In 1974, Roxanne Tremblay was 29 years old, a frazzled single mom supporting a young daughter. So when the pain came on, it presented a real problem.
The pain was sharp, a constant ache in her lower abdomen that doubled her over at work and left her breathless. Tremblay couldn’t risk losing her job, so she went to see her gynecologist, one of the most popular doctors in her small working-class river city of Lewiston, Maine.
This man had always impressed Tremblay and made her feel at ease, so when he announced that she would need an operation, she didn’t really question it. Tremblay never planned to get married or have kids again. She was done with all that.
Three weeks after her first appointment, Tremblay went into the hospital and had her uterus and ovaries removed – a total hysterectomy. After the operation, her doctor explained what he had found.
“It was what he called the seed of cancer,” says Tremblay. “It wasn’t cancer, but it had the potential of developing into it.”
Twenty-nine years old is pretty young for a hysterectomy. But in the mid-’70s in Lewiston, Maine, lots of women were getting them. Tremblay says she herself knew a bunch.
“My boss that I had when I worked at Kmart, she had one shortly after I did,” Tremblay recalls. “One of my friends that I’ve had since I was 6 years old, she had one. Just about anyone you talked to would say ‘Oh I had a hysterectomy!’ ‘Oh yeah I did – so and so did mine!’ There was a lot of them. I do remember that.”
Why So Many?
The question, of course, is, why were so many women in the city of Lewiston having hysterectomies?
At the time, this was not a question of interest to most Lewiston women. They were busy with lives and work and families. But there was someone, a health researcher in nearby Vermont, who was very interested in the Lewiston hysterectomies. A man who was interested in operations and health statistics of every kind.
His name was Jack Wennberg, and by the mid-1970s, he had compiled data about health care practices in cities all over the state of Maine, including data showing that in the city of Lewiston, an unusually large number of women were having hysterectomies. He projected that if the rate of surgeries continued in Lewiston, 70 percent of its women would have a hysterectomy by age 70.
Wennberg discovered Lewiston’s high hysterectomy rate, but he really did much more than that. Over the past 40 years, he has completely transformed our understanding of what’s going on in the U.S. health care system. His research on health practices eventually expanded. He started in Vermont, then moved to Maine, until eventually he studied communities throughout America. In the process, Wennberg led us to a clearer understanding of what doctors and hospitals are doing with their patients all across the United States.
So if you’re looking for a better understanding of what drives the cost of health care in this country, health researcher Wennberg is a pretty good place to start.
Though Wennberg is now a certified guru – a man whose insights underlie many of the arguments you currently hear about health reform – Wennberg didn’t set out to revolutionize our understanding of health care. It was an accident.
It all started in the late ’60s when Wennberg, fresh out of a residency at Johns Hopkins Hospital, was given a grant from the federal government to help improve the health care system in the state of Vermont. The idea behind the federal grant was that in states all around the country, isolated rural communities weren’t benefiting from many of the new, modern medical treatments that had been developed. Wennberg’s job was to fix this problem in Vermont. He was supposed to help bring high-tech medicine to poor rural communities.
En route to this perfectly laudable goal, however, Wennberg made a somewhat crazy decision. He figured that if he was going to improve health care in Vermont, he should really understand what was going on in terms of medical practice on the ground. And so he decided to collect information about every medical transaction of every person in every town in the whole state.
“We needed to know what was going on in home health agencies, what was going on in nursing homes, hospitals, doctors offices,” Wennberg says. “And for each patient, what their diagnosis was, what their treatment was, how much money was spent, and what the outcomes were in as far as we could measure them.”
Searching For Explanations
To collect these records, Wennberg hired researchers, people dubbed “the pit crew” who year after year were sent out to medical record rooms to collect records. It was a truly massive undertaking to gather every medical transaction in the state of Vermont. It took two years of road trips just to collect the records for 1969.
But once he had the information, Wennberg began to slice it and dice it in all kinds of ways. And what became clear almost immediately was that something was terribly off. “As soon as we set out to do the analyses we began to see these extraordinary differences,” Wennberg says.
Sitting at a table with most of the medical transactions in the state of Vermont listed before him, Wennberg was able to see just how bizarre the distribution of care was. People in one town would get their hemorrhoids removed five times more often than people in another town only 30 miles away. Ditto with mastectomies, prostate operations, back surgery.
This was even the case in Wennberg’s own town.
