How We Do Harm (35 page)

As a senior resident, Ransohoff did a rotation in cardiology at Tufts University.
By chance, he ran into a cardiologist named Steve Pauker, who was using the quantitative methods of decision analysis, developed at Harvard Business School and MIT, to diagram choices and consequences of medical interventions.
Pauker would become the founder of this part of the field of quantitative medical decision-making.
Working with Pauker on understanding medical decision-making was a kidney doctor named Jerome Kassirer, who would later become the editor of
The New England Journal of Medicine.
Ransohoff was captivated by this work.
It amounted to a dynamic model of decisions his father was describing to him as the two made house calls in Cincinnati.

The discipline of modeling clinical judgment was starting to take shape.
However, Ransohoff’s medical education was now complete and destiny was calling: he was expected to join the general medicine practice he was conceived and raised to inherit.
The problem was, he didn’t want to.

Instead, he wanted to pursue his interest in breaking down clinical judgment.
Had his mother laid on the guilt, he would likely have joined the practice.
Ransohoff realized that he had idealized the relationship between his father and grandfather.
True, they had a wonderful practice.
They treated university presidents and captains of industry, but in pursuit of social justice they structured the practice to include 15 percent of people who couldn’t pay.
They were on call twenty-four hours a day, seven days a week, and genuinely thought it was a privilege to work so hard.
They were good doctors, but they fought in passive-aggressive ways.
On rare occasions when Ransohoff’s father left town, his grandfather changed the medications of his patients.

The grandfather, Hiram Weiss, practiced till age ninety and died at age ninety-three.
He got an unforgettable eulogy at Hebrew Union College, where he had been the chairman of the board of governors: “No self-respecting Jew in Cincinnati would even think of dying without first consulting Dr.
Weiss.”

Rather than cause a confrontation—something they studiously avoided in their relationship—David’s father and grandfather accepted the myth that David would be joining them in the practice later.

*

IN
1977, Ransohoff enrolled in the Robert Wood Johnson Clinical Scholars Program at Yale.
The program, in its second year, was run by an internist named Alvan R.
Feinstein, a wiry, bespectacled, curmudgeonly man.
In a photo with Ransohoff’s class, Feinstein wears a conservative suit and an Ivy-prep striped tie.
He took pride in dressing English while thinking Yiddish.
Taking nothing on faith and examining everything, Feinstein was a lunatic-fringe teacher who was brilliant, articulate, and totally ruthless about evidence.
He was also famous for Socratic avalanches of questions that would make students and fellows cry.

Ransohoff was happy to learn that Feinstein had written a book on clinical judgment, David’s area of interest.

None of the eight fellows in Ransohoff’s group knew whether they would be able to find jobs, as their chosen field—quantitative clinical epidemiology—was totally out of the mainstream.
It was led by three titans: Feinstein, Thomas Chalmers, and Dave Sackett.

Feinstein was writing what would become the fundamental textbook on quantitative clinical epidemiology.
Also, he was politicking with the major academic medicine societies to allow clinical epidemiology papers to be presented at their annual meetings, which took place exclusively in May and exclusively in Atlantic City.

Feinstein’s vision was to create a science that would generate evidence that can be trusted in making decisions about treating patients
and
making national health policy.
(Similar principles should apply whether you are addressing health problems of an individual patient or setting policy for managing a large population of similar patients.)

Feinstein had a love/hate relationship with laboratory science.
Rat-turd grinders
was his term to describe lab scientists.
He impregnated that moniker with a surprising amount of complexity.
On one level, he was contemptuous of people who studied physiology for its own sake, without regard for application, often with the primary goal of advancing their careers.
Important insights were to be gathered by the methods that laboratory people used, but a decision about what’s right or wrong for a patient can’t be based on lab findings.
In a lab, you can develop a test that detects cancer early, but data that would show whether early detection leads to more good than harm has to come from a totally separate field—clinical epidemiology, where you study human beings, not animals or cell lines.
Ultimately, you have to perform a clinical trial or some other kind of study in people.

