s interest in the topic began two decades ago when he came to Dartmouth and met Black. At the time, there was powerful new imaging technology coming out, such as CT and MRI machines, that were being used to detect cancers that had never before been able to be seen.
Welch and Black began questioning whether there was some downside to these advancements.
In 1993, they co-authored a study in The New England Journal of Medicine that said advances in diagnostic imaging can actually confuse doctors when making decisions about the severity of a disease and in figuring out how well treatment works. They also argued that increased use of this technology would promote a vicious cycle, in which doctors would be tempted to provide more treatment with little or no benefit for the patient.
To this day, he believes it is one of his most important articles.
It garnered some negative reaction at the time, Welch said, but nothing like what he’s received lately.
“Here (in the 1993 paper) we’re outlining a general problem about diagnostic imaging. And there’s a couple throw aways on mammography,” he said. “A paper like I just had is right up front on screening mammography, the most high stakes test there is. ... It’s just a very charged issue.”
It’s more than emotionally charged. Mammogram screenings have become an accepted part of political and popular culture.
The Affordable Care Act included a requirement that insurers cover mammograms for women over the age of 40.
And some primary care physicians require their patients to have them done, Welch said.
As an example of how strong the messages on mammogram screening have been, Welch pointed to a poster from the American Cancer Society from the 1980s. It showed a woman’s face staring back at the viewer. The typed message said “If you haven’t had a mammogram, you need more than your breasts examined.”
“That’s pretty strong,” Welch said. “The subliminal message there is that if you’re not getting one, you’re crazy.”
A Change in Practice?
Welch isn’t focused on changing the politics of this issue. There are subtleties involved that he doesn’t expect politicians to take the lead on.
Instead, he’s concerned with shifting medical culture to a point where screening mammography isn’t seen as a public health imperative, but as a choice.
He wants physicians to discuss the pros and cons of the test with patients. His study doesn’t resolve the question over whether women should get screened for breast cancer. But it does suggest there’s more than one right answer, he said.
There has been some shift in this direction, particularly in the growing acknowledgment that mammography is imperfect.
“We agree that mammography has flaws,” said Andrea Rader, a spokeswoman for Susan G. Komen for the Cure, “but it’s the most widely available screening methodology available today.”
Even the American Cancer Society, which once suggested women were crazy to not get a mammogram, has softened its tone.
“The Society recognizes that overdiagnosis is a matter deserving of attention,” said Dr. Len Lichtenfeld, deputy chief medical officer for the group, in a statement released in response to Welch’s recent study. “There is little question that over time we will continue to refine and improve our ability to determine which women are more or less likely to benefit from different approaches to screening and treatment.”
When he saw Lichtenfeld admit that overdiagnosis was an issue, Welch said he could barely believe what he was reading.
e_SDLqUn-be-liev-a-ble,” he said, extending every syllable for dramatic effect. “That’s a huge change. To me, that’s a sea change. The American Cancer Society. Overdiagnosis is a problem. We acknowledge. Did you read that statement?”
An acknowledgement that overdiagnosis is a problem, however, is a long way from signing onto Welch’s notions of the risks of screening mammography for women. Indeed, most advocacy groups and physicians aren’t ready to agree with Welch on that issue.
Popular opinion among practitioners is that screening mammograms are still important, partly because early detection allows for less severe treatments.
“If you’re unlucky enough to develop breast cancer, it’s much nicer to be able to be treated with a lumpectomy and without chemotherapy than with needing to have radical surgery and to have chemotherapy to try to effectively treat your cancer,” said Dr. Richard Barth, section chief of general surgery at Dartmouth-Hitchcock Medical Center.
Barth questions Welch’s findings about the amount of overdiagnosis that occurs. So do plenty of other cancer specialists outside Dartmouth-Hitchcock.
Rachel Freedman, an oncologist at the Dana-Farber Cancer Institute in Boston, said she’d read Welch’s study but wasn’t planning to make any immediate changes in her practice as a result.
“I don’t think this changes the overall picture,” she said. “I don’t think we can say for certain that this is correct.”
Nevertheless, she and many other practitioners said it was a useful addition to the larger discussion around mammography.
Megan O’Brien, a family nurse practitioner at Gifford Medical Center, said she planned to wait and “file (the report) as there has been a question on the horizon” but was not prepared to change her approach to care.
She’s not an absolutist on mammography screenings, anyway. As a primary care practitioner, she said it’s not her job to dictate tests or certain treatments to her patients.
Those decisions must be made with someone’s personal health goals in mind.
“I don’t think my job as a primary care provider is to say yes or no, but to help patients negotiate the confusing literature and make an informed decision,” O’Brien said.
She drew a parallel between Welch’s questioning of screening for breast cancer and the shift in how medical professionals have come to view prostate cancer.
There has been a gradual change of attitude over the years about the value of treating prostate cancer. Many men diagnosed with prostate cancer are not likely to die from it and the U.S. Preventive Services Task Force recommends against a formerly common type of screening, the prostate specific antigen test, for men who don’t have symptoms of prostate cancer.
But the idea of not looking for certain types of cancer has been a difficult mental switch for medical professionals to make.
“We’re sort of all getting comfortable with not screening,” O’Brien said. “There’s going to be a lapse while we maybe change practices and see how we’re doing.”
Welch isn’t advocating a wholesale abandonment of mammography screening. He said he simply wants patients and physicians to question whether it is really needed. In some cases, women may decide that it is necessary for their peace of mind.
A difficult challenge in changing attitudes, he said, is the prevalence of people with personal tales of surviving with cancer.
It could be someone who has a friend, or wife, husband or child with cancer. It could be the patients themselves.
They might never have died from their cancer, but once diagnosed, it was treated. That likely put them through a difficult ordeal, filled with more tests, perhaps surgery, chemotherapy, hair loss and physical pain. Once someone goes through that experience, they understandably want to believe they’ve benefited from the process rather than been harmed.
Welch winces a little when he considers the implications.
“Look at all these survivors. Everybody’s living five years, so on and so forth, and that feeds back to more intensive screening,” he said. “That’s sort of a popularity paradox of screening. If we’re not careful, it drives us to do more and more and more overdiagnosis. I think that’s a problem. I think that’s a problem worth being honest about.”
Chris Fleisher can be reached at 603-727-3229 or firstname.lastname@example.org.