“We lived right on the boundary between Stowe and Waterbury Center, Vt.,” Wennberg says. “And if my kids had been going to the school system in Stowe, they would have had a 75 percent chance of having their tonsils out. If they had gone to the Waterbury School – where they actually did – it was about 20 percent.”
Why the differences?
There were two possible explanations.
The first explanation was that doctor behavior was somehow to blame. The second explanation was that it was the patients; that people in some areas were just much sicker than people in other areas, or maybe just wanted more services for some reason.
This was the next problem Wennberg needed to solve.
Carol Bradford lives in a small house on a tree-lined Lewiston street. She’s a church lady, the kind of person who takes in strays and carries food to elderly neighbors.
Bradford, like Roxanne Tremblay, had a hysterectomy in the mid-’70s. Fibroids prompted her doctor to remove her uterus, and, like Tremblay, Bradford is perfectly happy with the result. Also like Tremblay, Bradford was not at all surprised to learn just how many women were undergoing hysterectomy operations in the 1970s. To her it made perfect sense. Lewiston, she explained one morning over coffee, is mostly Catholic.
“Some women were having too many children,” she confided. “You know, there are families here with 10, 12 children. It’s a possibility that women came to the point where they just really couldn’t deal with any more children and were begging the doctors to do something about it.”
“You know,” Bradford said with a nod, “that’s my personal opinion.”
But it was not just Bradford’s opinion. It was everyone’s opinion when Wennberg first made his discovery.
In fact most people continue to assume that when a patient goes into a doctor’s office, the doctor is simply responding to sickness in the body, or the needs and concerns a patient has.
But in his work in Maine and Vermont, Wennberg demonstrated that it’s actually a lot more complicated than this. The women of Lewiston, Maine, weren’t having more hysterectomies because more of them were Catholic, or because more were sick. After his initial work in Vermont and Maine, Wennberg embarked on this enormous study of patients, which showed that in terms of sickness and demographics, the populations of different communities in Maine and Vermont were actually remarkably homogenous.
This could mean just one thing, according to Wennberg. “Because we could easily see that it wasn’t that patients were different between regions, so it wasn’t the illness that was driving this, this must be coming from the provider side.”
His insight: It was doctors, not patients, who drove medical consumption, and all kinds of things influenced the decisions a doctor makes when a patient enters his office. Sickness and patient preference play an important role, but a much smaller role than patients and the health care community had originally thought.
The Maine Experiment
The medical community’s response to Wennberg was pretty muted in the 1970s when he first began to publish his work about the underlying drivers of medical care. Some were supportive, but most in the medical community either didn’t believe him or simply didn’t grasp the significance of the work.
But in the state of Maine Wennberg actually found a well-placed fan. Dr. Daniel Hanley was the executive director of the Maine Medical Association and a towering figure in Maine medicine. Sometime in the mid-’70s Hanley was introduced to Wennberg, and according to Gordon Smith, current head of the MMA, Hanley was absolutely smitten.
“Jack really found a follower in Dr. Hanley,” Smith told me. “Hanley was fascinated with Wennberg’s work. He felt like Wennberg was making an important contribution to knowledge and that doctors needed to do something about it.”
What the doctors needed to do, Hanley decided, was figure out why the strange variations in medical care were taking place. The best way to do that, he figured, was to get all the doctors in Maine to sit down together on a regular basis, look at Maine city by city, and then hash out why the care they were giving was so different. The doctors would speak directly and honestly about their decision-making.
Bob Keller is a back doctor in Maine who worked on this project, and he says there was only one problem with Hanley’s plan: the doctors themselves. If the data were right then at least some of the doctors in the room were doing something wrong by providing care that wasn’t needed. That didn’t sit well with the doctors.
“No. 1, they were insulted,” says Keller. “They were angry because their judgment was being challenged, and that was not allowed. And in some cases they just didn’t believe it, so they would try to find holes in the data. One of the classics, they’d say, ‘Oh we have more workers compensation here, we have more heavy industry!’ But we were able to work through most of those things and demonstrate that wasn’t the case. Then they would say, ‘ Our population is older! More of them need prostatectomies!’ And we’d say, ‘Well, we adjust for age, so that’s not an argument anymore.’ And some doctors never could deal with that, and they would leave the study groups. They just said, ‘This is baloney; we’re not going along with this.’ ”
But in time, says Keller, most of the doctors in Maine did warm up to these ideas. “They began to accept the data, began to accept that indeed, different physicians were using different thought processes or decision-making processes in dealing with patients.”