Yet, Feinstein had immense respect for basic scientists because they were rigorous.
“We must be like the rat-turd grinders,” he would say, and the fellows knew what he meant.

The real world is a surprising place, where diagnosis is not always a benefit.
In one of his early classic papers, Dave Sackett found that if you identified people with hypertension, you would trigger an increase in absenteeism in people labeled as hypertensive.
If you are labeling people, they may end up missing work, and it’s not because the disease is giving them some sort of headaches or heart failure.
Most likely, they are missing work because you told them that they are ticking time bombs, and they are scared.
The label has a downside.

*

SOME
guys remember the exact spot where the girl of their dreams said yes.
Others remember where they were when they got the news that President Kennedy was shot.
Ransohoff remembers those landmark events with proper clarity, but he also remembers the exact seat at the Yale medical library where he stumbled upon Dave Sackett’s article on rules of evidence for screening.
The paper was published in a British medical journal,
The Lancet.
The paper argued that the burden of disease has to be high enough to justify an intervention, that the disease left untreated leads to a bad outcome, and that intervention leads to a better outcome.

“It’s simple common sense, but I remember being thrilled by thinking, my God, there are people who are thinking about when it is okay to do screening, when it is okay to do an intervention on a patient,” Ransohoff recalls.

Feinstein suggested a topic for Ransohoff’s fellowship paper: why not perform an autopsy of a cancer detection test called carcinoembryonic antigen (CEA).
This blood test had been purported to be useful in early detection of colon cancer—with almost 100 percent sensitivity and specificity—but was ultimately shown to be no good at detection at all.
Ransohoff’s paper, which set forth the methods for studying diagnostic tests, was published in
The New England Journal of Medicine
in 1978.

*

TO
the extent possible, the US government makes medical decisions based on science.
To get a research grant from the NIH, a scientist has to submit her proposal to a jury of colleagues.
To get a drug approved for sale, a company has to provide extensive data from clinical trials.

In 1984, Congress formed the US Preventive Services Task Force to make evidence-based recommendations on a wide range of preventive services.
For the next decade, the task force was pretty much the only voice in screening.

The task force spent the first years of its existence deconstructing the standard physical exam, piece by piece.
In those days, people did yearly chest X-rays, yearly urinalysis, yearly EKGs.
After the task force was done with assessment of whether the tests used were doing more harm than good, the institution of the annual exam was gone.

Each evaluation includes a comprehensive review of evidence—sometimes hundreds of papers—and in some cases lately reviews have included computer modeling.

Now, the task force consists of sixteen members who represent internal medicine, family medicine, behavioral medicine, pediatrics, obstetrics/gynecology, and nursing.
Subspecialties that perform tests aren’t overlooked.
They aren’t invited to actually make guidelines because their commercial motivation to sell tests and procedures could cloud their judgment.
They are consulted during the process.

The best my friend Ralph could do was listen to conflicting parties or make his own guess.
This couldn’t possibly serve him well, because the whole purpose of evidence-based medicine is to rise above guesswork.
Had Ralph checked the USPSTF recommendations instead of stepping behind a curtain at a shopping mall, he would have found that, at that time, the task force had concluded that there is no evidence to justify routine screening of men under seventy-five, and screening for men over seventy-five is not recommended.

Recently, I heard an interesting word for people diagnosed with cancers of unknown significance:
previvor
, as opposed to
survivor.
We have to accept the fact that we are harming many of these people.

Ralph was past seventy when PSA destroyed his life.
When you start screening at forty, the potential for harm is even greater.

Chapter 24

Saying “Enough!”

AS HER BREAST
CANCER PROGRESSES
and death comes closer, Lilla Romeo suggests to her husband, Tony, that he would make a good patient advocate.
She is almost subtle about this, except she mentions it a few times.

Tony says he wants to think about it.
He is not ready.
“We don’t have to do the same stuff,” he says.
“I can be there to support you and give you my input every now and then … I’ll tag along and maybe get involved sometimes.”

He knows that he is resisting because a good-bye is lurking beneath Lilla’s recommendation.
She is not being explicit, but it’s hard to miss that she is thinking that helping others would help Tony with the search for meaning that will begin in earnest the instant she is gone.

