How We Do Harm (37 page)

“You can’t possibly remember me,” Kay started.
Indeed, that was the case, Pat was a leader of the antiwar movement and Kay, a freshman.
Even now, in middle age, Pat was as conspicuous as she was on Bennington campus in 1969.
At the board meeting, Kay remembers Pat wearing glasses with a superbright aquamarine frame.
Pat was clearly not about to let metastatic disease get her down.
She would fight it the way she’d protested the war.
Now, she was building a grassroots network for the coalition—the first of its kind in cancer—and Kay was there to represent a Baltimore patient group.

That Kay was a top-notch epidemiologist had not yet seemed relevant.
Dickersin was the first researcher to describe the publication bias—the tendency to publish positive studies and not publish negative studies.
Though she wouldn’t understand this until later, Dickersin was at the foundation of
two
important institutions of evidence-based medicine.
In her day job, she was working on the Cochrane Collaboration, a worldwide collective of statisticians who pool data from existing sources for meta-analyses.

*

AT
one board meeting in 1993, Visco was startled by something Dickersin said about mammography.

“Wait a minute, I was thirty-nine, a mammogram diagnosed me, do you mean it didn’t do me good?”
Visco asked.

“Sorry, you might have found it the next day yourself,” Dickersin said.
Then she launched into an explanation of standards of proof for efficacy of screening.
Screening young women before forty would do more harm than good, she said, introducing public-health concepts.
You have to think of the denominator—the entire population—and consider the numbers of women you might harm and the costs you would incur in order to save one life.

Dickersin was thirty-four at the time of her diagnosis.
No one in his or her right mind would have recommended her to get a mammogram at that age.
Indeed, even at age forty, there is room for disagreement.
The benefit increases as women age and the harms diminish.

At the same meeting, the group sparred over high-dose chemotherapy with bone marrow transplantation.
Several women in the room raised questions about access to the procedure.
Some insurers were balking at paying, causing patients to sue for access.

“There is not a shred of evidence to support this,” said Love.
Any group started by Love would have to be about science.

At that time—as now—debate over mammography focused on women between the ages of forty and forty-nine.
Screening for women under forty would do more harm than good, and screening everyone over fifty saves enough lives to make the procedure basically noncontroversial.

In 2009, the US Preventive Services Task Force issued what I think was a poorly worded statement that said that the task force didn’t recommend routine screening of women in their forties.
The statement added that some women in this age group who are at high risk for breast cancer as well as women who are very concerned about the disease might still want to get screened and should be allowed to do so.

This triggered outrage among ardent supporters of mammography.
Having looked at the data, I favor mammography for women in their forties, but I am concerned that mammography has serious shortcomings, and that its power to prevent deaths is often overestimated.
The limitations of mammography are rarely discussed.
That’s unacceptable.
Women should be told about the known benefits of mammography as well as its limitations.

Breast cancer mortality rates in the United States have declined by nearly one-third over the past twenty years.
This is to say that a woman’s risk of death from breast cancer today is over 30 percent below where it was in 1990.
This decline is due to three factors, including mammography.
Improvements in treatment are another factor.
Women finding breast masses themselves and getting them evaluated is the third.
Twenty years ago, mammography likely contributed more to the drop in mortality than it does today, but there is still a contribution.
I think of these three factors the way I think of legs on a three-legged stool.
All are important.

Many people are lost in this national debate because it turns on statistics, which are hard for even a statistician to comprehend.
Interestingly, almost all experts who have weighed in on this controversy estimate that mammography among women age forty to forty-nine saves lives.

Here is what the controversy is about: The task force estimates that mammography decreases risk of death by 15 percent.
The most vociferous screening advocates have done studies suggesting it reduces risk of death by 29 percent.

What do these percentages mean?

There are 23.3 million women aged forty to forty-nine in the United States.
Most experts agree that screening all these American women in one year would result in 156,000 callbacks for further evaluation, and ultimately 79,000 would get biopsies.
Approximately 32,000 of these women would be diagnosed with breast cancer that year.
Eventually, about 8,000 to 9,000 of the 32,000 would die of breast cancer.

Essentially, the task force said that screening 23.3 million women would result in treatment that would save about 1,200 lives.
Using estimates of the proponents of mammography, screening could save approximately 2,000 lives.
The two groups actually see the data similarly.

Of the 32,000 women who are diagnosed, 23,000 to 24,000 will do well as a result of treatment.
Their disease will be arrested or cured.
These are women who are often diagnosed through mammographic screening, which is then credited with making the diagnosis.
However, in a world without mammography, most of these women could have detected their tumors at an early, curable stage by simply being aware of what their breasts should feel like and palpating an abnormality.

Breast awareness isn’t the same as the monthly breast self-examination.
Nearly two decades ago, many experts started to realize that teaching a woman to perform a dedicated self-examination of the breasts didn’t save lives, but did increase anxiety and the number of breast biopsies performed.
In lieu of the monthly self-examination, experts have been encouraging women to be aware of their breasts while bathing and dressing.
This is a cursory examination of the breasts daily.

Uncertainty about the quality of mammograms performed in the United States further blurs the picture.
Everyone ignores that a large fraction—perhaps more than half—of women in the United States get less-than-optimal-quality mammography.
More often than not, the mammographer does not have access to previous mammograms.
We need a better test, but satisfaction with and the belief in mammography causes complacency.
What we need is more support for research aimed at finding a better test.

As we argue over screening in the age group where we can at best save two thousand lives per year, we seem to ignore that 40 percent of women in their fifties and sixties don’t get mammography.
This represents more than five thousand lives per year that could be saved through simply getting people adequate care.