And so in the state of Maine began this incredible experiment. Four or five times a year each medical specialty would gather together for a kind Talmudic dissection of doctor choice conducted by the doctors themselves. They wanted to look at all the geographical differences, figure out why they existed, and then try to bring their medical decisions in line with one another. They figured that by doing this they could eliminate excess care.
Smith took part in the meetings, and he said that inevitably there was one thing above all that doctors believed shaped decisions: “The way you were trained. Maybe you were at a particular training program that does things a certain way and you bring that back to your community.”
Smith says to understand how their training shaped their decisions, the doctors who gathered would list on a white board all the criteria they had learned in school for doing this or that procedure. But what they found, says Keller, the back doctor who helped put the groups together, was that most of the time the doctors all seemed to agree.
“In a meeting they’d all say, ‘Absolutely I agree; you need to have a certain physical finding and if you didn’t have it you wouldn’t do it,’ ” Keller says.
“Well, that might be what they agreed on, but in fact when you were able to use data later you would find that it didn’t really work that way. That’s the criteria and the standard got tilted, sometimes pretty significantly.”
Somehow in the privacy of their own offices the doctors still enacted the agreed-on criteria differently. Why?
One reason some doctors mentioned was fear of lawsuits; some worried that if they didn’t do every possible thing they might get sued. Another reason was temperament – some doctors were clearly just more eager to take action than others.
Then there was the role of local medical culture. For example, even though it didn’t make sense and wasted a lot of time and money, pediatricians in some communities felt they absolutely positively had to send even mildly sick kids to the hospital.
“Families in small Maine communities were used to the fact that if their kid had a temperature of 102 and was vomiting, that kid was going into the hospital,” says Keller. “They’d been doing it for years, so they’d be aghast if they took little Tommy down and he had a temperature of 102 and the doctor said, ‘Well, go home and take this.’ Nobody did that!”
It was probably safer and better all around not to put the kids in the hospital, and the doctors knew this. But doctors, like the rest of us, are people, and therefore are subject to subtle influences.
For instance, it turns out that if you increase the number of doctors in an area, chances are that the use of medical services will rise. If there’s one doctor in a town with 100 patients, then he’ll schedule your heart checkups for once every six months, but if another doctor comes to town – and now the first doctor has 50 patients – the doctor will just schedule your heart checkups for once every three months. There’s a very simple reason why, says Frank Read, an eye specialist who participated in the doctor groups.
“I don’t want to be sitting on my thumbs all the time – I want to be busy. And that may unconsciously loosen my criteria for doing a procedure.”
Which brings us finally to the subject that incredibly was never directly discussed during the nearly 20 years the doctors met: money. Specifically, the way money affects medical decision-making.
Keller explained that this subject was completely verboten.
“It would have been a show stopper. It would have gone right to the question of greed, and you’re not going to keep a doctor at the table if you say that he’s greedy.”
Talking to doctors about money is difficult. It’s uncomfortable both for patients and for doctors to think that this most important and intimate service could be contaminated. But the truth is the decisions made by your physician when you enter his office are profoundly influenced by the way that doctors get paid in this country. “That’s just common sense. That’s human nature,” says Smith of the Maine Medical Association. “The payment system is an important influence.”
Most of the doctors in this country are not on a salary but are paid basically like pieceworkers in a clothing factory. This is called “fee for service,” and the way it affects doctor behavior is clear.
“If you pay people more, the more things they do, they’re going to do more things,” says Smith.
The U.S. health care payment system rewards doctors for taking action and doing procedures. This reality is so powerful that it hasn’t just changed the individual behavior of doctors. Keller says that the specialties themselves have changed, bending like flowers to the sun, moving toward the source of heat.
Consider dermatology. “In the old days, dermatologists just did medical visits,” says Keller. “They looked at people, they looked at their skin lesions, they advised them. Now they’re all doing Botox, they’re doing various procedures. So dermatologists are making very high incomes now, but they are doing procedures; that’s where they’re getting it. The specialties changed because it’s a procedural-driven world that we’re in.”
And the more complicated the procedure the higher the payment.
This is perfectly logical but has an unintended effect on care. Keller points to his own specialty of back doctor. He says one of the most frequent operations these days is a highly complicated procedure called an instrumented spinal fusion. When a patient has degenerative disk disease, the doctor inserts metal rods to stabilize the spine.