*

AS
options dwindle, Lilla’s doctor suggests Avastin.
Avastin is not a chemotherapy.
It’s a biologic agent that seeks to starve the tumor by blocking a protein called VEGF, which is overproduced by cancer cells.
VEGF—which stands for vascular endothelial growth factor—is important for the formation of blood vessels.
By blocking this supply line, Avastin can deprive the tumor of oxygen it needs to survive.

Avastin, produced by Genentech Inc., an American subsidiary of the Swiss drug maker Roche, is one of the most expensive drugs on the market.
In breast cancer, the cost of the drug for the full course of treatment was close to $100,000.
That’s based on prices that would have been charged to Medicare.
The hospital would have charged the insurer a higher price.
Avastin is usually used concurrently with a chemotherapy drug called Taxol, generic name paclitaxel, which is derived from the bark of yew trees.

Paclitaxel is a generic and therefore it’s cheap.
However, Lilla has had a good response to paclitaxel in a more expensive form, Abraxane.
Unlike regular paclitaxel, Abraxane doesn’t use solvents and doesn’t require additional medications to prepare the patient to receive it.
The downside is cost.

Effectiveness of this treatment choice is uncertain.
Avastin at the time has an “accelerated approval” for breast cancer from the FDA.
This is, essentially, a provisional approval, which requires additional studies to determine effectiveness.
The randomized trials that had been done up until then were done in a population very different from Lilla.
Avastin was found to be effective in patients who had just been diagnosed with metastatic disease, front line.
That’s where Lilla was in 2000, and in the decade that has elapsed, she has received about a dozen different treatment regimens.

Extrapolating these trial findings to her is a leap of faith.
Also, the value of benefits that accrue even to front-line patients is the subject of controversy.
The debate is, in part, philosophical.
Avastin isn’t shown to increase survival.
Women who get it live as long as women who didn’t.
However, one initial study shows that the drug pushes back the time of disease progression.
Lilla’s disease is already in relentless progression.
(Subsequent studies would demonstrate a smaller delay in progression.
At this writing, the FDA is seeking to revoke the drug’s accelerated approval in breast cancer.)
Of course, Lilla and Tony understand all this.
They ran the same literature searches I would have run.
“Lilla’s doctor had been thinking about it before,” Tony says.
“If you look at the biological properties, it wasn’t totally irrational to think Avastin could possibly work.”

Cost doesn’t figure in the decision, Tony says.
Certainly, the insurance company pays, but even if it balked, the Romeos would have used their own money.

“I can’t deny, you get to the point where there are no other options,” Tony says.
“We thought it’s worth a shot.
We did it with the understanding that we were really rolling the dice.
But the alternative for us at that point was nothing.
And we just were not ready for that yet.”

Avastin fails Lilla.

As the end nears, the Romeos no longer fear death.
They fear death from brain metastases, the crushing blows to identity that precede all else going dark.
Lilla has whole-brain radiation to keep brain mets in check long enough for her to die of something else.
Luckily, everything else goes first, the lungs, the liver, the whole system.

Lilla dies on June 9, 2010, at sixty-three.

*

LOSS
is a strange animal in 2010.
Traces of Lilla, breathing, speaking, are on the Internet.
Tony clicks on a seminar at NYU and watches her speak.
He does this only once.

He puts the Share men’s support group on hiatus.
“I don’t want to be depressing to the guys,” he explains.
“It’s one thing to be involved in the community, but to facilitate the group of guys who are living with women who are metastatic or in serious stages of disease, I am probably not as reassuring a force as I was, so I am going to opt out of it for a while.
It would be tough for me personally, and I certainly would not be an inspiration.”
It’s a blessing that he has a job.
(He runs an Internet start-up company.)
Not quite three weeks after Lilla’s death, her friends at Share put together a tribute to her and ask Tony whether he would be up to coming.
He does.
After the meeting, Musa Mayer and Amy Bonoff ask Tony to join them for a chat in the corner.
Their suggestion seems logical enough: would Tony be willing to step into Lilla’s advisory role at the NCI brain-metastases center?