Chapter 25

Project LEAD

IN PART
because of her expertise—but also because she lived close to Washington—Dickersin was frequently called on to show up on Capitol Hill.
In the early days of the coalition, only one other member of the board—Love—understood science.
Alas, she was chronically overscheduled.
Dickersin wanted to spread out the responsibility of talking with congressional staff members.

“I am going to suggest something,” Dickersin said at one of the early board meetings.
“We’ll do a teach-in, just like we did in the Vietnam era.”

Everyone at the table could be presumed to be a Vietnam protester, including Kay’s teach-in instructor from the 1969 moratorium, who sat across from her at the table.

Visco remembers that discussion.
“Kay Dickersin said, ‘If we want to influence NIH, if we want to influence science, all of us need to know what we are talking about.’
And that certainly resonated with me.
When you are a trial lawyer, you don’t walk into that courtroom, you don’t walk into any argument or deposition, unless you know everything.

“You have to know the strengths, the weaknesses, you have to understand it all.
I knew when I was practicing law that if my client had a business that made windows, I had to learn about that.
Now I had to learn science.”

The first teach-in, conducted in conjunction with a board meeting, was on basic science.
After the second teach-in, on epidemiology and biostatistics, the board was asked to hear a presentation by a drug-company executive.
The slides included “median survival time” on the drug in question, compared with placebo.

“Median survival means that this is how long patients lived on average,” he said.

One of the women who had just gone through a daylong course on epi-bio raised her hand.
“No, it doesn’t,” she said.
“Median is not the same as mean.”

Indeed, median survival is the point where half the patients have died and the other half live longer.
Mean is the same as average.

The drug-company executive acknowledged the screwup—how can you not?—then showed a slide with two curves separating.
He noted that some patients were found ineligible and were excluded from analysis.

One of the members raised her hand.
“Wait a minute, you mean you didn’t perform intention-to-treat analysis?
You are introducing a bias.”
Indeed, excluding patients after they have been assigned to groups potentially allows you to cherry-pick and is therefore verboten in good clinical trials.

“They really got it,” said Dickersin.
“It was so cool for me as a teacher.”

At the end, Pat Barr turned to Kay Dickersin, saying, “We must take this on the road.”
Kay agreed.

*

THEY
did, ultimately developing a core curriculum and a group of teachers—all of them top-notch scientists—who go on the road with the project.
The first session was held in 1995, and starting in 1998, they have been held four times a year.
The curriculum has changed over time, sometimes to reflect new science, and sometimes to add new areas of emphasis.

Teachers don’t get to bring their own slides; the slides are provided.
Uttering a sentence like “I tell my patients…” will bring on a talking-to.
Stick to the evidence, Doctor, thank you very much.

The course is designed to push the patients outside their experience, beyond their cancer, to the heights where science and policy soar side by side.

Every now and then, other patient groups inquire about adopting Project LEAD, even licensing it.
They get a firm no.
Even ripping off a syllabus and the slides, you will not end up with anything even close to Project LEAD.
To produce Project LEAD, you need skepticism and a cadre of tough, dedicated people who safeguard it.
“You need a culture,” says Love.

Graduates of Project LEAD get to do some amazing stuff.
Working alongside scientists, they review grants for the National Institutes of Health and the Department of Defense.
They sit on institutional review boards that monitor ethics of clinical trials.
They sit on data and safety-monitoring boards that decide whether clinical trials have met their objectives and whether they should continue.
Also, they offer guidance to pharmaceutical companies as they develop drugs and to the Food and Drug Administration as it decides whether drugs merit marketing approval.

Sure, patients who haven’t done Project LEAD do some of those things, too, but a Project LEAD credential means that you are dealing with someone who understands the fundamentals.
Some Project LEAD graduates are dazzling.
For example, Ransohoff can’t get himself to delete his correspondence with Share’s Helen Schiff.

“I’ve got a bunch of detailed e-mails going back five years, dealing with a broad range of topics,” he said recently.
“The detail of her interest is really striking for a civilian.
Most physicians don’t know half of what she knows about clinical-research methodology and clinical trials.”

*

ALL
NBCC events begin with a moment of silence in honor of an advocate who died of breast cancer.
Every board meeting, every advocacy conference, every scientific-consensus meeting, and every Project LEAD starts in exactly the same way.
A picture of an advocate is projected on a screen and her story is told.

The 2010 Project LEAD in La Jolla began with Fran Visco’s tribute to Lilla Romeo.
Tony knew it was coming.
He had been asked in advance, and he provided the picture.

It was a good thing, though it hurt like hell, or because it did.
“Lilla had died two months before,” Tony recalls.
“I was not at my most raw state, but not that far along in the mourning process.”

Then, breaking the silence, Fran simply moved on to her introductory talk.
The juxtaposition was powerful.

“Here was the emotion—here is the people connection that advocacy is all about—and then Fran is off and running, talking about evidence-based medicine,” Tony says.
“We recognize that this is about lives, about people who struggle with that disease.
Now let’s get down to business.
Not losing sight of the personal element, but also getting into hard thinking that’s required for advocacy to be effective.”

Fran often reminds advocates that true advocacy is about learning from your experience and rising above it.
The science at Project LEAD was stimulating enough, but Tony experienced a second track also, which ran privately, entirely inside his skull.
It seemed to be akin to a health-care economics conference for one attendee, a rigorous examination of decisions that had been made in Lilla’s care.

The hemoglobin-building drugs were an information problem.
Lilla got most of these drugs without the benefit of being asked or a discussion of pros and cons.
When the questionnaire was finally administered, it was disingenuous, an obfuscation.
And, of course, no one bothered to mention the system of incentives at play.
The system was manipulating information to protect itself.
“I wish we had known that, because I wonder whether it stimulated the tumor growth, and I know Lilla wondered about that,” Tony said.