Keller says in the old days, doctors used a much less complicated and safer operation. But the more complicated one had a clear advantage: “The surgeons could charge more because they could do these complicated procedures. So they were putting the screws in and they billed for putting the screws in. They were putting the plates in, and they billed for putting the plates in. So you had this whole new high-tech procedure, and it was enormously attractive to spine surgeons and it literally took off in this country, at the same time as most good spine surgeons will admit that they had no research to support what they were doing.”
In fact, says Keller, one high-quality study demonstrated it wasn’t so positive. “It showed that it isn’t so great, actually as people thought it was,” he says. “And they also showed, interestingly enough, that the old-fashioned non-instrumented fusion was as successful as the instrumented fusion – which was a real blow.”
More Is Not Better
And this, in miniature, is one of the big problems with the way that the current system is set up. It’s a problem some call “more is not better.”
Doctors exist in a fee-for-service system that encourages – and really because of the threat of malpractice and having to battle insurance companies – in some ways actually forces them to do more. More surgery. More tests. More of everything.
And while most Americans just assume that more care is good, it turns out that more isn’t always better for patients.
Natasha Saint Amand is 25 and wears her brown hair curled into loose ringlets. Her makeup is picture perfect. Amand has spent her whole life in the Lewiston area, but like most people who live there, had no idea that the area had an unusually high rate of back surgery. The rate has varied over the years, but between 2003 and 2007, the rate of lower-back fusions in Lewiston was around 107 percent higher than in the rest of Maine.
Five years ago, Saint Amand was in a car accident and after that had serious pain in her back several times a week. Her doctor strongly recommended a fusion, and so she got one. And then another. And then another. And then neck surgery.
Now Saint Amand is in pain every hour of every day. She can’t bend at the waist or at her knees. Not for the original sickness, she says, but from the cure. She is disabled, and she says really understanding the reality of that took a long time.
“I think it really hit me after after my third back surgery and after I had my neck surgery,” Saint Amand says. “It really hit me that, wow, there’s really not much I can do. My leg is all nerve-damaged. My lower back is nerve-damaged. I have nerve damage in my left arm. There’s really not much left that doesn’t hurt.”
It would be easy to dismiss what happened to Saint Amand as poor-quality medical care; surgeries that failed because they were badly done. But it’s probably more complicated than that.
A couple of years ago Keller and some colleagues did an elegant study of one kind of back surgery in Maine, a procedure called discectomy. Keller found communities in Maine that had high rates of this surgery, communities with low rates, and communities that were somewhere in the middle. Then he followed patients who had had surgery in those communities over a five-year period to see how they fared. Keller says the conclusion was undeniable.
“In the high rate of surgery overall, the patient outcomes were the least good of those three categories. In the middle rates, the outcomes of the patients were in the middle. And in the low-rate areas – less frequent operations per capita – the outcomes were the best.”
The reason that areas with more back surgery did worse, Keller says, is that doctors in those areas were operating on people whose issues were less severe; that is, patients who might not have been good candidates for an operation. So the problems associated with the surgery probably outweighed the problems of their actual sickness. For them, more wasn’t better.
But this essential dynamic – that more isn’t better – applies to a lot more than just back surgery.
In 2003, there was an enormous landmark study published by a Jack Wennberg protege named Elliott Fisher, who works at Dartmouth College. Fisher compared Medicare recipients with similar levels of sickness in areas throughout the whole United States. Fisher looked at places where elderly people used relatively few health care services and compared them with places where elderly people used a lot of health care services.
“The patients in the high-spending regions were getting about 60 percent more care; 60 percent more days in the hospital; twice as many specialist visits,” Fisher says. “And yet when we followed patients for up to five years, if you lived in one of these higher-intensity communities, your survival [rate] was certainly no better, and in many cases a little bit worse.”
This is probably because of a something called fragmentation of care. In high-use areas, it’s often the case that many different doctors play a role in the care of a patient; many specialists are responsible for overseeing only a small part of the person. This increases the amount of treatments, tests and hospitalizations that people get, and exposes people to more risk of harm from medical error and side effects.
For most Americans this is an incredibly difficult idea to accept: It’s hard to understand that more care isn’t necessarily better for you.
But study after study has borne out the truth of this completely anti-intuitive conclusion. In fact, Fisher and other researchers estimate that almost one-third of the care given in our country today is that kind of care – care that isn’t really helping people.
The United States spends more than $2 trillion on health care every year. So the cost of that 30 percent unnecessary care annually? $660 billion.